#pc-gaming

1 messages · Page 74 of 1

full flint
#

its for those that want the best, period, and makes no sense for basically anyone

smoky flame
#

The Division 2 is an AMD featured title btw so I don't think it would be favouring Nvidia hardware but at least AMD don't go adding features that wreck the competition's performance

full flint
#

performance favors nvidia,slightly is what i was saying, the 6800XT falls a couple fps shy of the 3080 there, where it beats it in most other titles

#

AIB cards with an extra 300+ mhz sustained clockspeed? different story of course

#

ampere does really badly at 1440p, compared to how it does at 4k

smoky flame
#

all I can say is I'm glad 1440p and 4K are now beyond "playable" at ultra settings

full flint
#

1440p has been beyond playable on ultra for awhile

#

4k is only starting to get there

#

5700XT/1080Ti/2070 super has been a viable 1440p ultra performance tier for a long while, its only starting to slip recently

#

btw, subtract like, 2 - 4% from these figures for a 5600X being used instead of the 5900X

#

minor, not worth caring about

smoky flame
#

and honestly, I'll take a hard pass on ray tracing for some time yet so it won't influence my decision on what card I go for. I can wait for as long as it takes for ray tracing to mature to the point where the impact it has on performance eventually becomes no heavier than some of the heavy hitting graphics settings today, ray tracing, for comparison being on a whole different league even to these settings

#

I think it is pretty evident what card I am now going for, as if it wasn't already clear enough though. Really taken a liking to AMD GPUs. Something that up until July this year I wouldn't have said

full flint
#

don't let LieutenantAmerica see you say that, he'll say "bu..bu..but DLSS" for the grand total of 15 games that support it, and ignore the fact that amd announced a DLSS equivalent that will be fully open source and supported on the consoles, meaning it'll win in terms of market adoption just like freesync did

#

1440p ultra 240fps FH4 is INSANE

smoky flame
#

yeah, big ouch there!

full flint
#

240hz on ultra....

smoky flame
#

ngl, this has me seriously considering 4K tbh

little dome
#

jesus, thats an excessive amount of MLCC caps you got there on that 6800XT amd, guess you don't want instability due to crap capacitors and an incorrect voltage curve 😂
@full flint poscaps are worse for this

full flint
little dome
#

lol i missed this one 😂
@full flint damn, thats a big difference from nvidias top card

#

slightly less impressive 153 on ultra at 4k
@full flint 153 is still impressive

full flint
#

i mean, i guess its all relative, 153 on ultra at 4k is still bonkers, and 167 for the 6900XT

#

i know poscaps are far worse for it Squirky, but i didn't quite expect amd to use that many mlcc caps, its overkill, in every possible way, totally unneccesary, but still awesome

smoky flame
#

ok, The Division 2 at 4K is still horrid. Maybe I'll wait on 4K a little longer. 1440p is plenty for now I think 🙂

full flint
#

then again amd reference designs are ALWAYS overkill... you could run two vega64s on the vega64 reference vrm, with room to spare

#

75fps isn't horrid SUB

#

and your AIB model of choice will be between 5 and 20% faster than these results depending on exactly how high they manage to clock it

smoky flame
#

that's about what I get now, maybe a little lower on my current system. I'd rather see more FPS than I'm getting right now

full flint
smoky flame
#

maybe I'll think about 4K after RDNA 3 or RTX 40 series launches. Surely by then 4K will be how 1440p is now with these cards

full flint
#

6 core zen3 is faster than 8 core zen2, 8 core zen3 is faster than 12 core zen2, and faster than the 10900k, even in multithreaded work

smoky flame
#

I do wonder if it is worth me maybe considering going up a step with my CPU upgrade or do you think the 5600X will remain fine for some time yet just gaming?

full flint
#

Considering Zen2 is ~40% faster than Zen+ (never mind original Ryzen), Zen3 is in effect 2x faster than Zen+ – a 2x (96%) improvement over just 2 generations, while core counts remained the same (unlike Intel that just increased core counts). Also consider you can now get a 16-core/32-threads AM4 CPU (which originally only went up to 6-cores), it is like having a 32-core Ryzen in the same AM4 slot – a 5.3x increase in overall performance. heh, from the start of am4, until now, they've doubled performance at the same core count, and at the maximum core count there's a 5.3x performance increase vs zen1

#

in 2 generations, that's madness

#

the extra $250 for the 12 core will give you low single digit gains in games, and quite literally straight up double performance everywhere else

#

that's your call, but purely for gaming? don't bother. there's only a 4 - 6% difference between a 3600 and a 3900X in gaming too

smoky flame
#

ok, 5600X I stick with 🙂

full flint
#

what else are you going to be doing?

smoky flame
#

purely just gaming really

full flint
#

then yeah, 5600X is fine for a few years

full flint
#

hey Dark, did you see the new benchmark data for the RDNA2 stuff? and the zen3 review i posted above? 😄

#

😂

#

like the driver nvidia had to release to downclock ampere to keep it from crashing constantly?

#

😄

rough ice
#

guys help i cant find the reason as to why my Xbox Series X is slower than my PS5

#

they said it would be faster

full flint
#

first off, you're in the wrong channel, you want #general-chat , second, you're impersonating linus using him as your pfp, and third, neither machine is out yet so why you be trolling yo?

analog sapphire
#

bc it's made out of your imagination, which is malfunctioning

rough ice
#

damn ok @analog sapphire

devout remnant
#

Looking forward to the 6800XT & 6900 XT 🙌

dry wolf
#

Hey guys can someone help me

obsidian bobcatBOT
#

Please refer to #233805809346150400 for support related issues. Do note that this is community sourced support and there is no official support here on discord. As such, people will assist if/when they are able to but if you have issues regarding your account and/or billing, please reach out to official support for assistance.

analog sapphire
#

oh that was like 2 hrs ago whoops

full flint
#

😂 i do that all the time

full flint
#

at 1440p, the RX6800 is only 5.22% slower than the RTX 3080 on average, the 6800XT is faster than the 3090, and the 6900XT is faster still, on average

#

if this holds in a larger suite, uh yeah... nvidia's in trouble, i mean, they're in trouble anyway, the worst these things are going to be is on par, if you slant the suite towards nvidia on purpose

#

if these things have any oc headroom at all, or if aib cards come in at the clockspeeds that got leaked last week..... all three rdna2 cards will beat nvidia's flagship at 1440p, instead of just two of them 😂

tender bloom
#

couple questions guys, appreciate any help you can give

  1. what should my gpu mhz boost be in afterburner (2070 super mobile)
  2. crashes related to undervolting wont hurt my cpu, right?
  3. my cpu seemed stable at -89.8mv, should i try to work my way back up (but not all the way up, cause it was crashing)
  4. should i make custom fan curves in silent option fan control for my cpu and gpu, (i7 10750H and again, rtx 2070 super mobile respectively), and what should they be?
    thanks
fervent nimbus
#

should i get a gtx 1660

full flint
#

why not just get a 3080

surreal moss
#

Yeah why not ez pz

smoky flame
#

Lieutenant, I'm still waiting on my 3080 that I ordered back on launch day (September 17th) and judging by how poor things seem to be going for this GPU I don't expect to receive it until early 2021. Nvidia, partners and retailers might only just manage to fulfil all existing backorders by the end of Q1 2021 too

full flint
#

46 fps beats the 3080, if you're curious, 1440p result this guy got is heavily cpu bound as its only a 3500X, which is such a weird cpu pairing for this lol

#

he did 1440p too, result this guy got is heavily cpu bound as its only a 3500X, which is such a weird cpu pairing for this lol

smoky flame
#

SotTR doesn't use DLSS 2.0/2.1 right? Not that it can be used in this instance but it could be used to save the 3080 if it is DLSS 2.0. Even at that I'm really liking the sounds of AMD's DLSS rival that will be open source and can be enabled in any game (not just those with ray tracing) while DLSS is limited to mostly games with ray tracing. A few exceptions and fewer still that have yet to update to DLSS 2.0/2.1

#

I'm guessing AMD's version would just be an option that can be toggled on or off within Radeon Software either on a global level (probably not ideal that) or per game profile

full flint
#

at least according to computerbase it beats it. nobody else really has any new test showing ampere in this title with RT on, only turing, and sotr has been updated since the turing days, so cant really compare

#

nah, amd's option still has to be implimented in games by devs, but the difference is, its fully open source, can be used on any platform and by any gpu of sufficient featureset, including nvidia and intel gpus, and on the consoles. being open source means devs can work with it without having to request help from nvidia, and without sending in game data for nvidia to crunch on its supercomputers if they decide to help you

#

devs can operate entirely independantly of both gpu vendors and still use it, and because of this, it will win. because it will be heavily used on the consoles, it will win

#

it will see really widespread use on series X, so they can go subnative and gain a ton of gpu overhead to use for better graphics. all the games that use it there should also be using it on the pc versions, for the same reasons. and unless nvidia pays all these devs (which would be illegal) to use dlss and not to use amds solution (which would be extremely illegal), amd's super resolution has already won because of that

#

and its consistent with the first source, great we know this is accurate performance now

#

even while being cpu bottlenecked a bit by the 3500X as you can see in the graph

#

only 73% of the time spent gpu bound, so it'll score higher with a proper cpu

#

ayyyy, hey @smoky flame , this is funny 😂 check this out

#

despite being cpu bound 25% of the time in that test, a 3070 with DLSS enabled is comprable to a 6800 non XT (its direct competitor) with no DLSS in this title at 1440p ultra with RT on (6800 would actually straight up just win if tested with a proper cpu)

#

so even DLSS can't save nvidia 😂

smoky flame
#

omg 😄

full flint
#

and they're redoing some testing, he's disabling cores and threads on his 3700X to turn it into a 3500X, and he's also running non DLSS performance on the 3070 as we speak

#

so he's gonna update that chart soon 😂

floral quail
#

and they said 3070 is better than 2080ti

full flint
#

so yep, DLSS isn't going to help nvidia at 1440p here, when the slowest rdna2 card is still roughly matching its dlss performance with RT enabled

#

and they still haven't retested with the 3500X yet, that's next (disabling cores and disabling SMT on the 3700X will get you a 3500X), which will drop the numbers for the nvidia cards in that chart, and become apples to apples with the 6800 numbers

full flint
inner rampart
smoky flame
#

Nvidia never said the 3070 was better than the 2080 Ti (I think) equal, likely to trade blows more likely which for what $499? is still a bargain considering what people paid for their 2080 Ti when it launched ($999)

