🔌│tech

[astral haven]

thats obviously false

[vagrant pagoda]

Proof?

[astral haven]

take that to the extreme

memory costs $24

take all of it away and spend $24 more on cores and the gpu is obviously definetly non functional

even if you leave 1gb its still useless

[cursive summit]

Oh.. no.. https://mixed-news.com/en/nintendo-rumored-to-be-working-with-google-on-a-vr-headset/

They really going the PSVR route

[vagrant pagoda]

I'm not talking adding bigger modules I'm talking adding additional modules for more memory width

[cursive summit]

Can't wait for PSVR to fail

[vagrant pagoda]

Show me a GPU with two cores and different widths that dose better

[astral haven]

1060 3gb vs 6gb
same bus width, 3gb gets absolutely slaughtered because it has fewer cores and less capacity

and less memory

despite the same bus width

[spring walrus]

how do i SS in this keyboard?

[vagrant pagoda]

Okay so you just can't read now I know why.

[cursive summit]

Please don't go german

[vagrant pagoda]

I literally have been talking about width this entire conversation

The question I asked was about width

I literally reiterated the message before that that it was only width not bigger modules

[astral haven]

yep and it answers your question

[vagrant pagoda]

And your answer is bigger modules

So the answer is you're either dumb or can't read.

[astral haven]

if you think that is my answer, you need to ask clarifying questions

[gusty kiln]

lol AMD subreddit whining about the "positive" 7800XT review from HWU and GN

[vagrant pagoda]

I asked a very specific question and you gave an answer to an entirely different question that I specifically didn't ask for

Please explain to me how you need clarification for that

This entire argument is only about width

Yes a lot of cards are memory deprived

Yes a lot of cards are memory bandwidth deprived

[astral haven]

basic algebra fam
take the variable and set it to be equal and see if your statement still holds

[vagrant pagoda]

But the reason that a mid-range GPU does not have more width is because they design a GPU in blocks

And in order to get more with on a smaller GPU you would have to redesign the block

[astral haven]

insanely broad statement that has previously been proven false in this conversation

[vagrant pagoda]

Yes and saying that more memory does better dosent do anything for a width argument

They both have the same width

[astral haven]

i think you understand the basic concepts, you just have not fit the puzzle peices in the correct order

[vagrant pagoda]

You have literally never shown that increasing the width while keeping the core count the same improves performance

[astral haven]

they have the same width, and different core counts

[vagrant pagoda]

You also have it shown me a single GPU that goes outside of blocks in it's range

[astral haven]

your claim that the blocks include everything from core to imc is false and has been previously disproven

[vagrant pagoda]

I didn't say that one block contained all of it

It's that it's built in blocks

And if you don't have enough core blocks you can't put any more memory blocks

[astral haven]

this statement

[vagrant pagoda]

Yes

[astral haven]

is false

[vagrant pagoda]

Prove it

Show me a single GPU of any part of a range that breaks that

[astral haven]

burden of proof is with the claimant fam

[vagrant pagoda]

Yes and I'm saying that it doesn't exist

Proved me that it does exist

Prove to me that every single block diagram is wrong

[astral haven]

1060 3gb has the same bus width as 6gb despite different core counts

[halcyon willow]

https://tenor.com/view/michael-jackson-eating-popcorn-enjoy-i-like-nom-nom-gif-18780245

[astral haven]

heres the leap to conclusions right here

[vagrant pagoda]

Because they are different dies

Lmao

One is cut down 1080

On is cut down 1070

[astral haven]

https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-3-gb-gp104.c2926
https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-3-gb.c2867
heres 2 gpus with different dies that perform identically

https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-6-gb-gp104.c3250
https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-3-gb-gp104.c2926
and these two gpus use the same die with the same memory bandwidth to acheive different performance

[vagrant pagoda]

Also I'm pretty sure the only 1060 with a different bus with is the goofy ass 5 GB

[astral haven]

and despite that, the gpus have vastly different performance??
how is that possible surprise picachu

[vagrant pagoda]

They all have the same fp

Other than the ones with less cores

And again they have the same width

[astral haven]

exactly!!!

[vagrant pagoda]

So they same the same amount of cores and they have the same amount of with

So the ratio is the same

They also have the same amount of cash

Almost like they're the same number of blocks

[astral haven]

but they dont have the same amount of cores

look fam this is basic formal logic

[vagrant pagoda]

I'm talking the two different dyes used the fact that they sell the same thing cut down under the same skew doesn't matter

It's the fact they build them in blocks

It is nothing to do with performance

I don't know why you think performance matters in this

I'm saying they build a GPU in blocks

This just is a GPU built with blocks

The fact that they disabled part of those cores in the same skew is just scummy not different

[astral haven]

again you keep saying true things until you suddenly make up a random conclusion is the problem

[vagrant pagoda]

They both have the same amount of cash in the same amount of with because it's the same blocks

They have the same die size

[astral haven]

well that happens to be incorrect in this case

[vagrant pagoda]

No you're just cherry picking

[astral haven]

but it doesnt have any effect on the argument

[astral haven]

cherry picking examples that prove you wrong? yes

[vagrant pagoda]

There are both 1070 and 1080 based 1060s on 3gb and 6gb

They all use the same blocks

[astral haven]

1070 and 1080 are the same die

mostly

[vagrant pagoda]

Not always

[astral haven]

yep

[vagrant pagoda]

It's been a long time since I looked at Turing but there are two different dies for the 1060 one is a cut down of one GPU and one is a cut down of the other GPU

All of them have the same cache and memory width

Because they contain the same number of blocks

[astral haven]

and you are 100% correct in the fact that they use blocks

but the blocks are alot more granular than you think

[vagrant pagoda]

There's just cause lasered off

Again linking two gpus with the same width is not disproving my argument.

You need the same GPU with two different widths

[astral haven]

https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-5-gb.c3060
https://www.techpowerup.com/gpu-specs/geforce-gtx-1060-6-gb.c2862

same core count
same l1
different l2
different bus

[vagrant pagoda]

Yeah because they removed a memory controller/cache block

[astral haven]

BINGO

[vagrant pagoda]

Now show me the opposite and then you'll prove me wrong

I'm saying they can't add more width

Because it's built on blocks that you have to stack

You could definitely take a 4090 and give it one memory controller it would just run like shit

You can take a 7900 XT and put one MCU

You cannot take a 7800 XT and put 8

[astral haven]

subtraction and addition are reversible operations fam

[astral haven]

I cant definetly
prove amd cant

[vagrant pagoda]

Also the 5 GB has all of the silicon there

It's just lasered off

And there is a missing module

With the 3gb they had to laser off a whole sp

[astral haven]

and yet the memory bus is the same size

[vagrant pagoda]

Which is why when you remove that whole block

[astral haven]

but the block doesnt include the imc

3gb vs 6gb and 5gb vs 6gb are proofs in both directions of that fact

[vagrant pagoda]

No but the blocks are 3 sp on pascal

So they removed part of one block

[astral haven]

and here is a claim that requires proof

[vagrant pagoda]

Which is how the other blocks already are.

