#offtopic_hardware
38 messages · Page 25 of 1
Interesting problem with m$ defender and Intel pre 12th generation CPUs https://www.techpowerup.com/forums/threads/windows-defender-can-significantly-impact-intel-cpu-performance-we-have-the-fix.295877/
lol
I doubt casual people will ever realize that bug
Btw, you should post that in software channel too, some people probably wonna know it since Defender is being used by a large majority of freeware users these days already
5.8 is more realistic 🙂
also something in your article that I don't believe
"max clocks(1T)" is ~5.8 GHz for both amd and intel
and nobody believes that
especially not with the v-cache variant
AMD needs both an insane cooling solution for the cores and seperate voltage for the cache for that to be a reality
in other words
your article is full of shit 🙂
It is likely that Intel will retain the prices of its chips moving forward. Intel only bumped the pricing up by $50 US when they doubled the core counts from 8 on Core i9-11900K ($539 US SEP) to 16 on the Core i9-12900K ($589 US SEP). Since Raptor Lake isn't a doubling and is mostly a refinement of Alder Lake, we can expect the pricing to remain the same and at most, hit the $599 US mark. AMD's flagship is likely to cost around $700-$800 US which means that Intel may have a big edge in price to performance segment.
if you even try to digest this pile of shit
good luck to you
EXCITED FOR AMD NEXT GEN!!
https://framechasers.org/ - Discord and Support
https://framechasers.org/consulting - Consults
https://twitter.com/ChasersFrame - Twitter
ht...
This assumes that 7900xt will be the highest model 🙂
judging by what you're writing, I can conclude, that you haven't watched the video
Welcome to the LIAN LI 2022 Digital Expo 2.0!
This is Part 1 of our 2 part series. Stick around for part 2, to see more updates, new product announcements, and a giveaway at the very end!
Watch Part 2: https://youtu.be/efxxP9tOwwk
Enter the LIAN LI 2022 Digital Expo 2.0 Giveaway: https://gleam.io/fdZ6r/lian-li-2022-digital-expo-20-giveaway
F...
navi 1 had bad drivers, this isn't news for anyone with a brain
I suppose CPUs became thinner and Mainboard designs did not consider that at all. Reminds me of iPhone6 bending w/o real pressure
I hope this problem is not on all mainboards… Don‘t wonna take the extra effort first to buy in EU just to get return right covered for this bullcrap
It’s not about the cpu becoming thinner, it’s about the cpu getting larger as there are more pins, and intel uses a design from a previous era meant for far less pins.
In 12th gen intel jumped from 1200 to 1700 pins
Almost 50% increase without increasing the locking mechanism to match
Oh I didn‘t pay attention to the pin numbers. The increase is indeed large
They cheaped out on the locking mechanism
The same mechanism used by $750 12900ks on a $2000 asus Maximus board is shared by a $70 pentium with $70 motherboard
AMD has an appropriate socket design so they don’t have this fix
And now for another fun fact
Der8auer worked with TG to develop this fix into a product
There are much cheaper alternatives on websites like aliexpress
Ali recently removed listings of the cheaper mechanisms for German visitors to the site
It’s not even something you can patent
1800 pins physically present and only 1700 of them are used by Alder Lake.
all of 1800 will be used by Rocket Lake
so 1800 pins vs. 1200 pins is 33% more
0 issues for me, since I'm not using the original locking mechanism.
actually I have nothing that locks the CPU inside the socket.
so no bending of anything, no problems with memory channels
Nice
You didn’t watch the video
Then you’d understand what I mean by rotations
no bracket/frame used
my waterblock is what pushes down my CPU.
so pressure is optimal and even
why would I spent money on a piece of metal (bracket/frame) if there is no need
watercooling 🤮 only water cooling I need is on my cnc machine
If you have an airconditioned room with stable temp., it is no problem to stay on air cooling, regular overclocking is no problem too
It depends
Air cooler will heat soak much faster, about 90 seconds when appropriately sized.