#

newer games optimized for Ampere rather than Turing may show advantages for the 3070 but the 11GB on the 2080 Ti is almost an equalizer, especially in higher res gaming

full flint
#

yup

devout remnant
#

I’m pretty sure Nvidia said 3070 is faster than a Ti in their briefing.

full flint
#

they did, actually, but its actually slower in a lot of games

full flint
#

lol nvidia's announcing the 3060Ti on november 17th to desperately try and shift the news away from rdna2 and back onto ampere

#

it didn't work for the 3070, and its not going to work for that

#

https://www.anandtech.com/show/16210/intels-discrete-gpu-era-begins-intel-launches-xe-max-for-entrylevel-laptops intel's first discreet gpu... is slightly faster than an MX350, which means that ryzen 5000 series APUs and their rdna2 based gpus will completely demolish these things. some of the features are pretty cool though, and have more implications in datacenter than in laptops, its just a bit weird they're launching them in mobile instead of actually datacenter atm

smoky flame
#

ehh, when it comes down to it paper specs alone should put the 3070 ahead of the 2080 Ti, even factoring in the 11GB it has over the 3070's 8GB. Too bad games can't make proper use of CUDA cores (and AMD's equivalent) because if they did we'd definitely be seeing improvements and both the 3080 and 3090 would be way out ahead

full flint
#

SUB, the problem with that statement is assuming that all shaders are created equal. the reason ampere is under utilized is because if there's any int instructions at all on a given cycle, half of the cuda cores can ONLY be used for int work on that cycle, even though int only accounts for about 30% of instructions for gaming workloads. so a third of the gpu is idle basically all the time while gaming. there are other issues and bottlenecks at play as well, but that's the big one.

#

if you run a purely FP workload on ampere, like tile based 3d rendering, the performance uplift over turing is roughly linear, but for games it isn't and can never be, due to how they designed ampere

smoky flame
#

I see

full flint
#

ampere was designed first as a datacenter card, that design limitation isn't a problem in the datacenter

#

nvidia made the same mistake amd made almost a decade ago

#

it tried to use the same architecture for computational work as it uses for gaming, amd tried that with GCN, and while it worked great for a bit, they eventually fell behind, and utilization is a big problem on GCN era amd cards

#

on paper they're monsters, and for compute they show it, but games can't actually use most of that power

smoky flame
#

what about AMD though? How do their cards fair if games could fully use the available cores within a GPU

full flint
#

rdna2 seems to be laser focused on gaming performance, and they've split and are continuing evolutions of gcn for compute (cDNA is a gcn derived compute only architecture)

#

utilization seems to be improved, thanks in large part to the infinity cache being able to actually keep the gpu fed, especially at lower resolutions which is why it gains ground relative to nvidia at 1440p, it has the bandwidth to keep the card working, and to keep data fed to all of those shaders

#

ampere simply doesn't, even with their 320 bit GDDR6x interface (3080)

#

infinitycache is nearly twice as fast as a 384 bit G6 bus, and like 75% faster than the 320 bit G6X bus on ampere

smoky flame
#

ooh

full flint
#

its something dumb like 1.7TB/s bandwidth

#

RTX 3080 has 760GB/s bandwidth, 3090 has 936

#

1660GB/s for infinity cache

#

so i was right, its something dumb 😄

#

i was wrong though, its more than twice as much bandwidth as a 3080, though its a small pool of cache, its enough to keep dispatching work while you do regular memory fetches for things

smoky flame
#

nice, really looks like AMD never needed HBM to achieve amazing speeds with memory. This infinity cache is some computing wizardry though. How on earth does so little (128MB) do so much? rooWow1

full flint
#

its also enough to store the bulk of the bvh heirarchy in cache, rather than thrashing memory while doing RT work constantly. which explains why its doing so well in tomb raider as seen earlier today, better than it should on paper

#

because its ridiculously fast, and ridiculously low latency compared to graphics memory, and it lets the shaders move on to new tasks very rapidly

#

they still need to pull data in from memory, this is true, but they can start work very quickly, instead of waiting for a memory read before they can do anything

#

if the data a given shader core needs is sitting in cache, from having been computed by another shader a few cycles beforehand, it can begin work without a lengthy delay for a memory write and another for a read operation to get the data to the next shader

#

this is why cache is always so useful, its bloody fast, even if its small

#

its also a big reason for zen3's performance increases vs zen2, each core has direct access to twice as much cache, and can talk to the rest of the cores on the die without ever going to memory first, where with zen2, if the core that needed the data was on another CCX, you had to travel to memory, and that's a 100+ ns delay

#

the cache is expensive though, SUB, in terms of the die area cost of putting that much of it on there

#

roughly 8 - 10% of the die area appears to be used for infinity cache, which is not insignificant

#

more memory controllers for a wider traditional memory implimentation would be nearly as large, to be fair

smoky flame
#

this is all pretty cool and you clearly know a lot

full flint
#

i read a lot, and hang out in extremely nerdy discord servers, like the graphics programming discord

#

but yeah, this cache is definitely why rdna2 is scaling so well as resolutions fall, where you're not utilizing all of your shaders anymore, but need to be able to feed fewer shaders with more and more data

#

1080p performance should be even higher relative to ampere, where ampere is barely faster than turing at 1080p

#

obviously 1080p performance isn't a focus for cards of this calibre, but we'll get testing done on it by the usual outlets

smoky flame
#

hmm idk but I see potential for AMD to merge the two 64MB infinity caches in a future GPU or is it set that way for a reason? Seeing them like that isn't it almost Zen 2-like as you mentioned before?

full flint
#

no, they have to be arranged that way, because all data that comes in from memory takes a trip through it too

#

and your memory controllers must be at the edge of the die, they just have to be, and you can see them there

#

there are 4 shader engines here, 4 groups of up to 20CUs basically, each with 1MB of L2 cache, and each bolted to 64MB of infinity cache

#

all controlled by a central group of command processors and schedulers and other fixed function nonsense in the yellow block you see in the middle there

smoky flame
#

so this is almost chiplet like or at least CCX and CCD

full flint
#

yes but also no, because the link between all of that crap on-die is several TB/s, and latencies are obscenely low

#

data that fits in cache can be sent anywhere else on-die without a trip to memory

#

there's obviously always some latency involved, its traveling some distance and the speed of light is finite

#

each CU has its own local L1 cache, each shader engine has a block of shared L2 cache, and then at the level above that is a giant block of L3 cache on this design. each cache level has dramatically higher latency than the level below it, but memory is magnitudes slower

smoky flame
#

hmm maybe there is potential for AMD to merge or rather double up on these shader engines then. 4 of these engines doing what they do now vs 2 with double what they have. Idk but it all looks really cool and looks as though there is room for further improvements yet. Really hope Navi remains a thing for some time

full flint
#

actually, the next step is going to be to split all of these shader engines off of this die, and put them in chiplets. once amd's active interposer designs are ready to go, they'll run all the fixed function stuff, and memory controllers, and L3 cache in a central control die, and then they can run multiple chiplets for any given shader count they need. nvidia is also working on this, but amd has a decade+ headstart on this kind of tech, they've been working towards it since fiji (fury-x)

#

the only thing stopping them right now from doing this for a gaming gpu is that current passive interposers don't have the bandwidth needed to pull it off

smoky flame
#

hmm, so we could see chiplets carry the equivalent of the above, or more yet and chiplets being chiplets I guess we'll see more than one of them in some future GPU?

full flint
#

yes, precisely. as shrinks become harder, its the only way gpu tech can proceed. we've already seen it happening with cpus, and its possible with cpus because link bandwidth between chiplets and i/o are much, much slower. infinity fabric is something like 50GB/s on zen2 (bi-directional, so you an send at 50GB/s while also recieving at 50GB/s)

#

you need more like, 5TB/s to do this with a gpu, and that requires an active interposer being driven by some kind of asic, because you can't build a wide enough bus to achieve that with passive traces

#

fun fact, amd's purchase of xilinx will be critical to achieving that goal. its about more than just getting into new markets

smoky flame
#

yeah, I heard about that deal. Did it go through fine? Nvidia's deal with ARM seems to have been held back

full flint
#

it hasn't closed, but there are no real regulatory hurdles here like there are with nvidia's bid for arm

#

nvidia sells mobile cpus and self driving car SOCs, buying arm gives them an unfair competitive advantage in that they can keep the latest arm core designs for their own chips and make their competition fall behind, thus squeezing everyone else out of those markets

#

xilinx make fpgas and other things that amd has no market in at all right now

smoky flame
#

that would be very unfair. I'll be surprised if Nvidia can actually pull this deal off based on that

surreal moss
#

This rocket lake flagship is looking really interesting

#

An incomplete userbenchmark scored just short of the 10900k with only a 4.2ghz boost clock

#

Strangely enough the memory attached was a single 31GB stick at 2667

#

Who is running this... thing... of a pc lol

full flint
#

31GB stick makes no sense 😂 its a fake or bugged entry

#

tons and tons and tons of fake nonsense gets uploaded onto userbenchmark constantly

#

also, a 10900k running actually in spec only runs 4.2 all-core sustained anyway, so a 4.2ghz 8c rocketlake matching it would mean intel's double digit ipc gain claim is true. The bigger issue is intel's disclaimer that margin of error on all performance claims is +/- 15%, which means they shouldn't even be making statements about performance in the first place yet, because the chips nowhere near ready. 15% error margin is obscene

smoky flame
#

this rocket lake is meant to be the 10nm backport/last of 14nm right?

full flint
#

yes, its a 14nm backport of willow cove (icelake)

#

it wont be enough to overtake zen3, because clockspeeds will take a hit due to the much larger FPU

surreal moss
#

Calm down there fanboy lol

#

It’s a 32gb stick by crucial

#

BL32G32C16U4B.M16FB1

full flint
#

also, gaming performance testing is going to be interesting. the 6900XT seems to be the fastest gaming gpu going forward, and that means that intel's at a 5 - 10%+ disadvantage due to smart access memory alone, when comparing the two for gaming

devout remnant
#

I’d be wary of that though

full flint
#

ok but how is it reading as a 31 gig stick?

devout remnant
#

AMDs benchmarks aren’t 100% accurate.

surreal moss
#

It was via an msi z590 too which is neat

#

It had something indicative of native AVX512 support so that’ll be neat as well

full flint
#

native avx512 support was a given as soon as we learned they're backporting cove cores

surreal moss
#

Cpu support yes

#

Native Mobo support no

full flint
#

mobos dont need to support avx512, its a cpu instruction set

surreal moss
#

Now we can mostly assume it won’t try and interfere with the instruction on launch mobo