And by remove I mean laser disconnect

[astral haven]

🤷 still doesnt matter how it happens, the proof is that they are able to do that

[vagrant pagoda]

1060 always had 2 gp with 5 sp on both

The 3gb just got 4+5

For who knows why

[astral haven]

what is preventing them, besides having an iq higher than 80 of taking a 7900xtx and slicing it down to 12cu without cutting down on the memory bus?

[vagrant pagoda]

It would cost the same.

[astral haven]

🤷 thats not a statement that has any bearing on the point

[vagrant pagoda]

Yeah because it wouldn't be mid-range lol

You'd have this product that cost more than a 7900gre that is slower than a 7600

If it would even work

[astral haven]

🤷 it has half the core count of a midranger, and double the memory capacity and bandwidth
the fact that doesnt be good doesnt mean its impossible to create

[vagrant pagoda]

Who knows if the blocks even work that small

[astral haven]

again tho we arent discussing that

[vagrant pagoda]

But the price would put it out of mid-range

It would cost more than a GRE to make

[astral haven]

again bringing statements with no bearing on the argument

dragon range exists and it only has 3 cus

[vagrant pagoda]

Price range is what makes mid-range

Because it's literally middle of the pricing range

[astral haven]

midrange-ness has no bearing on the argument

[vagrant pagoda]

My argument is why there doesn't exist a mid-range card with more with

So price has 100% of the argument

If it's priced out of mid-range then the argument is true

They could completely redesign the entire core but that would cost so much money that it would be out of mid-range as well

Which is literally my argument that they build gpus in blocks and so it's cheapest to just use less blocks

[astral haven]

yep and im not arguing with that
im just proving that its possible, regardless of how monumentally stupid youd have to be to make it exist

[vagrant pagoda]

Redesigning the core for every skew would balloon the cost

[vagrant pagoda]

It is not possible because it wouldn't be mid range

[astral haven]

yep
and being stuck putting a core with a specific amount of IMC would also balloon the cost

everyone learned from maxwell!!!!

[vagrant pagoda]

No it doesn't because that's how they build gpus

They use the same blocks

[astral haven]

ion

please fam

think

[vagrant pagoda]

You are moving goalpost ever statement fam

The only thing I've argued the entire time

Is that you can't have a mid-range card with more width.

Because of how they designed gpus

Which is in blocks

And in order to have a mid-range GPU with more with they would have to redesign the blocks

And that development cost would put it out of midrange

[astral haven]

and my argument is that you are making insane overgeneralization

[vagrant pagoda]

Yes because not everybody here knows how the fuck gpus are laid out

And you do and somehow you're arguing that something exists that doesn't

Every desktop GPU made is built in blocks

[astral haven]

but because of your insane overgeneralization, you are making these false statements

[vagrant pagoda]

There's nothing untrue about the fact that gpus are built in blocks

And that all of the skus in the same stack use the same block

You haven't shown any example of this not being true

[astral haven]

yep, and if you think thats what im arguing against, you havent been paying attention

[vagrant pagoda]

I know what you're arguing against and it's not what I'm arguing for

So I have no fucking clue why you keep arguing something that has nothing to do with the conversation that I had

[astral haven]

and therein lies the problem

[cursive summit]

Still going, ionstorm? What's the conversation even about that it's worth arguing for the last hour for

[astral haven]

you are making incorrect statements, i have refuted them, and you havent provided any reasoning for why the refutations would be invalid

[cursive summit]

Ah, Java 21 is out https://timefold.ai/blog/2023/java-21-performance/

[vagrant pagoda]

So me where I was wrong then. You have never shown a GPU that isn't built using the same blocks as the rest of its family.

The whole desktop rdna3 is the same.blocks in the same ratios

[cursive summit]

Wonder if Minecraft will get any perf boosts from Java 21

[astral haven]

who is moving the goalpost?
i just claimed that you dont know what is contained in a block

i know they are made of blocks

the blocks however are provably smaller than you think

[vagrant pagoda]

The blocks are built out of blocks I know that

But the blocks have to connect to each other

The way that they are designed you cannot have a wider memory bus because there's nothing to feed it into

Do you not remember the 970?

They hung one memory controller off just the crossbar and the performance was fucking horrible with that section of memory

[astral haven]

after this statement you make the insane leap of logic

[astral haven]

start working back from here making direct statements that lead to the conclusion

[vagrant pagoda]

That's just a fact go read a GPU brief

Nvidia has really good ones AMD has terrible ones

[astral haven]

yep
and after that statment you went completely off the rails and into the ocean

[astral haven]

so im rerailing you back where you were correct
Right here

[vagrant pagoda]

Again you're saying I'm wrong without providing any proof

So we're both wrong until one of us shows something

The easiest proof for me is the fact that the 970 exist

If you hang a memory controller off of nothing but the crossbar it doesn't work

[astral haven]

and the many iterations of 1060 prove that they solved that problem

[vagrant pagoda]

How they have this same GPC to memory layout

[astral haven]

with different bus widths on the same core count

or different core counts on the same bus width

[vagrant pagoda]

Yes less bus

[astral haven]

both prove this statement incorrect

[vagrant pagoda]

The whole argument is more

Wider

Not narrow

I know with your intelligence level it's very hard to understand the English language

[astral haven]

6gb is wider than 5gb

[vagrant pagoda]

But narrow means less wide means more

[astral haven]

3gb is wider than 5gb

[vagrant pagoda]

Yes so they took a core and shrunk it

Meaning less

We're not talking about memory size we are talking about widtg

So the 3 GB and 6 GB are the same

The 5 is less

[astral haven]

yet with different core counts

[vagrant pagoda]

All the 1060s have the same amount of gpcs

[astral haven]

that you claim are tied together

[vagrant pagoda]

They are all on the die

They lasered off cores out of the gpcs

[astral haven]

yet they are cut off, and the gpu still functions

[vagrant pagoda]

But the cash is still there so the gpcs are still there

Because the memory controller count is tied to the GPC

You can have less memory controllers but you can't have more

[astral haven]

some of the cache is there, some of the cache is cut off

[vagrant pagoda]

Yes because even the gpcs are blocks

[astral haven]

yep

[vagrant pagoda]

And if you disable too much of the foundation you can't stack more shit

[astral haven]

blocks in blocks in blocks

[vagrant pagoda]

I never said you can't go smaller I said you can't go bigger

You cannot take a GPC and add more memory controllers

[astral haven]

7900xtx is big

[vagrant pagoda]

Yeah and you can't attach 10 memory controllers to

[astral haven]

7800xt could have as many memory controllers as that

[astral haven]

as a bare minimum

[vagrant pagoda]

No because it's less blocks

[astral haven]

and so they can add more blocks if they want

[vagrant pagoda]

I'll bet money when the die shot comes out that there's no connections for the extra memory controllers

But also just look at the scaling of the 7900s

[astral haven]

🤷 and that doesnt mean they couldnt have put them there

[vagrant pagoda]

Don't even worry about the 7800

Look how core count cash size and memory width scales

[ashen spindle]