Meanwhile a 240mm AIO can take 200 seconds to heat soak
420 AIO might take 370 seconds
And of course custom water cooling will have longer time (versus AIO) to heat soak for a variety of reasons, such as copper rad, thicker rad, pure water instead of ethylene mixture, better tubing and pump for higher flow, better cold block for better transfer.
Lets rent a gaming room inside one of those former nuclear power plants and just activate their cooling system again lol
The cooling towers are epic
With a veritable spate of published patent applications over the past week, AMD has potentially revealed a lot of new insights into upcoming technologies in GPUs and CPUs. On June 30 alone, 54 patent applications were published. It remains to be seen which patents will ultimately be approved and which will actually be found in products. Regardless of this, the applications provide interesting insights into the technological approaches pursued by AMD.
of particular interest for the article, is the patent regarding better shader utilization for MCM
And Intel is being pushed down the patent ranking each year lol
Nvidia has a major oversupply problem that may delay Lovelace, & Intel has delays as well…
[SPON: Use “brokensilicon” at https://shakerandspoon.com/brokensilicon for $20 off of a sub!]
0:00 I bought an RTX 3090 Today…
1:00 Nvidia Ampere Oversupply Leak – Prices are NOT done dropping!
5:15 Important advice for GPU Shoppers
6:15 Because of RDNA ...
I see you are a fellow man of culture yourself.
The guy is absolutely underrated IMO
He has had in depth leaks that turned out correct so many times by now, no idea why his videos dont get more views
Because the entire market is based on what you wanna believe. The others either don‘t care about leaks or busy with their own opinions 😜
Because a lot of them were wrong as well and then fanboys got burned and angry with him
he’s regularly made fun of by people with actually good info
No idea why they are desperately focused on leaks, they should make good news in general first
Fun page of improvements in game from RAM OC on vermeer and vermeer x https://github.com/xxEzri/Vermeer/blob/main/Guide.md#performance-gains-from-ram-overclocking
AMD fixing their software, shocking
holy
1080 ti light is flashing orange slowly, I don't know what I did to make that happen.
why dont you ask on tomhardshardware, linus, your manufacture forum...
maybe even read the manual
What is the best headset for playing arma?????
the most comfortable one, without usb and open back design
Like astros?????
lol no
AKG, Audio Technica, Sennheiser, Beyerdynamic, comparable companies like them, Hyper x is the best low priced option
avoid the RazorLogischit what have you
Steelseries
samething as razorlogicshit
It’s funny how much effort intel will go through to advertise e cores while enthusiasts do not care about e cores, if anything they are more e-waste for enthusiasts.
Userbench is heavily biased
To the point they’re frequently banned from being posted on tech forums
Also
The 12600k will be better for games
But it doesn’t make much sense over a 12400
12400 is substantially more cost effective if what you care about is in game performance per dollar
well I keep my computers for a long time so a 10 core cpu seems pretty future proof, and would help me with multitasking
what is better benchmark site
Depends on the workload
Told ya, still same old habit. They did not communicate in advance, software industry is still not ready for this, Intel realized a decade later „oh hey, mobile CPUs have this amazing power efficient concept, lets copy and push everone“
And depending on what we understand of „enthusiasts“ I see them as much as overclockers, most of them are everything but sane. The whole mobile business works on e-cores already and some of them even have 3 different cores. So yes, there are proof of concept starting with Apple Bionic, Qualcomm Snapdragon or ARM and in terms of efficiency neither Intel nor AMD would stand a single chance. That is why they are heavily investing in mobile gear to not lose more grounds.
Pawgers 😮
Official Geekbench account responded saying these results are probably fake because it looks exactly like 12900k at 5.5
Unless of course the additional cache is useless, then it’s not fake 🙂
or you're simply not smart enough to understand that it was running on Z690 with no official support and not Z790 with full support and that it's not final, thus performance is not as it should be
and saying that additional cache is useless is also stupid, because it can simply be that a specific program doesn't take into account the added performance coming from more cache, like Cinebench doesn't care at all whether you have DDR4 2133 or DDR4 4700 MHz - results in exactly same number of points
Damn that’s some impressive copium