#

No motherboard plays into that greatly especially when doing auto voltage adjustments when running them

full flint
#

rocket lake will go sub 3ghz during avx512 operations across all 8 cores, otherwise the cpu will melt

#

sustained clockspeeds are taking a huge nosedive specifically because of avx512

smoky flame
#

how big of a deal is this AVX512?

full flint
#

if the 125W PL1 is enforced, clockspeeds during floating point work are going to be very low, and if its not enforced, we're not talking about stock performance anymore

surreal moss
#

I’m excited to see the boards tbh. The extra lanes will be very noice for direct storage attachment

full flint
#

AVX512 is very useful for specific types of compute workloads, and not at all for consumers on a mainstream socket

surreal moss
#

I’ll get a 980 pro and we’ll be good to go

smoky flame
#

so it is more a prosumer instruction set?

full flint
#

not really even that, its for specific types of scientific compute work, and some datacenter stuff

#

prosumers don't really even benefit from it outside of very specific types of workloads, like h.265 video encoding, where you can see some nice speedups if you have memory bandwidth to match

smoky flame
#

damn, what's that even doing on a consumer CPU then?

full flint
#

unfortunately, with a 125W PL1, avx512 clockspeeds will be very low, sub 3ghz for sure, and that will erode much of the performance gains that they otherwise could have had

#

its on a consumer cpu because removing it requires redesigning the core, and that's counter productive

smoky flame
#

Sounds a pretty redundant feature but then again it is Intel we're talking. Still on 14nm over 5 years later rooSmug1

full flint
#

they could in theory simply disable it on mainstream chips, but that too is kinda pointless, you can count the number of consumer programs that support avx512 on one hand, and that isn't really going to change

surreal moss
#

I’ve told you this before

#

The whole 14nm bad thing is a terrible stance and looks like you’re reciting what you read on Reddit

full flint
#

intel's 14nm is very impressive, its just not impressive enough anymore with what amd are now accomplishing. they badly need to shift nodes going forward, but they aren't ready yet

surreal moss
#

Node size equates to efficiency, but not how well a company is developing. There’s a reason these 14nm chips are still competing with 7nm chips that are HALF the size

#

Rocket lake is a 10nm standard back port and alder lake will be the full 10nm superfin

full flint
#

XLR8, not when run at intel spec they aren't. clamp a 10900k to 125W after 56 seconds and sustained turbos nosedive to 4.2ghz for int work, 3.8ghz for avx workloads and nowhere near the 4.8 all-core they do when you overclock them to allow them to break spec

#

rocketlake isn't 10nm at all, its 100% purely 14nm, they took the 10nm willow cove and ported it to 14nm

#

alderlake is supposed to be superfin, you're correct about that though

surreal moss
#

Nobody cares about that because the 14nm node is matured enough to run at PL2 at all times

smoky flame
#

PL2 at what 250w 100% of the time?

surreal moss
#

It’s literally a 10nm part inside a chip with 14nm supplementary components

#

It’s not the 14nm back port you’re thinking of

full flint
#

if we're talking stock performance, then running the chip at actual stock speeds should be a thing that is done. its fine if you want to compare OC vs OC, but when testing stock, you test with the PL1 duration and power limit in place otherwise its not actually a legitimate test

surreal moss
#

Again nobody cares about that. You put the chip in the mobo, the mobo overrides that because the chips have so much more headroom than that, and your experience isn’t even close

#

I get what you’re saying, but literally nobody in the pc world cares about stock testing legitimacy

full flint
#

if you're reviewing a processor and claiming that you're testing at stock, testing at actual stock should be required otherwise you're literally lying to your viewers

surreal moss
#

I mean if you pull the chip out of the box, put it in a mobo, fire it up without changing anything, and that’s how it performs, then that’s going to be the “stock” experience for most people on the planet

#

This sounds like saltiness more than anything else

full flint
#

by the way, find me one quote from a tech news outlet that says rocketlake is 10nm, because every single thing i can find says its straight up 14nm across the board, no 10nm silicon inside of it at all

#

different motherboards enforce different boost behaviour though, if you put the chip in an asus motherboard you're getting intel stock unless you explicitly bypass the power limits, for example, which reviewers of course did, thus making it not stock for those boards again

#

if you put the board in a budget motherboard, it'll also be clamped down to intel spec

smoky flame
#

speaking of Asus, what is that multicore enhancement feature they have on their motherboards?

full flint
#

testing at true stock is the only way to represent the actual performance on any compatible motherboard across the board

#

MCE runs all cores at the single core turbo clockspeed, by dumping whacktons of voltage into the chip and allowing it to guzzle unfathomable amounts of power, its an overclock, a very bad one, but loads of reviewers use that too

smoky flame
#

wow, glad I have that off then

#

I have everything at auto and MCE off

#

on my 6700K. Still not even sure if that is stock or what

full flint
#

yeah, that'll be stock

devout remnant
#

I finally turned off the LED on my ASUS motherboard after 1.5 years

#

Didn’t know you could.

full flint
#

intel's quad cores were incapable of drawing enough power to violate spec until the 7700k came out (and running that with PL2 duration unlimited will let it pull 150w+ in some workloads)

smoky flame
#

heh yeah, took me just as long (probably longer) to figure that out too, EnVii 😄

full flint
#

EnVii, its been a thing for like, a decade 😂

devout remnant
#

Arghhh I didn’t knowwww but that orange LED was driving me nuts 😂

full flint
#

lol

smoky flame
#

can AMD CPUs be set to all auto (stock) too or is that not recommended? I hear Ryzen likes to run at a pretty high voltage

full flint
#

auto is fine, the voltage isn't actually running that high, the instant there's any kind of load the voltage hits what you'd expect. its a side effect of the way they power gate idle cores (shutting them off completely), when monitoring software pings the cpu to figure out what voltage the cores are running at, they wake from idle, and that transient spike is what's being recorded

#

when a cpu core powers on, a burst of voltage hits it so it can quickly ramp to full speed to respond to incoming load, amd tweaked the way the sensors report data at some point last year and you dont even read those 1.5v spikes anymore in the first place

surreal moss
#

MCE just activates the all core OC and removes the time limits on the power phases

#

And chat just updates

#

Nice

full flint
#

MCE runs all cores at the single core turbo, or it did prior to this generation, did that change with 10th gen?

#

if you enable MCE on a 9900k, every core runs at 5ghz, which is the single core turbo clockspeed

#

it also eats like 275w, so there's that, yay overvolting the crap out of a cpu, good job mobo vendors

smoky flame
#

wow

surreal moss
#

All MCE really does is removes the time limits on the power phases. The clock speed is always overriden by whatever you manually set (which if you’re in the bios messing with stuff, you probably have one set) If you don’t set anything it just runs the single core boost on all cores.

#

So if you tell a 10700k to run at 5.2 all cores, it’ll allow it to run at 5.2 all cores with unlimited usage of PL2

smoky flame
#

my 6700K doesn't seem to pass 4GHz yet it is said to boost up to 4.2GHz, which I guess is a single core boost

full flint
#

4ghz is its all-core turbo

#

your 6700k is behaving as expected

smoky flame
#

good to know. How differently do AMD's CPUs behave if I was to set them to auto?

surreal moss
#

Well

#

They usually pull back a little bit when actually running at 100%

#

Just like .1-.2

#

Oh nvm

#

Auto is really the best setting. OCing them doesn’t get you much

full flint
#

amd cpu's don't have fixed turbo tables depending on how many cores are active. they have a maximum turbo, a defined maximum power limit (142w stock), and the chip clocks as high as it can for the load its under while keeping under that predefined power limit, and a temperature threshhold, so long as the motherboard VRM is reporting acceptable temperatures

surreal moss
#

I run the 3900x on auto. I’m able to OC to some pretty good spots, but they get very hot very quickly at not much benefit for what I use it for.

#

So cool and quiet is my choice

full flint
#

that 142w limit is for the 105w cpus i should mention, 65w cpus have an 88w maximum defined allowable power draw, assuming the same conditions are met (motherboard vrm able to cope, cpu temps and voltages not passing certain threshholds)

#

basically SUB, its complicated and hard to really define in certain terms, because it will vary cpu to cpu, and with your particular setup

smoky flame
#

really can't wait to get my hands on the 5600X

surreal moss
#

It’s like the 3080 GPU tables we talked about when they were crashing because of improper tabling

full flint
#

if you have a terrible board and ambients are like, 30c+, and are using a crap cooler, it wont boost as well as if conditions are better

#

zen2 and beyond behave like a gpu in terms of how boost works basically, rather than like an intel cpu

#

intel will adopt this behaviour eventually too, because rigid turbo tables are starting to bit them on the butt (which is why they've stopped reporting anything but the single core turbo and the base clock)

#

boosting as far as you can for the environment and load conditions you're facing is the way to do things going forward

smoky flame
#

I think sticking to stock is how it should be. Minimal issues with temperatures and games that don't play nice with overclocks. Such games are likely rare but still. All that time spent overclocking a CPU only for a game to either break or not run because of it? no fun. Same applies for GPUs in my eyes

full flint
#

source : me, just now lol

smoky flame
#

nice, that's pretty low (I think?)

full flint
#

not really, nor is it particularly high

smoky flame
#

almost certain my 6700K is running just shy of 1.3V and that is meant to be stock. Was told a while back I'd be able to lower it to 1.2V

full flint
#

stock voltage is supposed to be 1.2v on the 3600, mine wants 1.26v under heavy stress, that's acceptable and well within a safe range for zen2

smoky flame
#

I've heard some Zen 2 or Ryzen CPUs overall reaching 1.4v, possibly even passing that. ngl, when I read that it scared me but seeing this it doesn't bother me so much

#

Hopefully Zen 3 plays nice with B550

full flint
#

its designed explicitly for b550 and x570 from the start

#

some x470 and b450 boards already got their beta bioses to support zen3 like 3 months early, so thats cool

smoky flame
#

are bios updates any hassle on Ryzen?

surreal moss
#

Depends on the board

full flint
#

they're exactly the same as anything else

#

put file on flash drive, put flash drive in usb port, go into bios and tell it to flash bios. alternatively, put flash drive into specific usb port and hit usb flashback button while no cpu is installed, and wait a few minutes