You can have 12

[vagrant pagoda]

If it was possible it would exist

[astral haven]

yes ion i understand what you are saying, and 90% is correctr
its just you are taking true statements, tieing them together, and then suddenly jumping to an incorrect conclusion

[vagrant pagoda]

It doesn't exist because it doesn't work

The only time it ever happened in the 970 and they got suit over it

[astral haven]

thats the most insane of logical fallacies

[vagrant pagoda]

Because it didn't work

[ashen spindle]

No it doesn’t exist because it was not adding any benefit since the shaders got fucked so the base config was more than enough memory

[vagrant pagoda]

No the extra memory controller for the last memory chip is hung off just the crossbar

That memory controller is not connected to any l1 cache or cores

So the only way to access that memory is through the memory crossbar

Which ties up the memory controller that you're using it for

Because they're built in blocks and there's no way to connect that memory controller to the core without redesigning

Now they redesign the entire chip to have wider l1 that supported more memory controllers than they could have done it

[astral haven]

again extrapolating from maxwell to literally anything designed afterwards is beyond stupid and i proved you incorrect

[vagrant pagoda]

Because they're all designed the same

The ratio changes

But the fact that they are built in blocks doesn't

[astral haven]

1060

[vagrant pagoda]

You can't just hang blocks off of no foundation

Yes the 1060 exist and then they made it smaller

They did not make it bigger

[astral haven]

970 wasnt bigger than 980

[vagrant pagoda]

The 1060 has the same bus to GPC ratio as a XP

[vagrant pagoda]

Gpc to width yes

[astral haven]

wrong

the 970 is to the 980 what the 1060 3gb is to the 1060 6gb

[vagrant pagoda]

192x 2 is what?

[astral haven]

not the same as 160x2

[vagrant pagoda]

What is 192 x2

[astral haven]

320

[vagrant pagoda]

384

How wide is the XP bus again?

[astral haven]

not 320

[vagrant pagoda]

Yes 384

It's almost like the 1060 has half the memory controllers as a xp

Odd

Oh look 1.5 vs 3mb of l2

[astral haven]

so a gpu with 1/2 the titan xp cores and not 1/2 the memory controller would prove you incorrect?

wait a goddamn second
titan xp and titan x pascal have different core counts
🤔 🤣

ooooooh look same memory width

what a shocker 😱

[vagrant pagoda]

Same die

[astral haven]

970 and 980 are same die

[vagrant pagoda]

The 1060 die has 6 gpc

The og one

[astral haven]

or 8gpc

[vagrant pagoda]

No 4

My bad I can't type on phones

I'm elbow deep on a jetski

[astral haven]

oh wait its 4 gpc all the way up and down

1080 down to 1060

tho the 1060 can also have 3

[vagrant pagoda]

No

[astral haven]

oh my bad it has 2

[vagrant pagoda]

Ti nd titans are 6

1080 is 4 which is what the original 1060had

[astral haven]

wrong OG 1060 has 2
"In its full implementation, the GP106 features 10 streaming multiprocessors (SM) with a total of 1280, single-precision CUDA cores, arranged in 2 graphics processing clusters (GPC). Each GPC includes a dedicated raster engine and five SMs with 128 CUDA cores, eight texture units, a shared memory unit, and some L1 cache. "

[vagrant pagoda]

Enaabled and on die are different

The original 1060 was a 1080 die

[astral haven]

wrong
the first 1060 i posted to disprove your point was however

anyways i have a job to do, please think about your statments
if you think a statement i said is monumentally stupid, try to rephrase it in a way that isnt monumentally stupid, because thats probably what i meant...

[vagrant pagoda]

It's a redesign though

[astral haven]

🤷

same exact gpc blocks

[vagrant pagoda]

And same cache

1060 sits on a single block of l2

They never increased the width

[fickle ore]

where is 7800 xt benches

[astral haven]

But the capacity is different for different 1060s

[vagrant pagoda]

All over

[cursive summit]

Hub and GN did theirs

Hubs bench is.. weird

It seems like essentially just a 6800(XT) but on a "better" node and extra software features

But sometimes it gets really left behind even by a 6800XT

You'd be better hoping that the AMD Fluid Motion Frames tech will not suck and give a major boost in all games

Can't wait for this gen to end, it's such an underwhelming generation from both parties

[fickle ore]

easy win on cost/frame yoo

[fickle ore]

ty

[cursive summit]

Well it's better than what Nvidia offers

Nvidia's midrange is way too expensive so AMD is better there

[fickle ore]

if 4070 drop price by 20% I wud get it instead

[cursive summit]

But perf is just sad cause you'd hope a 7800XT would be better than its namesake of 6800XT, but it's even worse at times

[fickle ore]

yeah

[cursive summit]

Treat the 7800XT as a 6700XT upgrade, essentially

[cursive summit]

4070Ti imo is where Nvidia's performance begins to be good

Which is sad cause after that you really only have 4080 and 4090

I have my doubts that Nvidia will make RTX 5000 insane either, since they are allocating resources into AI sector instead

[fickle ore]

its like 5% ahead of 6800 xt

at 1440p which is good enough for me

[cursive summit]

Not amazing since it's about the same price as 6800XT

You really only get the new software features, HYPR-RX and AMD FMF when that releases

Also at times 7800XT loses against 6800XT miserably

[fickle ore]

only DP 1.4 on 6800 xt so no dreaming about those upcoming 480hz oled panels :p

[cursive summit]

They cut those CU's and it's showing

[cursive summit]

I don't know who needs that anyways so, not a big loss?

[fickle ore]

The thing is, it's a variable hz/res panel. It goes 240hz 4k to 480hz 1080p, so u can choose by the game you play

[gusty kiln]

hellhound looks to be the best model

and it's at msrp

[cursive summit]

Oh, 7900XTX is getting the 6900XT situation

Quite a bit below MSRP here in EU

Hope that means it goes down to 700$ next year

4090-ish levels of gaming perf for 700$ would be very appreciated

[gusty kiln]

all the cheap 7900XT are gone at microcenter

[gusty kiln]

more so 4080

[cursive summit]

Eh, it's weird with that card

But 4080 is still 1200$ at least so, even then it's amazing

[fickle ore]

All In all I think 7800 xt is a very nice model. Inflation adjusted price is A LOT less than 6800xt back then. Around 5 % better perf. Antilag+, DP 1.4 -> DP 2.1. Uses 15-25 % less power.

[cursive summit]

Like, it's 7800XT really depends on how you look at it

Compared to last gen? It brings very little new other than software features

But in general, it's good

[fickle ore]

Thing is 500$ now is like 400$ a few years back lol 😛

[cursive summit]

It can game at 4K for 500$

It's really really incredible

[gusty kiln]

AMD subreddit are still whining about the 7800XT

[cursive summit]

But compared to last gen? Underwhelming

So, it's a bit of a weird one

[fickle ore]

Yeah but 500$ is jack shit atm

with inflation

keep that in mind

[cursive summit]

Meh, 500$ is a good price, but when it falls to 400$ and below? Pretty amazing

[gusty kiln]

6800XT stock should dry up any moment now

[fickle ore]

4070 beats 7800 xt easy on power draw tho

[cursive summit]

Didn't watch the reviews on 7700XT, is it overshadowed by 7800XT?

[fickle ore]

yea its garbo

[gusty kiln]

techpowerup

[cursive summit]

50$ difference yet the specs are so massively different, you mainly lose 4GBs of GDDR6 which is sad

7800XT is nice, but I hope for price cuts to 300$ eventually and it'll be great

[fickle ore]

Too bad HUB didnt put 7700 xt here

[cursive summit]

7900XT should eventually settle to about 500-600$

And XTX idk, ~750?

[fickle ore]

U get free starfield if thats ur thing with purchase btw

[gusty kiln]

doubt

it'll be like 700/900

[fickle ore]

Wish I didnt, bc I dotn want that

[cursive summit]

Well, if we will copy what happened to RX 6000 series, XT - ~550 and XTX - ~700

[gusty kiln]

it only happened to the 6000 series because they over produced

during the whole mining craze

[cursive summit]

Yeah the covid effect hit that one hard as well

[gusty kiln]

it won't go down that far until the next gen launch

[cursive summit]

But we aren't seeing any new consumer cards until 2025

I think both AMD and NV are skipping 2024 yeah?

[gusty kiln]

they most likely are just going to refresh

[cursive summit]

Per Herkelman, RX 7000 series is done

So, no refreshes eh?

[gusty kiln]

only the GPU dies are done

they can make any SKU they want

[cursive summit]

I saw a very interesting thread today

Where if you give a 7900XTX more memory bandwidth and raise the power target, it will beat 4090 in raster

Could be the eventual 7950XT

[gusty kiln]

that's old news

and needs like 600W

no one wants that

[fickle ore]

https://www.youtube.com/watch?v=AdWv3GwS-kY

[cursive summit]

Maybe they could bring back the ol' dual-GPU tech and combine 2 7800XT dies into a megaGPU

Look, I just like the dual gpu thing alright

The fact they did one for Apple not too long ago made me happy, too bad it'll never come to desktop, well, it's old already anyways

[fickle ore]

https://www.verkkokauppa.com/fi/product/892648/Sapphire-PULSE-RX-7800-XT-Gaming-16-Gt-naytonohjain

whats the cheapest atm? found 600€

[cursive summit]

RX 7800 is not a thingy

[gusty kiln]

get the hellhound if it is available

[cursive summit]

https://www.ldlc.com/fiche/PB00569976.html

569

[gusty kiln]

quietest one with still decent cooling

[fickle ore]

609€

[gusty kiln]

looks good to me

[fickle ore]

gotta wait for some german shop deal

[gusty kiln]

crazy that it competes with the Nitro+

[cursive summit]

@fickle ore Hellhound for 568 https://www.computeruniverse.net/en/p/90932874

For same price you can also have Sapphire PULSE and some jiggybyte

[gusty kiln]

yeah, written is nice to view at work

[fickle ore]

sO UHH

Is hellhound better than a pulse

[cursive summit]

Wait so why is the Hellhound the best one?

Sapphire usually is king

[fickle ore]

RNG

[gusty kiln]

it's quieter and cooler

according to techpowerup's roundup

[fickle ore]

Powercolor has lower RMA rate afaik atleast a week back there was statistic from some rando shop

[cursive summit]

Hellhound does look cute

Dual slot 322mm card? That's weird

322mm is too big for some ITX cases already

AMD absolutely loves to make their cards long

[gusty kiln]

they basically took the 7900XT hellhound cooler and made it thinner for the 7800XT

so the cooler is a bit overkill for the 7800XT

which is nice

[fickle ore]

noice, thats longevity

[cursive summit]

Hellhound looks to boost less than other cards

Although a few extra Mh is not much

[gusty kiln]

probably using stock PCB

red devil will boost higher

[cursive summit]

Ayyy Yeston is doing their thing again

Man these are cute

[gusty kiln]

they never been able to top the cute pet

[cursive summit]

Ngl, I didn't really like the cute pet

The colours are not my favourites, and the design was a bit uncute?

[cursive summit]

Ok, now THAT is hella cute

https://vi.aliexpress.com/item/1005005880234947.html?algo_exp_id=014d0e2c-d6d4-4693-babc-849f182a44bb-34

Only one cooler on a 4060?

These designs are still just adorable

You can prolly make a handbag PC with such sizes

Ay, Fym, bestie, how have you been

[gusty kiln]

great for deshrouding

smol

[cursive summit]

R9 Nano might have some proper competition

I will tear you to shreds then

[gusty kiln]

wat, 16" MBP is a great laptop if you can work with MacOS

[cursive summit]

What's a carmen?

The opera thing?

[gusty kiln]

universally praised for the hardware as well

best laptop speakers on the market

[gusty kiln]

bon

[cursive summit]

As much as I dislike apple and will potentially never buy anything from them

Macbooks are really well designed and built

Especially the low tier options for sub-2000$

The high tier ones? Very disappointing value

I just would need them to embrace proper dedicated GPUs for gaming to even consider switching out of Windows

[gusty kiln]

if macs could game, I would be all in on apple atm

the ecosystem is just too nice

[cursive summit]

Although ngl Win11 left me so sad that I would rather MacOS torture myself than look at Windows 11 design ever again

[gusty kiln]

https://www.xreal.com/

get one of these

me and carmon enjoy using em

[cursive summit]

Do you need a portable device?

[gusty kiln]

it comes with a lens that you can take to your eye doctor

[cursive summit]

Cause I dislike laptops myself, only desktops

But those macs are so badly priced

Mac minis are fine though, if only they had GPUs at the same price point

"personal life and professional pentesting".. is there a difference?

[coarse cloak]

LOL

[gusty kiln]

kek

I can see 379 msrp after a few months

[steep creek]

Congrats on the Mac, hope it brings you lots of benefit

[gusty kiln]

the bigger question is...

how's the notch?

[steep creek]

Very good

[gusty kiln]

I would do the blackout thing if I have it tbh

[steep creek]

Yeap worst case I heard you can also create a fake black bezel in the settings

[gusty kiln]

@prisma crag Lower Decks is back this week

[reef patrol]

I hate that it's a weekly show. They somehow manage to end every episode on a frustrating cliffhanger

[steep creek]

hah

[gusty kiln]

new Castlevania at the end of the month as well

[reef patrol]

This meme has been made with only the finest of pigments.

[humble hazel]

PC:
-Z790 Gaming Pro Wifi
-I9 13900K
-RTX 4070TI Aorus Master
-1000W Pure Power 12M
-32GB DDR5 (6000MHz) Corsair Vengeance RGB
-NZXT H7 RGB Flow
-NZXT X63
-500GB NVME M.2 (Atlas OS)
-1TB HDD

Monitor:
-Alienware aw2521h 360hz
-Philips 242V8

Headset:
-Sennheiser GSP 600
-Epos GSX 300 with Epos H6 Pro

Tablets for Osu!;
-Wacom CTH 680
-CTH 480s
-CTL 472

Mouse;
-Razer Viper Mini

Keyboard;
-Alloy Origins Red Switches
-2K Keyboard all mx
-Wooting 60HE

Microphone:
-Shure MV7x
-Scarlett Solo 3.0

what u guys think about my setup build it today

[reef patrol]

SMH making a huge post, pushing up my amazing meme.