#

most B550 and X570 (nearly all) boards have usb flashback on them

surreal moss
#

1.4 won’t hurt them btw

#

It’s a lot but it’s not gonna kill the thing

full flint
#

1.4 under sustained heavy stress is pushing it, and you only see 1.4v in single threaded during torture loops anyway, with auto voltages

#

which is perfectly normal

smoky flame
#

ahh, that sounds easy then. Idk what I was reading some time back that had me all worked up on Ryzen. Something that required AMD sending out some update kit

full flint
#

multithreaded you'll never see it go over 1.3v stock really

surreal moss
#

I never kept you updated but I hit 4.4 all core on that 2700x with around that voltage

#

It was like 1.38

full flint
#

oh, that was only if you had a board that lacked usb flashback and came with a bios that didn't support a newer generation cpu, and the same problem would have been a thing for intel owners buying a, say for example Z170 board and a 7700k

#

flashback was rare on 300 series boards, its a common thing now, and included on both boards i reccomended you for your own build SUB

smoky flame
#

nice, panic avoided again. I really like this. Things seem to be piecing together for me, maybe too nicely

full flint
#

highlighted usb port and that handy dandy little button is what you need

surreal moss
#

All bios since like 2014 allow you to update it directly within the bios if you don’t have a flashback button

#

It’s nothing to worry about

full flint
#

yeah, but if he gets a b550 board that somehow doesn't have a new enough bios out of the box for zen3 support, it still doesnt matter because any board i'd reccomend him has flashback anyway

#

as seen above

surreal moss
#

That’s impossible for a 550

full flint
#

not true, you need a bios that supports combopi v 1.0.0.8 or above for zen3 to post on a 500 series motherboard

surreal moss
#

Nevermind I read gen3 not zen3

smoky flame
#

I can't remember the last time I done a bios update tbh. Even with the supposedly high end Z170 board I have it hasn't received an update in at least a year, maybe two now

full flint
#

its not really a problem, i think only like, 8 B550 boards lack flashback, and 2 of them are biostar boards and the other 6 are extremely low end boards with no vrm heatsinks that you dont want anyway

#

and this board is unfit to be used with ANY cpu, and biostar should not be allowed to sell it

#

oh my dear jesus they charge $98 for that thing 😂

#

how are they even still in business?

smoky flame
#

I asked around on the other server I'm in that's really into tech about the pcpartpicker build you done and it was all good from them too. Though someone updated it a little by adding the 5600X and 6800 XT as custom parts. Something I think they mentioned could only be done when logged in to pcpartpicker

full flint
#

i mean, yeah, i was just too lazy 😂

#

and i didnt think you needed help remembering those two parts

smoky flame
#

no but I just wasn't aware that could be done until I got the link lol

full flint
#

yeah, you can

smoky flame
#

anyway, the crazy thing is that is about as much as I spent way back in 2016 on my PC. Back then I got a 6700K, 16GB DDR4 2133MHz, Corsair H100i v2 AIO, 980 Ti, 250GB SSD + 2TB HDD. Not completely breaking the bank and getting pretty much a whole system for about the price of a 3090 FE

full flint
#

yup, 3090 is dumb

smoky flame
#

only thing I may have to add is a new PSU yet but I'll see how my 2016 Corsair RM650i holds up before spending any more than I may have to

full flint
#

Anyone know if I can play a GFWL game without the cd key?

#

No intention of playing online

storm urchin
#

Rtx 3070 or rx 6800

full flint
#

RX6800

quiet swallow
#

Ryzen 5800x and RX 6800xt 😻

full flint
#

Wdy guy think about swtor , i played with different characters but didnt finish the campaign with one, the grind gets boring, and the story can dip

#

If i played it earlier then maybe i would appreciate it but not now the sith probably has the best story out of the bunch even the starter planets

#

MMOs usually suffer pretty hard in a lot of areas since more developmental focus is used up in more specific areas. Wasn't a huge fan of SWTOR for a lot of reasons (that I can't remember, it's been too long), but that's pretty much par for the course with how I feel about MMOs.

#

I havent touched a proper mmo since mxo days, then i stumbled upon swtor. It was ok for the most part but it does lack in other areas. The next mmos i would play would be ff14 or something as its been well received, and monster hunter those are it

#

FFXV and MH aren't MMOs.

#

I think you might be getting MMOs and RPGs mixed up.

#

Nah ff is mmo also mh

#

How is a single-player game (FFXV) an MMO?

#

MMO stands for Massively Multi-player Online. Like World of Warcraft, for example.

#

Ffiv

#

Ah. FFXIV. I getchu now.

#

😅

#

Was super confused there. MH doesn't really fall into the MMO category either, though.

#

Monster hunter online

#

Ah.

#

I'm clearly not up-to-date on what MMOs are available nowadays. XD

#

Those 2 are quite popular in the mmo community

#

😂

#

Yeah I stopped playing MMOs before 2010, honestly.

#

Why was that

#

¯\_(ツ)_/¯

#

Moved on to greener pastures. Single-player RPGs are usually infinitely better.

#

Because you became a nihil

#

😅😭

#

Nihil?

#

Nihilist

#

I don't think nihilism applies here.

#

Wdym

#

All hail ¯_(ツ)/¯. In its 11 strokes, the symbol encapsulates what it's like to be an individual on the Internet. ... ¯_(ツ)/¯ represents nihilism, “bemused resignation,” and “a Zen-like tool to accept the chaos of universe.” It is Sisyphus in unicode.

#

I'm not sure how to respond to that.

#

Thats the most stressed out looking face ive ever seen ngl

surreal moss
#

I’m a narcissist

#

Because I’m the best

full flint
#

I'm not surprised, XLR8. 😂

surreal moss
#

I live to make people feel bad about themselves

#

Not in a self depreciation way

#

But in a “I wish I was brad” way

full flint
#

I'm pretty sure I have a Messiah complex.

surreal moss
#

Brad is the best

full flint
#

Probably explains a lot about me, if I'm honest.

surreal moss
full flint
#

😂

remote vapor
full flint
#

Western digital black SN 850 boi

still wolf
#

What

devout remnant
#

@smoky flame

#

I feel this pain ^

smoky flame
#

lol

full flint
coarse pond
#

So board partners came out about the 3070 delay and said "they wouldn't have had a single card" according to a video steve just released

#

also has more information about the 3080 20gb and 3070 16gb cancellation

full flint
coarse pond
#

The other day I guess. Still new to me and no one else was talking about it here

#

Someone here mentioned that that wasn't the reason that the 3070 was delayed but apparently it was

full flint
#

yeah, nvidia messed this launch up in every possible way, and its worse than i've ever seen a hardware launch before

coarse pond
#

When you're on top for so many years I guess they just get too big of an ego. Similar to intel too.

full flint
#

wasn't really about ego, they got outmaneuvered by amd. they thought that even on a cheaper, inferior process node they would be far, far, far ahead of anything amd could produce. to compound the problem, amd has been buying basically 100% of spare tsmc 7nm wafer supply for over a year now, and have made themselves tsmcs single largest customer, so nvidia can't fix this in any reasonable amount of time now even if they wanted to. Nvidia should have delayed their launch by 6 months, but doing so would mean amd would be uncontested this cycle. So its lose-lose and no choice nvidia could have made here was a good one once it became clear that amd had not only competitive cards, but seemingly outright faster ones

#

its also the reason for the dramatically lower priced flagship vs last gen ($700 instead of $1200), and the tricky marketing tactic of calling the flagship the 3080 instead of the 3080Ti, to make it look like its only replacing the 2080, and not the 2080Ti

coarse pond
#

I just meant that they thought they would be okay and still would still have the best product, I guess. Them being on top for so long I think they didn't think AMD was going to have anything close to them ready and more supply.

full flint
#

more supply was something amd was guaranteed to have, these gpus are going to be their like, 9th design on the 7nm node, where ampere is nvidia's first on samsung's node, and it shows with the problems they're having manufacturing them.

#

the node sucks, and nvidia didn't have the benefit of multiple generations to work out the kinks to make the node work

coarse pond
#

Yup! I bet they are really regretting that now too though

full flint
#

whether they regret it or not, they didn't really have a choice

#

amd is buying all of tsmc's spare capacity, and that leaves nvidia out of luck doing any kind of mass production at tsmc right now

coarse pond
#

If they had the foresight that amd had on just going with tsmc.

full flint
#

amd has leverage too, they're tsmcs largest customer, and a ton of their patents are in tsmc 7nm and even more are in tsmcs 5nm node, so nvidia will have to wait in line, tsmc needs amd, they don't strictly need nvidia

#

that's exactly the issue, nvidia can't just go tsmc for a high volume part, there's no supply for them, because amd has secured most of it

coarse pond
#

I know. I meant if they had the foresight like last year or something of going with tsmc

full flint
#

nvidia would have loved to go tsmc for everything, despite what they say, they've never done a gpu that WASNT on tsmc, outside of some low end cards like the 1050Ti being made by samsung

#

wouldn't have mattered, amd have been buying as much 7nm as tsmc has to sell since before 7nm came out, and they're already buying all the 5nm they can get ahead of putting their products on 5nm next year

coarse pond
#

Basically they were just too late to pick up on it before amd did, since they were already with tsmc they should've picked up on it way ahead of time

full flint
#

they literally couldn't

coarse pond
#

I mean I don't think tsmc would've just not supplied them if they both started picking up on them at around the same time? they would've at least got some of it

full flint
#

amd patents made tsmc's 7nm node possible in the first place, and their 5nm node, and future nodes. without amd, tsmc would have gotten beaten to the punch with 7nm by global foundries and samsung both. amd made tsmc the world's leading fab, and gave them an enormous lead over intel, who previously were the world's most advanced fab. amd has all the leverage in that relationship, and the cash flow to justify buying all of the wafers they can get

#

nvidia can bid for wafers, and they'll of course get some wafers, but tsmc is not going to give nvidia priority over their single largest customer, and the company that gave them their status as the best fab on earth

coarse pond
#

Welp that's just an unfortunate case for nvidia. I didn't know, I was just looking at 20 series launch and it was 10 times better compared to this.

full flint
#

nvidia's total output is miniscule in terms of volume compared to the entirety of amd, they're not even a drop in the bucket in terms of the amount of silicon the likes of amd, qualcomm, and apple move every year (qualcomm is on samsung nodes going forward it seems, so that's even more supply amd likely bought up recently)

#

nvidia's a huge company, and worth a lot more than amd, but they make most of their cash on low volume, ungodly high margin parts like the 100k+ DGX boxes for datacenters