[fickle ore]

rose to 583, not cool

[cursive summit]

?

Still the same price

[humble hazel]

its ddr5 that is why bro ❤️

[steep creek]

local conversions i guess

[fickle ore]

for me its 583, must be atuomatic update for VAT or some shit

[cursive summit]

excl. VAT

[fickle ore]

wtf??

[cursive summit]

EXCL. VAT

[gusty kiln]

6000 is also kinda slow for a gaming intel system

[steep creek]

aha i missed no vat, nevermind then

[fickle ore]

yo u wanna buy me Ill pay 500€? Make 30€ extra

😄

[cursive summit]

EXCLUDING TAX

[steep creek]

import duies as well would make this cost a lot more lol

[fickle ore]

is that non eu shop

[cursive summit]

Computer universe?

[fickle ore]

y

[cursive summit]

Compuni is germany based

EU customs free shipping

[steep creek]

issue is im not in the eu

[cursive summit]

Too bad, too sad

https://tenor.com/view/so-sad-oh-no-trump-sad-gif-20568390

[steep creek]

but my neighbor country is, ive read of people who just import to them and go pick up at the border lol

[cursive summit]

Not sure why Amazon has no new cards yet

[fickle ore]

NBB is 569€ me but mfrs want me to mke an account https://www.notebooksbilliger.de/powercolor+rx+7800xt+hellhound+16gb+grafikkarte+814000?nbbct=4001_geizhals

[humble hazel]

❤️

[fickle ore]

@gusty kiln how big was the hellhound btw?

My lian li lancool II mesh performance: VGA Length 384mm

[humble hazel]

Just for Osu but the hardware prices are low at the moment

[gusty kiln]

think 322mm

[fickle ore]

should fit I g

[gusty kiln]

add another 25 for fan, so 347mm in total

yeah should fit with a little wiggle room

[fickle ore]

;_; but 600€

[reef patrol]

Essentially. Bit of hand soap or dishwasher soap maybe.

Soft cloth or just rub with hands and fingers. Wiggle carefully to get most water off, and leave to dry bottom side up for a while.

[sharp badger]

MT/s are probably more Accurate measurement of ram but MHz is fine, also with ram speeds of the Intel setups some people seem to have some problems with running 7000+ stable if they haven't fixed it

[crisp granite]

7700XT being significantly better than the 7800XT in averages
and the 7700XT having the highest 1% lows

it's absolutely weird

[vagrant pagoda]

Linus isnt used to the D

[gusty kiln]

https://www.techpowerup.com/313309/cooler-master-raises-the-bar-with-the-q300l-v2-matx-case

https://www.youtube.com/watch?v=hYDmcSE9KII

Looks like COoler Master finally took GN's advice

[fickle ore]

speed holes?

[gusty kiln]

Yeah, bigger holes

Though the porosity still doesn't look that great on the new front panel

Holes aren't close enough

[fickle ore]

can you call it porosity when there's nearly more holes than matrix

[gusty kiln]

kek, I feel they can squeeze another row of holes in each the X/Y direction

[fickle ore]

Lol maybe

On another note.. Wheel of Time >Season 2 has arrived

gg's

[vagrant pagoda]

so

i have the atx version for my portable vr rig

i just cut wholes and slapped fan grills

yeah

[gusty kiln]

https://xtia.design/cdn/shop/products/1-1_a5a498bc-5ea9-4898-9ac7-872027a14250_750x.jpg?v=1641314140

^ATX

[vagrant pagoda]

pic

thats for the aio

280

[gusty kiln]

https://xtia.design/cdn/shop/products/9-2_750x.jpg?v=1641314144

AIO suffering, but looks cool for a test bench

[vagrant pagoda]

if i throw that in my back seat

it would die

portable atx

here was mocking up the other fans and hardware

[gusty kiln]

thermaltake took that orientation and ran with it

CTE cases

[gusty kiln]

I'm guessing these were spare parts?

that's why ATX?

[vagrant pagoda]

they make it in atx

same outside other than rear panel

gn reviewed the atx

q500 atx, a300 matx

same size

i know because i got a matx one second hand that was damaged for the handles

handles

[vagrant pagoda]

look at the front io

it isnt even on right

not atx psu

[gusty kiln]

yeah they honestly made it more boring, looks fixed now

that case is way more expensive

those cooler master cases were like under 70 at the time

[vagrant pagoda]

the 500 is bigger

but can fit dual rads and all pcie slots

and a atx psu

still nice case

yeah blocks bottom slot with the rad

not even the fans

and gpus block front rad

it was slightly to small for me

i do

i had 2

now it has one

but i need last slot

are you just linking because they look cool?

it was a list

sfx allows top rad

so atx = no top rad

long gpu = no front rad

need bottom pcie slot = no bottom rad

nvdia founders maybe

will a aorus master pcb even fit in there?

[hasty thunder]

well, at least a pure white option now for AMD GPUs

but i'd imagine they'll charge a pretty expensive extra for that

[steep creek]

beautiful cards, though my case is under the desk in a mostly black box so i wont notice it much lol

[hasty thunder]

idk, doesn't seem beautiful to me, especially when i know it's basically a Pulse under the hood... Their Pulse line up is their least supported one

[steep creek]

think its because ive mostly only seen black gpus, so this one seems striking at the moment

[hasty thunder]

There's the Gigabyte Vision

[steep creek]

yep that looks nice, i also liked the aorus white one

[gusty nexus]

What's the problem with their Pulse line up ? 😅

[hasty thunder]

poor QC, even worse RMA support

[gusty nexus]

oh I see 😅

[hasty thunder]

i.e. my RMA for a 5600 XT Pulse got denied, because i damaged the "Warranty Void" sticker on a screw. Despite the fact that these are illegal and can not be enforced upon european citizens

[reef patrol]

Then reply and point that out. And when that fails, threaten them with consumer rights filings in court.

[hasty thunder]

That's the reason why i had to open it in the first place

for some unknown reason there was solder on the copper cooling plate and i've sanded it down

[reef patrol]

That uh... I hope you didn't tell them that.

[hasty thunder]

the GPU then died, because my first X570 Aorus Master had a fault

[hasty thunder]

ofc i didn't tell them

but the QC must have been absolute garbage if such a thing doesn't get recognized

[haughty vector]

can someone help me rq in wooting60he channel

[hasty thunder]

like you can clearly see grey, sanded down spots

and that can't be copper to begin with

anyways, i had sued them and they didn't appear to the court date. So the court punished them with 500€ for damages

since them i'm on their blacklist

can't buy any products anymore from them directly

[vagrant pagoda]

lmao

they wouldnt rma it?

[hasty thunder]

nope

[vagrant pagoda]

what reason?

[hasty thunder]

^

[vagrant pagoda]

must be a eu thing then

never had issue with them in us

ive send pcb and cooler seperate and got rma with sapphire

[hasty thunder]

thing is that their warranty void can't do shit

i'm allowed to disassemble my products for trouble shooting and even fix them as long as it doesn't cause any noticeable changes like component changes or further damage

but yeah, i'm avoiding Sapphire as best as i can since that incident

6800XT Aorus Master Type C, because it was the only USB C supporting 6800XT

and so far i'm not regretting that choice, because it almost got into the top 100 best 6800XT's on 3dmark

despite being purely cooled on air

the only reason it didn't become a top 100 GPU were 2 small clock drops from 2798 MHz to 2680 for a second

[vagrant pagoda]

on hwbot?

[hasty thunder]

but yeah, it's a damn solid GPU, got freaking lucky in the silicone lottery so far

[hasty thunder]

3Dmark in total

[vagrant pagoda]

no one submits to 3dmark

check hwbot

[astral haven]

Hype!!!!

[astral haven]

Hwbot scores are screwed up thanks to the rules
They are entirely different scores

[ashen spindle]

you say hype I say its just a steam deck die shrink like ps5

[hasty thunder]

you're automatically submitting to the 3dmark database as long as 3dmark notices an online connection

[astral haven]

Galileo?

[astral haven]

The hwbot scores can't automatically be submitted since it makes modifications to the 3d settings

[hasty thunder]

tell me something new pls

[ashen spindle]

family I think is just everything and sepiroth might be a die shrink

maybe its deckard im just coping

[astral haven]

It's still 3dmark, but you will not see the high end overclockers on the regular scores outside of hwbot
The professionals will submit to both for sake of completeness tho

[astral haven]

Why is it called "valve" tho?
Normal socs would be called AMD in their patches

[ashen spindle]

I think the jupiter entry also says valve, idk

might be because valve does actual work on SW side for it and its not amd alone

[ashen spindle]

it's a valve SoC so its theirs to submit is what a distro dev said to me

[astral haven]

Nah amd was who submitted it

They just put valve codenames on it rather than their own

[ashen spindle]

¯_(ツ)_/¯ I dont linux so I take the word of someone who works on it

[astral haven]

Oh wow Bradley has a pretty interesting theory

[ashen spindle]

did not get around to watch it yet

[astral haven]

its live now

as in a livestream

[ashen spindle]

ye I know still in VC we just finished raid so gotta watch later

[astral haven]

bradley's idea is deckard will actually just run deck games natively

just with alot larger screen and ar and stuff

[ashen spindle]

hmm neat idea

[astral haven]

supporting vr ofc, but also heavy focus on updates to the steam deck compositor with vr supportish

and arm support

[fickle ore]

low gfx vs ultra gfx

[vagrant pagoda]

https://www.youtube.com/watch?v=PpxMKtbI8G8

so it draws more power

gets less frame

and cost 50 bucks less

where is that supposed to be

[astral haven]

rip
no iphone for me

fkn 85mm

real chiplets from intel ahoy

[rough osprey]

what do the other cameras have

[vagrant pagoda]

https://www.newegg.com/asrock-radeon-rx-6800-xt-rx6800xt-pgd-16go/p/N82E16814930049

how can a 7800/7700 compete with that XD

[astral haven]

my s21 has a 240mm telephoto and a 75mm telephoto

[rough osprey]

is 240 the 10x?

[astral haven]

yup

[rough osprey]

so the trash one

haha

[astral haven]

🤷 it can see better than my eye

[rough osprey]

true probably

its useful for taking pictures of the board in lectures

😄

[astral haven]

no as in its literally better than my eyes
around 2k pixels horizontally in 8 degrees fov
thats around 250PPD

thats above the optical limit of the retina

oh wait no physical limit of the retina is 300 ppd

highest measured human ppd is 240

20/20 vision is 80 ppd
and average human vision is around 120 including glasses

[fickle ore]

except it doesnt see anything, it just a camera

[astral haven]

that is not a provable statement

however if I see, it sees

[final flame]

Gotta love vending machines, I got charged 5$ for a bag of pretzels, I'm guessing they're gonna give me back the rest of the money once it gets fully processed

[fickle ore]

ur eyes are again the limiting factor when ur looking at the feed of the camera.

[astral haven]

also no, the magnification from the taking of the photo magnifies it to a point where its easily visible to me

[final flame]

Nah, when I'm checking focus on a camera I always magnify the preview display to the max, then apply the focus

[fickle ore]

yes but ur still limited from the theoretical max view clarity, ur not a hawk

[astral haven]

also this is assuming 2k pixels,
the camera is 3648x2736, and the fov is calculated horizontally

[astral haven]

again no
you need to be aware of a thing called magnification

[fickle ore]

naw

[astral haven]

if i take x that is too dense for me to resolve, and magnify it up to a larger fov, i can then resolve it

[fickle ore]

resolve to what clarity?

Can you see the viruses lurking in the magnified photo?

[astral haven]

329 PPD over 8.3 degrees horizontal

[fickle ore]

yea so cucked

imma sleep bbl

[astral haven]

better than your vision lel

[cursive summit]

https://www.macrumors.com/2023/09/06/app-store-safari-and-ios-designated-gatekeepers/

Good

[ashen spindle]

yay on package ram for no upgrade options and nearly no gain /s

[fickle ore]

But this convo about whether camera can see was actually interesting

[cursive summit]

Apple has been getting leashed by the EU quite a bit

[gusty kiln]

oh yeah still wondering why MSI was left out

[ashen spindle]

they are making it later just like with the original launch

why idk, maybe just how they allocated stuff

[final flame]

Lessgo, can't wait for the Macos through virtual machines on ipads

[hasty thunder]

Not even worth reviewing, ngl

the 7800XT is just a 6800XT in disguise with a lower power consumption and lower launch price

[astral haven]

solid disagree
everyting needs to be reviewed in the end

[ashen spindle]

I think you are putting too much on the name and ignoring names dont matter but prices do

[hasty thunder]

it's literally like the 3060 vs 4060

[ashen spindle]

if a 5080 would be a 4090 exactly in everything but 300$ that would be actually insane

the 4060 literally had a higher msrp than the 3060

[astral haven]

https://tpucdn.com/review/amd-radeon-rx-7800-xt/images/relative-performance-rt-2560-1440.png
also a 7800xt is alot faster in RT
faster even than a 6900xt

[ashen spindle]

wait nvm got 60ti results mixed int here

still not a lower price for same cost though

anything beyond high end is usually the same as last gen at cheaper prices

so just because its not called a 7700 this time doesnt make it a bad product or not interresting

[astral haven]

well $500 isnt as exciting as it could have been

[gusty kiln]

kek 500 used to be flagship

[astral haven]

that was a looooooooooooong while ago

so far back that you actually need to take inflation into account otherwise you are super duper off

[gusty kiln]

GTX 580 was 499 in 2010

[astral haven]

oh my bad last 500 flagship was 480 in 2010

[astral haven]

wasnt the flagship

[gusty kiln]

really?

oohh

Dual gpu

[astral haven]

590 was $700

[ashen spindle]

whats 500$ in 2020 where the 6800xt came out for 650?