#

meanwhile amd is supplying the SOC for both consoles, and making hundreds of millions of zen chiplets a year for use across their entire cpu lineup, from $50 cpus through to $7000 datacenter cpus, they all share one chip design (their cpus i mean)

#

its not even close in terms of numbers of chips moved 😄

#

could they have worked out some kind of deal to get supply with tsmc? yes probably, but it would have been inadequate for the volume they tend to ship with geforce, so they're in a bad spot regardless

#

Nvidia's 20 series launch was a lot smoother because they launched them on a relatively ancient node, and just shipped ungodly huge chips

#

massively increased pricing made that the best play at the time, financially, and supply wasnt an issue because they are the ONLY customer for tsmc's 12nm node

azure flicker
full flint
#

thats a big mouse pad 😄

ruby spear
#

I have one like that which goes across the entire length of my desk.

full flint
#

i use a QCK+ myself, its like 10 years old now, still as good as new

ruby spear
#

They always stink when you first open them. They smell like they've come from a smokers home for some reason, but it goes away after a couple of days.

full flint
#

yeah, that's HUGE

coarse pond
#

I bought an arozzi desk that came with a giant mousepad. It's not bad.

azure flicker
#

Okay now that's huge

#

Kinda want one now

#

Do they stick to the desk because I'd like that so the tempered glass top doesn't get scratches from now on

full flint
surreal moss
#

It’s about time

#

Been rocking multiple glorious 3xl for years because they’re one of the few who made one for the whole desk

full flint
#

you getting a 5XL? 😄

surreal moss
#

No

#

My desk is 47x23
The mat is 48x24

#

I need nothing

full flint
#

oh, that works out well then 😄

azure flicker
#

i need to measure my desk

surreal moss
#

@azure flicker you got an iPhone?

worldly ravine
#

Use the measure app on the iPhone

azure flicker
#

No why?

full flint
#

iphones have a built in measurement tool

smoky flame
#

I'd like to see a mousemat that can somehow cover my whole desk. Awkward thing is its a corner desk so it won't be perfect. Even my PC and monitor on it look pretty small with plenty space for another monitor or two. Never been in this position until I got this desk back in June I think it was. Should have seen how I had it set up before. Was embarrassed to take any pictures of it, or even call myself a PC gamer

ruby spear
#

They have ones like that.

smoky flame
#

thing is my desk isn't shaped like that either. I can't remember the name of it but it was discontinued by Ikea before I got it. Predecessor to the Bekant I think

#

found an image of it and the name of it too! Ikea Galant

#

hold on, that's not quite how mine looks. That one doesn't appear to have a third leg where mine does

#

looks really small in the picture and that caught me out but it is huge. That or this being my first desk just has it seem huge

floral quail
#

I have an issue with my new monitor

#

I pluggeed it into my laptop, power is on, windows detects it, but the monitor wont display anything

#

but on startup it works

full flint
#

GPU drivers

floral quail
#

I fixed it

#

the monitor input was set to DP but my laptop doesnt have DP so I was using HDMI

full flint
#

5600X faster than not only the 10600k which has the same core and threadcount, but faster than the 8 core 10700k in single and multithreaded both, and slightly slower than a 3700X for multithreaded, while stomping a mudhole into EVERYTHING single threaded, yep, that's about right

jolly timber
#

No
@surreal moss you are ben 10

surreal moss
#

He gets yelled at in general so he goes here

full flint
full flint
full flint
#

nice, that reccomended list is clearly for 1080p60 or 1440p60 though

#

and not on max settings, as max settings include raytracing

full flint
#

It looks like reskinned warframe but gearbox

autumn sluice
#

Gonna have to wait on peer reported sys specs b4 I even consider it

#

Gah I'm so mad I got a 2600 instead of a 3600

full flint
#

I’d wait on peer reported reviews in general tbh

autumn sluice
#

Uuuugh I wanna upgrade but I still can't find a job cause covid is worse than April where I am

full flint
#

lol, the devs explictly say that 4k ultra REQUIRES 12gb+ of vram, meaning 3070, 3080 are out of luck

#

so is 2080Ti

waxen axle
coarse pond
#

Yeah I think i'm just gonna pick up a 6800xt and ryzen 7 5800x

#

ain't gonna wait for a card that isn't gonna be back in stock for months

#

i wanna see the aib cards tho

full flint
#

whoever's idea this was....BLESS

#

Snack, gamersnexus said AIB cards are 1 - 3 weeks from reference card release, depending on which AIB and which model is being discussed

#

we know asus has a model that has a 2.57ghz boost clock, with sustained in-game clocks above 2.3ghz 85%+ of the time, which is likely the strix liquid cooled card. i'd expect an air cooled strix to be 100mhz or so slower than that, give or take

#

so its like a 3700X for multithreaded stuff, and beyond everything except the higher end zen3 skus for lightly threaded and single threaded, looks like amd weren't making it up

full flint
#

lol this is dumb, a 4.7ghz 5600X beats 6ghz+ 8700k's, you can do 4.7 all-core 5600X on air (this guy did), you.... can't do 6ghz 8700k without LN2

jolly pond
#

can you guys recommend me a gpu for a bit less than 500

quiet swallow
#

Rx 5700xt

quartz hinge
#

is 85 degrees normal for a laptop gpu?

tepid fable
#

abit hot but its not like it hurts it, usuallymine stays in the 70-80 degree range, its dangerous around 100

#

my brother ran a92 degree one for around 7 years

#

if its doing something id say thats pretty normal temp

tender bloom
#

85 degrees is like the max before it throttles

#

should be safe, if you can get it cooler, thatd be great

devout remnant
#

Yeah I think i'm just gonna pick up a 6800xt and ryzen 7 5800x
@coarse pond I can’t wait to pick up a 6800XT and then a 6900XT. I just hope there’s stock.

full flint
#

hey Penguin, hows the cooler holding up? any complaints after having it about a week?

spring walrus
smoky flame
#

no idea, was just showing off what my desk looks like. Was the closest image I found online

full flint
#

Ubisoft+ anyone

smoky flame
#

what about it?

full flint
#

Watch dogs legion ac valhalla day one I think

smoky flame
#

yup, I've been playing Watch Dogs Legion since it launched (have Ubisoft+)

#

Had they kept the 3 days early access that was included as part of the ultimate edition that would have been a little sweeter

full flint
#

And a lot more broken

smoky flame
#

still, the early access is nice. Broken game or not. I'd think a game is in a pretty good state if devs decide to give 3 days, or even 7 days early access when paying for a higher tier edition of the game. Otherwise why would they include it or even release the game if it wasn't ready?

tidal meteor
#

on pc how would you use a wired headset? ive seen headphone jack ports on laptops but not pc..

thorny cypress
#

desktop PCs should have headphone jacks too

#

I'm not aware of desktop PCs that don't actually

full flint
#

Celes, i think he meant 3.5mm combo jacks, rather than the traditional 2 jacks

#

@tidal meteor you can get a $3 adaptor to plug your single wire headset into your pc's mic and headphone jacks

thorny cypress
#

Ohh, that would make sense 🙂

full flint
#

yup, a lot of headsets are single wire these days

analog sapphire
#

@heavy vault this is a good place for the questions you were asking me earlier :)

full flint
#

Yeah those are quite cheap those splitters
You can also get ones which have a mic too, lets you use any headphone and is way cheaper than the modmic
I'd say its fine for gaming maybe not for streaming and whatnot

smoky flame
#

OMG. Warframe is finally getting support for DirectX 12! https://forums.warframe.com/topic/1233916-directx-12-beta-support-coming-soon/ Shame it will probably end up a crash happy mess

Warframe Forums

Tenno, We are pleased to bring you beta support of DirectX 12 this week! This beta arrives alongside The Great Ensmallening: Part 2 build. What to Expect: The DirectX 12 driver should look graphically alike to the standard DirectX 11 driver, but the new driver exposes several ...

full flint
#

Swish!

full flint
#

end up
what's new

#

Warframe had DX9 support for the longest and still crashed

smoky flame
#

I never tried the older APIs so wouldn't know but since I got into PC gaming the only APIs I have really been aware of were OpenGL, Vulkan, DirectX11 and DirectX 12. Any I didn't name I likely haven't played a game that uses these other API or where possible through selection I go for the newer and (most would assume) better option

surreal moss
tender bloom
#

nice
what resolution, size, hz

full flint
#

360

#

if that's the sku I think

surreal moss
#

1080p360

tender bloom
#

nice
do you play comp csgo or something lol

surreal moss
#

I make people cry in mcc

#

Or quit, losers choice really.

full flint
#

so when are we doing swat doubles XLR8?

surreal moss
#

No doubles is doo doo

#

My favorite is when they think I’m on controller

#

Controller is not op ur just bad

full flint
#

input-indicators can't come soon enough

proud monolith
#

the xbox app is trash companion was much better

analog sapphire
#

what has that to do with PC gaming?

surreal moss
#

He’s talking about the PC apps

proud monolith
#

the pc app for xbox

analog sapphire
#

he's also not keeping it to one channel

proud monolith
#

is flaming garbage

analog sapphire
#

give reasons why pls

proud monolith
#

its not even hd dood

#

the format

full flint
#

berome wat

surreal moss
#

But regardless sorry you feel that way and aren’t aware of the future the new app has as it will be constantly and steadily improved and most likely surpass the console companion.

proud monolith
#

whyd u end companion

surreal moss
#

Why didn’t you read what I just said

proud monolith
#

the xbox app is harder to navigate

#

u cant chat while finding games

#

or on someones profile

surreal moss
#

Alright well have fun

proud monolith
#

wtf

#

i got a warning for saying_e

analog sapphire
#

lol

ruby spear
#

Format your messages normally then ¯_(ツ)_/¯

proud monolith
#

but i d i d

ruby spear
#

A space between each letter isn't how people normally write, is it?

proud monolith
#

orly

full flint
#

the game bar and xbox app are synergistic

proud monolith
#

im a roblox gamer its normal ok

full flint
#

you can chat while playing games, set up LFG, upload/view/share Screenshots and captures, among other things

proud monolith
#

why is there a PS in the thumbnail thing

analog sapphire
proud monolith
#

oh wait thats a mini

#

mb

#

does oculus hook up to xboxs

analog sapphire
#

no

proud monolith
#

bruh

analog sapphire
proud monolith
#

is there xbox vr

analog sapphire
#

no

proud monolith
#

thats not the crap 360

analog sapphire
proud monolith
#

o

smoky flame
#

@full flint any idea what Nvidia GPU you have on this rather odd performing system? Doesn't sound burnt out to me

spring walrus
#

@full flint What laptop do you have?

full flint
#

Hold on.

spring walrus
full flint
#

Just starting the PC.