[gusty kiln]

wow 700 for dual gpu is kinda nice

[ashen spindle]

or what would 650 be today from 2020 if the calculator only goes one way

[gusty kiln]

that's basically the current lowest 7900XT pricing

[astral haven]

yup

[ashen spindle]

so the 7800xt is a 500$ card replacing a 750$ card (sorta, not really but kinda)

[astral haven]

yup

[gusty kiln]

the only question is how much 6800XT and 6950XT stock is left

[ashen spindle]

probably not much thats why its launched now

[final flame]

Lol

Apple making up every possible excuse not to implement NTFS support

[astral haven]

https://play.google.com/store/apps/details?id=org.futo.voiceinput

the voice input app for android that rossmann was promoting is out now!

[final flame]

That's cool, I'm assuming it can completely replace Google assistant

[astral haven]

no

[final flame]

Oh it's only for voice to text

[astral haven]

yee

full local STT

multilingual too apparently?

gonna try it out after work

sure my coworkers wont want me saying a bunch of random stuff while they are working

[final flame]

Just found this

[astral haven]

samsung has it as part of the onboarding, when setting up an apple you can only migrate from another ios during the OOBE

[final flame]

Interesting

[final flame]

I don't like transferring stuff, I like having a fresh start when I use a new phone

Just transfer all the personal files over onto my PC and start fresh on the new device

[astral haven]

yea no dice, i have hundreds of apps that use on a monthly basis

17 of these icons are folders

do i just happen to have 42 icons on my homescreen?

not including the widgets

wow funny coincidence

[final flame]

Most of these are my bare minimum

I'm using a a14, so entry level soc, didn't want to bog it down with a bunch of crap

Still running as fast as when I got it, so I think I'm doing good

[astral haven]

yea ive got a ton of stuff running in the background

full (user accessible) filesystem backup to google

[astral haven]

Okay, so let's give this dictation app a shot. Let's see how fast it is and how good it is.

Okay, unfortunately it's not real time so you can't preview what you're dictating while you are talking to it, but once it's done, it just pops into the box.

Wow, that is actually extremely impressive. I would be shocked if it made it just to understand what I'm saying now by my obfuscation of what I am saying.

Wow, it is completely nailing it. 100% so far, including all of my mistakes.

There was someone who I was chatting with ages ago in this chat who really needed a stt app. Does anybody happen to remember their name?

[vagrant pagoda]

8% in RT isnt worth 20% everywhere else lol

[astral haven]

Ain't the same price tho is it?

[vagrant pagoda]

yep

https://www.newegg.com/asrock-radeon-rx-6800-xt-rx6800xt-pgd-16go/p/N82E16814930049?Item=N82E16814930049

6800xt for 499 with starfield

ironicly starfield is one of the few titles the 7800xt is faster

[astral haven]

6900xt

The 7800xt is faster than the 6900xt in rt

[vagrant pagoda]

its 8% faster than a 6800xt in rt

read what i wrote and you responded to.

[astral haven]

Oh I see
Yeah that still doesn't make it any faster in rt

[vagrant pagoda]

i didnt say it was

6800xt is faster in some games

slower in others

for the same or less money

[astral haven]

oh my god @ashen spindle (sorry random person i pinged first) valve pushed a steamvr update while bradley was livestreaming

[sharp badger]

Yeah I see what you're saying. DDR5 6000MHz -> 3000MHz because of the cycles. But 6000MHz(Clock rate) & 6000MT/s(Effective Speed) in this context is the same. Should the manufacturer use the latter? Yes

[shadow minnow]

Same network same speed same location, my Slovak SIM card always calls it LTE and the UK one always says 4G strange

[icy cedar]

can anyone suggest me one keyboard between two? im thinking of getting the tuf k1 and vapoo v500, which one should I go for? both are same prices

[modern sphinx]

I'd go for Rapoo purely on the basis that it's a mechanical keyboard, whereas the K1 is membrane

[misty saffron]

what was that app somebody had recommended here to stream music from ios to pc?

nvm probably any airplay one

[cursive summit]

https://www.eurogamer.net/nintendo-demoed-switch-2-to-developers-at-gamescom

A more reliable source talking about these rumours

Although, Breath of the Wild Remaster for Switch 2?

That sounds... Easy to do and ez money, could be interesting

[rough osprey]

i can't wait

[fickle ore]

gime switch2 botw2

[rough osprey]

mariokart 2

oh wait

[icy cedar]

alright then

[crimson wigeon]

omg give me starfield

[rough osprey]

starfield on switch when

[crimson wigeon]

I'd like to see that

[rough osprey]

just port it

should go fine

[ashen spindle]

If the Stemdeck can’t run it the switch won’t, I don’t think new switch has better hardware than steamdeck

[rough osprey]

believe in Nintendo

[crimson wigeon]

Ninty magic

[rough osprey]

a higher end stationary console would be dope from them

need da 4k60 Mariokart

[cursive summit]

https://www.videogameschronicle.com/news/sources-nintendo-showed-switch-2-demos-at-gamescom/

Another VGC source claimed that Nintendo showcased Epic’s impressive The Matrix Awakens Unreal Engine 5 tech demo – originally released to showcase the power of PlayStation 5 and Xbox Series X in 2021 – running on target specs for its next console.
The demo is said to have been running using Nvidia’s DLSS upscaling technology, with advanced ray tracing enabled and visuals comparable to Sony‘s and Microsoft’s current-gen consoles.

UE5.. DLSS... Ray Tracing.. Switch 2???

[reef patrol]

Switch TWO?
nintendo_executives_going_nuts.meme

[cursive summit]

I mean, this has been known for a long time

[crimson wigeon]

Switch U

[rough osprey]

3Switch

oh that doesn't work

[reef patrol]

I know. It was just the name of it and all.

PC? Where is PC2?

[rough osprey]

where is wooting two

oh wait

[reef patrol]

[digital cliff]

I'm on PC3 myself, you gotta upgrade friend

[reef patrol]

[rough osprey]

windows 12 next year

finally

[misty saffron]

windows 12? how about

[rough osprey]

im sry

[cursive summit]

Sad part it's probably not even a joke

Windows barely made 11 work, isn't that thing still in beta?

And 12 is coming up, already

11 is just a few years old

[rough osprey]

they probably not gonna redesign everything again

just changing stuff a bit and adding Features

[crimson wigeon]

More ads

[rough osprey]

ads?

[reef patrol]

Yes.

[rough osprey]

i only have seen the "preinstalled" apps if you can call that ads

[reef patrol]

Under EU law, yes

[gusty kiln]

Still got 2 years on W10 before eol

https://www.reddit.com/r/Android/comments/16bivne/google_pixel_8_and_pixel_8_pro_euro_pricing_leaks/

RIP people in the EU wanting a Pixel 8

[vagrant pagoda]

PC3 is old, you don't atleast have PC4?