#

It's a custom CyberPowerPC - Second hand from a freind.

smoky flame
#

wait, so it is a desktop?

full flint
#

No it's a laptop.

smoky flame
#

oh

spring walrus
#

how did your friend make a custom laptop?

#

it seems very hard

tender bloom
#

some companies let you pick out everything you want in your laptop

full flint
#

NVIDA Geforce GTX 860m

spring walrus
#

oh

full flint
#

He brought it custom.

spring walrus
#

oh cool

#

I have a 2015 Macbook Pro

tender bloom
#

yeah
cyberpower, clevo, brands that use a tongfang chassis

spring walrus
#

works fine, can play most games

tender bloom
#

bruh how

#

what games u playing

spring walrus
#

wdym?

smoky flame
#

ouch, that GPU is pretty weak. Should still be stronger than the Intel integrated graphics though

full flint
#

I keeps turning itself off.

tender bloom
#

ive tried to play games on a 2015 macbook pro, and it didnt go too well @spring walrus

spring walrus
#

I play Fortnite, CS:GO, Roblox, etc.

#

it works fine though

full flint
#

Like it can be playing a game, then gives up and turns itself off.

tender bloom
#

weird, i couldnt run anything on a macbook
except like starbound, which is a pixel game

spring walrus
#

No? that only happens if your CPUs get too hot. It never happened for me.

full flint
#

It was doing very consistently.

smoky flame
#

just sounds to me it is a pretty old system and maybe it is ready to give up

spring walrus
#

oh. maybe the one you were playing on had some faults

full flint
#

How is my 13 year old laptop still kicking?

spring walrus
#

just sounds to me it is a pretty old system and maybe it is ready to give up
@smoky flame Nah, it has good specs and works fine. Only thing I might have to change is the storage

full flint
#

I have to disable it or it will crash TF2.

spring walrus
#

i might switch to 1 TB

smoky flame
#

technology can be funny that way. Some old stuff can outlast newer hardware

full flint
#

And any game I try to play for more than an hour.

spring walrus
#

yeah

#

Plus I'm just a highschooler, don't need to game that much

full flint
#

But the Intel HD 4600 needs to be on or the PC doesn't even register the NVIDA

smoky flame
#

personally I could never game on a laptop. They do have their uses though. Just not for gaming - Cloud gaming is an exception though since it doesn't rely on the performance of the system but rather a cloud gaming system

spring walrus
#

the only hardware my laptop isn't so good is the graphics card

#

Intel Iris 6100

smoky flame
#

have you updated the Nvidia drivers?

full flint
#

Any reason why it's not reading the card at all when the intel 4600 hd is disabled?

smoky flame
#

I don't even know what architecture this 860m is on but I'd like to think you still have the option to update the drivers of it. May be you are using a really old driver if you haven't kept on top of it. Latest Nvidia driver is 457.09

spring walrus
#

ouch, that GPU is pretty weak. Should still be stronger than the Intel integrated graphics though
@smoky flame really? works smooth

#

hmm could be true

#

but again don't need ultra high specs

full flint
#

Trying to update it now.

smoky flame
#

before you update it check what version it is on by going into the Nvidia control panel

full flint
#

That's not even showing up.

smoky flame
#

sounds like you might not even have Nvidia drivers installed then

full flint
#

I'd laugh if it was something stupid.

smoky flame
#

try downloading the latest Nvidia driver and see how it goes

#

just checked and the 860m supports 457.09

full flint
#

Like the Intel 4600 was coding over the the NVIDA

smoky flame
#

I've heard laptops will do that. Sounds a lot of trouble if you ask me

#

Nvidia control panel should have an option to prioritize the GPU over CPU

full flint
#

My PC is running so slow.

floral quail
#

Like it can be playing a game, then gives up and turns itself off.
@full flint that happens to me too

#

but mine has a GTX 1650 and a i7-9750H

full flint
#

ignore the wallpaper

smoky flame
#

at this point idk. I lack any experience with gaming laptops. All I have heard from various places is they can be a real pain to get working right and by the sounds of it they are a lot more hassle than worth

full flint
#

That's no worries man.

floral quail
#

that happens to me too
im my case, the ants ate some foam that had to do something wiht cooling

smoky flame
#

fingers crossed you manage to solve it

floral quail
#

im my case, the ants ate some foam that had to do something wiht cooling
is that weird or not ?

full flint
#

oops

#

Operating System: Windows 10 Home, 64-bit
DirectX version: 12.0
GPU processor: GeForce GTX 860M
Driver version: 457.09
Driver Type: Standard
Direct3D API version: 12

#

Direct3D feature level: 11_0
CUDA Cores: 640
Memory interface: 128-bit
Total available graphics memory: 10199 MB
Dedicated video memory: 2048 MB GDDR5
System video memory: 0 MB
Shared system memory: 8151 MB
Video BIOS version: 82.07.27.00.06

#

system video memory seems to be a red flag

floral quail
#

yes

tender bloom
#

whats the difference between GDDR6 and GDDR5

floral quail
#

5 is slow and 6 is fast (idk alot so you should wait for others to answer too)

tender bloom
#

i googled it and thats right

#

lol

full flint
floral quail
#

sameee

full flint
#

i do be doing a hecking confuse.

floral quail
#

but now I got new monitor so I have more options

smoky flame
#

check under physx configuration and see what options there are. I think the option to select the GPU is there though I may be wrong

#

if not it may be under manage 3D settings

full flint
smoky flame
#

ok, that definitely confirms the system is prioritizing Intel graphics and not the GPU

full flint
#

how do I change that?

smoky flame
#

can physx be set to auto?

full flint
#

It doesn't do anything

smoky flame
#

That's what I'm trying to figure out. Honestly have no idea but I'm sure there is an option somewhere within the control panel that would allow the GPU to be prioritized. Try under manage 3D settings

full flint
smoky flame
#

open windows graphics settings

#

I think we are getting somewhere!

full flint
smoky flame
#

is that what the message from the Nvidia control panel sends you to? thought it would have looked more like this

full flint
#

I didn't even know this existed.

smoky flame
#

hmm try adding a game and see what it shows

full flint
#

Oh yes.

#

I love you.

smoky flame
full flint
smoky flame
#

is high performance set to the GPU?

full flint
#

Yep.

smoky flame
#

nice. Guessing it looked much like this when deciding? only differences being power saving probably showed the CPU and high performance your GPU right?

full flint
#

Yep.

#

Can't test it yet, but I think you just fixed it.

smoky flame
#

fingers crossed 🙂

#

also, it probably won't change anything but I think it is best you change physx to auto. System should be able to detect from there if games use CPU or GPU based PhysX and decide what is best to use

full flint
#

Sigh.

smoky flame
#

try setting it to use the advanced 3d settings

#

make sure to hit apply

full flint
#

Still nothing.

slate jackal
#

have you downloaded the Geforce experience to update your drivers? (it's a dumb question)

full flint
#

Yes.

#

Restart the PC?

smoky flame
#

could try that yeah

full flint
#

Ok it's back

smoky flame
#

nice, now try and set it to auto

full flint
#

Done, now load TF2 again?

smoky flame
#

did you make a high performance TF2 profile on windows graphics settings? make sure you do that before starting it up

full flint
#

I have.

smoky flame
#

ok, give it a go 🙂

full flint
#

Oh HECK IT WORKED.

floral quail
#

YAY

smoky flame
floral quail
#

Congrats 🥳

smoky flame
#

my lack of laptop knowledge is over! lol

full flint
#

It didn't just crash the entire PC!

smoky flame
#

wait, it crashed? 😦

full flint
#

No.

#

That's what it used to do.

smoky flame
#

oh, phew 😅

full flint
#

And it broke again :(

floral quail
#

oof

smoky flame
#

oh no 😦

floral quail
#

contact customer support an get it fixed ?

full flint
#

It's an old custom PC.

floral quail
#

oh god

#

i think your doomed, unless you fix it

full flint
#

It breaks a little more everytime.

#

This be doing me a hecking confuse.

smoky flame
#

thing I don't get is how physX decided to hide when you had 3d settings to let application decide. You could try that again but something tells me you will want it kept at what it is set to currently. Otherwise yeah, I think I have really run out of ideas. There is only one thing I have left to suggest

full flint
#

Is there a way where I can tell the PC to only use the Nvida?

smoky flame
#

that should have been the way

full flint
#

And it broke again.

smoky flame
#

this other idea I have is checking for CPU drivers and uninstalling them. Could be a conflict between CPU and GPU drivers. Only other thing I can think of

floral quail
#

im prretty sure you are doomed now if sub's Idea doesn't work

full flint
#

This is frustrating

#

Wait.

#

I changed some settings and I got TF2 to load.

#

Here's hoping.

#

Nah its broke..

full flint
full flint
#

6800 at that clockspeed should be 25% faster than a 2080Ti/3070, rather than ~16 - 17%

smoky flame
#

and this is before factoring in RAGE mode?

full flint
#

rage mode is just an increase to the power limit, like raising the power limit in afterburner on nvidia cards

#

the aib card appears to be using about 27w more than reference, to deliver what should be about 7 - 8% more performance on average. we don't know which model this is yet

smoky flame
#

oh btw, a store near me has Corsair's A500 air cooler on sale (early bird black friday deal) is it any good? £29.99

full flint
#

no, its awful

smoky flame
#

ahh I see. Just thought I'd check it out since it was quite a reduction in price (was £85 I think) and comes with two Corsair ML fans

full flint
#

gamersnexus called it impressively bad when they reviewed it in march

azure flicker
#

time to see if some sdds go on sell

#

and buy some longer display cables

autumn sluice
#

Ok so I'm gonna have some money (ty gov) what's a good upgrade from an r5 2600 and 1650 super?

indigo lance
#

hi

full flint
#

ADHD, a ryzen 5 5600X (if you have a b450 board you wont even need a new board, just a bios update whenever its available for whatever board you have). and, depending on gpu budget, the reccomendation for that could go a bunch of different ways. if you're in the $500 - 600 pricerange, its looking like the radeon RX6800 that launches on the 18th is your best choice. if you want to spend less, your options change a bit

full flint
#

changed now

full flint
#

still doesn't have RX6000 series gpus or a 5700XT listed, but that's a slightly less weird list than the first time the posted those

#

also i thought it was 250GB? what changed?

slate jackal
full flint
#

yeah, they announced it was 250GB the first time, that's weird 😂

slate jackal
#

they just wanted to make sure you had enough space kaishrug

thin copper
#

sa

azure flicker
#

in the next cod update will be at the very least 1tb

floral quail
#

My GPU is listed as the minimum 😄

#

GTX 1650

coarse pond
#

anyone know what time the cpus are suppose to be released?

devout remnant
#

@coarse pond I believe 1-2pm UK time.

floral quail
#

I went to my PC store and they said to the NZXT H710i that I chose "dont just choose whatever comes up first in amazon, first do research" and what they recommended me was antech nx800 (like idek what antech is)

#

I have never in my whole life heard of antech, is it supposed to be a popular brand or something?

full flint
#

antec has been around for a long time, they havent made a good case in 20 years

floral quail
#

ok, thats why I have never heard of em

full flint
#

it does this in.... most games in this review, intels only winning in RDR2 here, other reviews launch at embargo time, should be 5 hours

full flint
tepid fable
#

I’ll wait till my cpu is a bit older

smoky flame
#

I'll be having to wait a little longer to get my hands on a 5600X. Might be next month so hopefully stock isn't as bad as Nvidia's Ampere GPUs have been. I may be able to get my upgrade started before the year is out and if all goes right, I may even be able to finish it too

tepid fable
#

I’m using a 3700x currently so it doesn’t seem like a huge enough jump for me to upgrade

smoky flame
#

nah, in your case I'd wait tbh but you do have the easier upgrade I think. I'll be ripping pretty much my whole PC apart. Depending on what motherboard you have it may just be a matter of updating the bios so it supports a Ryzen 5000 CPU, take out the 3700X and install one of the new CPUs

#

btw, I don't know if it means anything but I got a notification from the store where I bought my 3080 from and they upgraded shipping from 3 working days to next day delivery free of charge. iirc next day was double the price of the 3 days

full flint
#

idk

#

reviews are going live for zen3 in a couple hours 😄

smoky flame
#

the order itself still hasn't processed though 😦

floral quail
#

I will probably upgrade in 7-8yrs

full flint
#

ryzen 5000 cpus go on sale in 6 minutes

#

reviews should drop then as well

#

yup both dropped

#

its roughly matching 10900k in some games, slightly slower in some, and obliterating it massively in some

coarse pond
#

Newegg shows in stock but when ordering says out of stock lul

full flint
#

yeah, its going fast

coarse pond
#

Now it shows as out of stock, been trying for 9 minutes

full flint
#

they'll be restocking quickly, this happened with zen2 as well, there was stock every week, and it lasted longer and longer each shipment, until it was finally just straight up available

#

holy crap this ones hilarious 😂

#

😂

#

Which cpus we looking at atm

#

the LTT review above is the entire stack, GN is the 5950x only, the rest of the reviews drop throughout the day

#

its absolutely one sided, intel might as well not even be in those charts

#

that should say plenty right there 😂

#

oof

#

its not quite that brutal in the rest of the games LTT tested, but its still rough for intel

#

overclocked, the 10900k catches up in about half of games, wins slightly in some, but still loses absolutely MASSIVELY in a ton of games

#

while drawing 4 -5x as much power as the 5600X stock, and you can OC zen3 as well, nobodys really gone into it too deep yet

coarse pond
#

guess it was pointless for me to get up early 😂 been up for all the launches so far

full flint
plain lily
#

hi

full flint
#

absolute bloodbath

smoky flame
#

omg, just contacted the place where I bought my 3080 from asking if I could swap the order rather than cancel it and they said it is possible. Do I dare go for the 5600X, motherboard and ram upgrade now?! Probably still works out cheaper than what the 3080 was going to cost me and I may actually get it all fairly quickly

full flint
#

go for it, then pick up a 6800XT after new years when stock is less crazy, or something

#

its quite clear that zen3 is the cpu you want, its less clear on the gpu choice, and stock will massively improve after christmas for both of them

smoky flame
#

only awkward thing is they might not have the 32GB kit I want so I'll have to check the alternatives they have

full flint
#

any 2x16 or 4x8 3600 C16, C17, or C18 kit will be fine

#

at 1440p, we're talking about low low low single digit differences between those latencies, and when gpu bound, no difference at all

smoky flame
#

motherboard should be in stock, ram too. CPU might be the only issue but should be resolved fairly quickly, unlike the 3080

full flint
#

do they have the motherboard?

#

yeah dude, grab all that, do your rebuild, then you can relax and decide on a gpu once all reviews are out and the holiday craziness is done with, that's my advice

smoky flame
#

if I go 5600X I'll likely be going for the 6800 XT. 3080 stock is hardly going to change

full flint
#

fair

#

still, stock will be easier on gpus after christmas, and cpu stock aint nearly as bad even now, except for the 5950X which was gone instantly due to low low low stock due to the extreme bins required for one, which is expected

#

5600X being the worst bins means stock will rapidly accumulate

#

yeah intel just lost a benchmark that they themselves created explicitly to crater performance on amd cpus (intel makes the benchmarks in the total war games, and sponsors them)

#

its done, wrap it up

smoky flame
full flint
#

get that 109 pound 3600C18 kit

#

you wont notice the difference between C18 and C16 at 1440p, and that's a drastic price gap

#

the patriot viper 4

smoky flame
#

any idea how they are or is ram jsut one of those things that should work and do what it has to?

full flint
#

it'll just work, unless you're trying to overclock it or tweak the subtimings for max performance, then specifics begin to matter in terms of what chips are being used

#

or if they're ultra high speed modules like the new DDR4 5000 kits, but those are all uber quality bins anyways

#

that kit will run at its XMP timings on zen3, no worries, and at 40 pound cheaper than the OTHER 3600 C18 kits, and 100 pound cheaper than C16 kits, get it, no question

#

now if you want rgb lighting and all that, you can get one of the rgb kits there, but thats up to you

smoky flame
#

something I noticed I get money back on the motherboard. Pretty rare taht in the UK I think. Almost sounds like one of those US rebate things

full flint
#

yeah

smoky flame
#

things are falling into place too well, what is going on?! 😮

full flint
#

268 for mobo + ram, cpu is what? 299 pound?

smoky flame
#

279.99

full flint
#

548 pound for the rebuild

#

you can use the cooler that comes with it for now, only upgrade it if its too loud for you

smoky flame
#

and the 3080 cost me another £200 yet! 😄

full flint
#

so 548 pound leaves you what? 200 pounds left towards your gpu when stock normalizes?

#

yeah dude, call them asap and get on that list for the 5600X

#

and grab that ram while its still ridiculously cheap

#

as for the mobo, if you have a flash drive you can do the bios update, i assume you do or can buy a 2quid flash drive off amazon

smoky flame
#

wait, waht about the NVMe SSD what was that again?

full flint
#

still 110 on sale, nice, 30 pound saved there

#

you have like 90 left, let me find you a psu

smoky flame
#

I might not need a PSU and tbh if I can save a little I don't mind

full flint
#

fair

#

ok, call the shop asap and get it done 🙂

#

put that 90 quid toward that gpu a couple months from now

#

damn, yeah, wrap it up for intel in gaming right now, and this is without running hyper tuned 3800C14 on it, which will slant things a further 5 - 8% in amds favor, and without smart access memory and an RX6000 gpu, which is another 5 - 10% shift in amds favor

#

those seeking ultimate performance period, now go all red, which has not happened in 14 years

#

@smoky flame what did you decide on? 😄

smoky flame
#

just doing it now, had to run and get some food. May have eaten it too quick lol

full flint
#

lol

full flint
full flint
smoky flame
#

idk if I'd be terrified of AMD but AMD is definitely the new Intel. At least their CPU stack is constantly improving generation by generation. If AMD never launched Ryzen guaranteed we'd still be sitting at quad core with Intel and £300+ at that

full flint
#

yup

#

did you call the shop and get an order in?

smoky flame
#

contacted them over Twitter. CS is pretty slow and I didn't make it. 5600X is out of stock now 😦

full flint
#

will they let you place an order for later?

#

you should still get that SSD and ram while it'll save you nearly 100 pounds

smoky flame
#

idk yet, they haven't got back to me since I said I'll go ahead. Their Twitter CS is due to close in like 10 minutes so I think I left it too late. Always tomorrow I guess 🙂

full flint
#

true

#

they don't have a phone number?

smoky flame
#

yeah but that would be packed with other people calling about the CPUs and whatever else plus I find it more helpful I can deal with them over Twitter rather than over the phone so that is pretty cool. Maybe I have a use for a Twitter account yet besides what I used to get up to (ripping into console fanboys)

full flint
#

😂

#

yeah if you can get on a backorder list for the cpu, you'll get it a lot sooner than you will a 3080

#

single CCD cpu? they can ship hundreds of thousands a week worldwide, and its the worst bin at that, so plenty of resupply for it

#

5950X will be the tightest supply for a few months

#

i'd say before january 1st, the 5600X supply will be fully normalized, with the rest getting under control in january

#

we saw this with zen2, where the 3600/X supply was pretty good after less than a month, and the 3900X and eventually (it launched later) 3950X supply took a bit longer

smoky flame
#

as long as the 5600X remains solid for a good few years I'll be happy. Only thing I will need to worry about will be making sure I can still get a GPU upgrade if not this year, asap in 2021. Really want to get that full AMD setup going fast

full flint
#

6800XT supply will be non existent until after the holidays, sheer demand guarantees it

#

5600X will remain good for the entire coming console generation

#

XSS has 7 3.5ghz zen2 cores available to it for games

smoky flame
#

depending on how things play out I may be able to get one ordered and bought when it launches

full flint
#

5600X is about 50% stronger than that

#

so you'd be good for the generation with the 5600X

#

ps5 also has 7 3.5ghz zen2 cores available, and XSX are slightly higher clocked

#

5600X still remains plenty

smoky flame
#

is it true the 6800 XT will also require 750W or is this AMD, similar to Nvidia covering their butts?

full flint
#

its a 300W gpu and they don't know what cpu and other components you have

#

you'll be fine with a 65w cpu like the 5600X

#

even burst spikes wont hit 750w with that setup, and that assumes you're running a torture loop on the cpu while running the gpu under load, because no game will get it anywhere near 65w while also pushing the gpu to its max

surreal moss
#

I this day and age just buy an 850w.