[gusty kiln]

Only PC4? PCP is the latest and greatest

[vagrant pagoda]

https://www.reddit.com/r/Android/comments/16c1kef/mediatek_successfully_develops_first_chip_using/

[stark loom]

Hey I’m new to the topic of keycaps and profiles.. but would the cherry profile keycaps be usable or compatible on the Wooting two he?

[vagrant pagoda]

yeah

[vagrant pagoda]

profile dosent really matter for the keyboard, layout and stem dose

profile is prefrence

[gusty kiln]

Profile only matter for north facing switches

Since cherry profile interfere with the caps

[stark loom]

Coz I was doing some research and read that Wooting ships originally with OEM profile

[gusty kiln]

Though a lot of newer switches avoid this problem

[vagrant pagoda]

there are north facing swtiches now

they will fit cherrys

[stark loom]

So as long as I have the entire set with the cross stems, cherry profile is good yea?

[vagrant pagoda]

but lekkers are slim compared to mech swtiches

[gusty kiln]

Yeah

Cherry profile should work

And lekkers are reversible on the pcb anyway

Ezpz

Buy any caps

[stark loom]

Okies, thank you guys..

[ashen spindle]

Oh looks like 7800xt can get decent overclocks that also scale

[knotty summit]

You lose RGB

[stark loom]

Meaning that if I change my lekker switches to south facing rgb is lost but the led still lights up?

[marsh glade]

It is almost never necessary to rotate the switches and we strongly recommend against doing so unless you have demonstrated that your specific set of custom keycaps has an issue

[stark loom]

Got it, so unless there's obvious interference we are advised not to rotate the switches..

[marsh glade]

Correct

[stark loom]

[mild dome]

https://tenor.com/view/it-prints-money-switch-nintendo-money-iwata-gif-10140242

[old marlin]

https://reddit.com/r/MouseReview/s/arxsBitYIB

the better g pro superlight 2

[knotty summit]

RGB is there but you will lose good portion of it. It will be only on the plate basically

[hardy idol]

If anyone ever asks why I don't want a perforated mouse:

https://reddit.com/r/pcmasterrace/s/wGiw5lDBNU

https://tenor.com/view/vomit-puke-ugh-gif-25512927

[halcyon moss]

Same... It seems so obviois

[hardy idol]

Tomorrow my new mouse arrives

And it's not perforated luckily

[wheat plume]

In 5 years all mice will be perforated
Except really expensive hand made ones cause non perforated is the niche now

[ashen spindle]

that claim doesnt even make sense why that would happen

[wheat plume]

:') I know it won't

Only for gestures...

But not for actually moving a cursor

I use them a lot also

The freaking touchpad gestures in gnome desktop
😍

[steep creek]

i got too accustomed to having a touch screen lol, i use it a lot on the laptop. but i prefer a mouse to be honest

[wheat plume]

The only time I wish my laptop was touchscreen was going through that list of cookies you need to manually disable one by one on some websites

[steep creek]

ahah, i just find its quicker to touch rather than use the trackpad, when the cursor is far away

[wheat plume]

Makes sense

[vagrant pagoda]

https://www.youtube.com/watch?v=5-rr_8JqIPU

[wheat plume]

Really??

[ashen spindle]

would make sense, steam deck cpu is hurting and this one has a stronger one

[vagrant pagoda]

and the ally runs it

z1 is just same soc binned

i hate the only the ally dose freesync

at 30 fps its super nice to have

and helps a ton

freesync with autotdp on the ally makes a big difference

i wish i didnt hate fps on ctroller more

so i could get a handheld

[wheat plume]

Using gyro on the Deck makes playing fps games fairly accurate for me

I bind activating it to left trigger so gyro is only enabled when I'm aiming

[thick elk]

Personally prefer trackpad aim but I end up arriving at the same issue as wooting analog has, namely a lot of games don't liked mixed controller and mouse inputs

[wheat plume]

https://www.techspot.com/news/99979-microsoft-accused-malware-like-tactics-again-attempt-push.html

[outer monolith]

got this notification earlier today

[pale sigil]

I've had that before

Lame

[crimson wigeon]

Hell yeah where do I sign up?

[hardy idol]

Definitely not

And if so

Watch me with my glue gun

[crisp granite]

There is a GitHub repository for emulating those gestures on windows

I use a Magic Trackpad with my windows machine when I’m not gaming

[cursive summit]

https://twitter.com/necrolipe/status/1699822182690439253

'Necro' Felipe #UnivNintendo (@necrolipe)

Pelas minhas fontes agora: A versão do DLSS exibida a portas fechadas pela tech demo do “Nintendo Switch 2” era 3.1 e não 3.5 como reportado ou apontado pela Eurogamer

Ray-Tracing de fato é possível e a memória RAM que eu fui informado era de 12 GB para o consumidor

(Rumor)

Likes

193

Twitter · Sep 7, 2023, 4:29 PM

ayo

Switch 2 has DLSS 3.1 and 12GB of RAM

What kind of supercomputer are they cooking up

Switch 2 sounds incredible already for Nintendo

[chrome estuary]

Isn't that basically a steamdeck

But Nvidia and not amd

[thick elk]

it will be arm based though and we don't know how powerful, just that it uses a more modern nvidia soc at that point.

[chrome estuary]

Oh yeah good point

[vagrant pagoda]

You mean Thor?

[cursive summit]

They aren't using NV Thor lol

[vagrant pagoda]

XD

you said supercomputer XD

[cursive summit]

It just sounds like an incredible upgrade over what they already have

Proper 60FPS 1080p gaming on the Switch

That doesn't look horrible

Imagine they also pull a major 360 on their ecosystem and make it a "Windows-like", so you can build your apps for it

[vagrant pagoda]

will prob be a ampere based soc

they have a bunch in the jetsons

[cursive summit]

Yeah it will be Ampere based

Which tbf is already quite old

[vagrant pagoda]

https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin/

[cursive summit]

Leaves a lot of room for a refresh model, or a Pro

[vagrant pagoda]

old

[reef patrol]

Heh... Baidu's clone of chatgpt will crash if you start asking it ... Problematic questions.

[vagrant pagoda]

its one gen back

[cursive summit]

Orin's already old

Like, closing in on 5 years

[vagrant pagoda]

????

they are ampere based now

and came out last year

and they have more skus this year

[cursive summit]

What do you mean

[vagrant pagoda]

the newest orins

are brand new

[cursive summit]

Yeah but Orin was announced 5 years ago

[vagrant pagoda]

and?

how long do you think it takes to design a gpu core?

they take a long time

nvidia knows when each gen will come out years in advenaced

[cursive summit]

There's still headroom there for them to adopt Thor, if needed

[vagrant pagoda]

orin is ampere

[cursive summit]

Ada Lovelace is a pretty big improvement over Ampere

[vagrant pagoda]

so its one gen back

[cursive summit]

Although I wonder if Orin will be able to use frame generation

While being on RTX 3000 series card arch