#

Then you just buy the parts you want and never have to worry, + the thing works for 10 years

smoky flame
#

I could do that but tbh I still don't feel it is a required purchase. At least not yet anyway. If I was hellbent on waiting on that 3080 and somehow still managed to upgrade to a 5600X I would have considered getting a new PSU but I think I will be fine for now. Max power usage on my system 5600X and 6800 XT factored in will probably top out at 465W - 65W on CPU, 300W on GPU and 100W or so on whatever else. I'm not worrying now nearly as much as I was

full flint
#

yeah, if you're buying a new psu right now (i still think you should) the extra 15 quid for an 850 over the 650 of a given series is worth it

#

you've had that one since 2014 correct?

smoky flame
#

2016

full flint
#

oh, you'll be fine for another couple years

smoky flame
#

Corsair RM650i

full flint
#

you're good, keep it for now

#

get an 850w when you do buy a new one

#

you do have enough cables hanging off the end of that for 2 8pin pci-e correct?

smoky flame
#

yeah, that will be my next step for sure or I may just go overkill and go 1000W instead. Then that way I'm ready for whatever the future may hold. 500W GPUs? 300W CPUs? Bring it on!

full flint
#

300w for a mainstream cpu isn't coolable, see non delidded 8700k or 9900k overclocks, from before intel literally sanded down the die by 70% to aid in heat transfer for the 10900k

#

those are pushing 250 - 275w and already weren't really doable for most consumer grade cooling setups

#

the only reason 300 - 500w is coolable on threadripper is because you're spreading the heat load across 9 dies and an area the size of a credit card, so heat per unit area is actually the same or less than 150 - 200w on a mainstream cpu

smoky flame
#

I can sort the cabling out while I upgrade. Got the second 8 pin cable handy and ready to go at any time

full flint
#

then you're good to go

#

i really like how the gaming edge looks top down, i might actually grab that for my rebuild (already ordered my 5900X, its on the way)

#

couldn't justify the extra $250 for the 16c chip

#

anyone just gaming shouldn't buy it, its a stupid cpu for gamers, and the best cpu for content creators

smoky flame
#

yeah, I couldn't even justify the pricing on the 5800X tbh. 5600X is honestly plenty for gamers and I agree with what you said earlier on it

full flint
#

5800X can be argued for, even for gamers, so can the 5900X, but only for those targeting 1080p 240 - 360hz experiences competitively

#

5600X is the best all around cpu choice for hobbyist creators and gamers

#

it offers 9900k beating multithreaded performance, and 10900k beating gaming performance, with a bundled cooler, for $300

#

5600 non X will become the default choice in a few months when they roll that out

#

there was no point releasing the non X skus for less money while demand was going to outstrip supply at launch regardless, so i dont fault them for waiting

smoky flame
#

that's the awkward thing I need to keep in mind. AMD will likely launch a 5600 and that will come in even cheaper

full flint
#

yeah it'll be $50 cheaper, probably 229 pound

#

but i dont expect it until CES in january, and demand for that will be high for the first several weeks too

smoky flame
#

only difference will be slightly lower clocks right?

full flint
#

here's how a 5600X behaves (not with its stock cooler, you need to put it on a high end air cooler or an AIO for it to do this)

#

i expect the non X to cap out at 4.4, and ride around the 4.1 - 4.2ghz mark under multithread loads

#

the 5600X will be faster enough to justify its pricetag (its faster than the 3700X 8c part from last gen)

#

the 5600 won't have that honor, though it will still be excellent

#

5600X is the only zen3 cpu that just kinda exists at a single clockspeed for a given workload type, the rest lose clockspeed as threadcounts increase, the 3600 was the same way

#

anyways, if you think a 30 - 50 quid savings is worth waiting a few months, go for it, but your ram and SSD are each that much cheaper than they normally are right now due to sales, so you might end up not actually saving on the total build at all by doing so

smoky flame
#

now idk again. As much as I want to upgrade the CPU now I'm still asking myself can I wait just a little longer, perhaps be in a better position and get myself the 5600 that will work out even cheaper yet while still getting about the same sort of performance out of Zen 3

full flint
#

right, but as i said, the ssd is 40 pound cheaper than it normally is, and the ram is 45 pound cheaper

#

waiting means losing out on more savings than the cheaper cpu nets you

#

you could always order those two parts while they're on sale and wait on the cpu of course, but passing on 85 quid in sales to save 30 - 50 on a cpu 2 - 3 months from now doesn't really work

#

and there's a 20 quid rebate on that motherboard too, that definitely wont exist months down the line

#

so over 100 pound savings atm, effectively makes your cpu 179 quid anyway

smoky flame
#

I'll chance my luck and instead change the 3080 to a 6800 XT. Save some money there that can then play into the CPU upgrade. Money saved on the GPU should still be greater than the savings made on the ram and SSD

full flint
#

maybe

#

if you can spare the extra to order that ram while it's 109 you should probably do so though

#

you saw the price on the rest of the kits, 149+

smoky flame
#

I think I've made up my mind. After all that fuss, I'm going to wait a little longer 😄

full flint
#

ayyy newegg paul posted his review finally https://youtu.be/NS5uWnF496E 😂 he used to be the public face of all newegg product videos before going off on his own to do paul's hardware

AMD Hurts Intel BAD - Ryzen 9 5900X Review!
⇨ Sponsor: Corsair 4000 Cases | Buy: https://geni.us/d5mAFn | info: http://bit.ly/PHCor4000
Zen 3 launches today! Ryzen 5000-series CPUs based on Zen 3 include the 5950X, the 5600X, the 5800X and the 5900X -- the 12-core 24-thread CP...

▶ Play video
smoky flame
#

5600 should still be quite the impressive upgrade over a 6700K. All that single core and multicore improvements (that I probably should read too much into) then there's smart access memory

#

silly question but how does the 5600X compare to the 6700K?

full flint
#

in what sort of workload?

#

for gaming? it'll be 15 - 20% in favor of the 5600X, intel quad cores don't really do so well in a lot of modern games right now, especially the 6700k that can't overclock worth a crap relative to all the later skylake revisions

#

in cinebench R20 (which is a reasonable multithreaded test), it scores 2200 points, 5600X is like, 4800 points or something dumb?

#

@smoky flame the 3300X in this chart easily beats the 4c/8t i3 10100, which is roughly exactly the same performance as your 6700k

#

if you want a decent reference of where the 5600X stacks up 😉

#

i'll leave you to find the 3300X in the chart, and place the 6700k a few fps below it in your mind

#

and yup, that's amds budget chip beating intel's flagship, in gaming

#

that about answer your question? 😂

tepid fable
#

Guess I’ll hold off for now looking at the charts, 720p with an rtx 2080 ti isn’t my style anyway lol, I think my gpu would need the upgrade first (2060super)

full flint
#

720p with a 2080Ti is neccesary to avoid gpu limits in some of the games tested, at 1080p the 2080Ti caps out at 155fps, so everything above the 7960X on that chart would be truncated and look like a tie

#

im not sure why they didnt just test with a 3080 or 3090, but whatever

#

you'll need to upgrade to a gpu that's 3070 tier or better (or probably RX6700XT tier or better for amd) to notice the difference for 1080p gaming overall. for 1440p? you'll need a flagship

#

some titles are more cpu bound than others, you'd REALLY notice it in microsoft flight sim, even with that 2060S, zen3 is dramatically faster than zen2 in that title (and faster than intel currently)

#

but with a 3700X you're better off waiting until ryzen 6000 series next year, on the new platform

#

new gpu next year too if you're happy with the 2060S at the moment

tepid fable
#

Yeah I’ll likely wait

#

Unless they say came out with a 144 FPS 1440p card that’s not expensive

#

Didn’t look at amd stuff

full flint
#

depends if you consider $400 - 450 expensive

tepid fable
#

Yeah it’s pretty pricey to me for it yet

full flint
#

the 3070 is $500 atm, the 6800 non XT is $579, and seems to be 17 - 25% faster (depending on whether you test reference cards or AIB cards, we'll find out a lot more on the 18th when it launches). the 6700XT should be just under 3070 performance, but it should be $400 - 450 if pricing makes any kind of sense as the stack continues

#

or about what your 2060S cost you

#

maybe a little more, but for like, 50 - 60% more performance or something depending on the game

tepid fable
#

Right now I can’t do much with higher FPS with my 60hz display so it’s more then just the gpu too

#

But works for now

full flint
#

yeah, save up and do a full system replacement late next year

tepid fable
#

Not like I go below 60 usually

full flint
#

new gpu, whichever is the bang for buck king at that point (probably an amd card the way things seem to be shaking out right now), an entry level 1440p 144hz monitor, and the new cpu/mobo/ram

tepid fable
#

Might wait for drr5 for a new mobo

#

Got enough ddr4

full flint
#

that's what is expected next year

#

ryzen 6000 series is already confirmed to be on a new socket, and is slated for a launch next year

#

DDR5 support seems likely, or they wouldn't bother with a new socket yet

tepid fable
#

Hard to say without confirmation, it should be soon though

#

Could buy it piece by piece anyway

#

Right now I’m running 3200mhz 32gb ram so I’d say it’s still good but no idea how fast ddr5 will be

full flint
#

DDR5 starts at 4800 speeds, will go up to 6400 and beyond

#

it will be EXTREMELY expensive early on

tepid fable
#

Yeah I’d say dd4 still is pricey even today

#

Launched worse

#

The board I got is pretty cheap being 130ish so it’s not a big deal to replace it

#

The ram is a bigger hit though

past tiger
#

can someone help me

#

it’s kinda specific so if no one wants to it’s alright

#

if you could

#

can you tell me the best parts for gaming pc with what it is

#

for example

#

Motherboard: Gigabyte Z490 Aorus Master

#

and the total price of a grand

#

i don’t know much

full flint
#

$1000 will get you an entry level gaming PC

#

especially when factoring in peripheral and case costs

snow blade
#

does anyone know how to fix "check your account" after opening game bar

#

can't use it basically

full flint
#

riri, $1000 is not enough, you're better off with a ps5 or xbox series X than with any pc that costs less than $1500 today, and i'm 100% serious

coarse pond
#

I really don't want to have to wait till the end of the year for my pc build and it was suppose to be a first for me. still kinda upset about it all lol

full flint
#

Snack, there are still some stores in the US with 5600Xs in stock, they didn't all go instantly, which is a good sign

#

which cpu were you tryin to get?

coarse pond
#

5800x

full flint
#

yeah, that one is more constrained