Anyone impressed - or surprised - with what AMD is showing off at the 2022 CES? Do the new products stand up well against Intel’s newest chips or those in the offing?
See https://www.techspot.com/news/92835-amd-ces-2022-roadmap-zen…
Anyone impressed - or surprised - with what AMD is showing off at the 2022 CES? Do the new products stand up well against Intel’s newest chips or those in the offing?
See https://www.techspot.com/news/92835-amd-ces-2022-roadmap-zen…
Anyone impressed - or surprised - with what AMD is showing off at the 2022 CES? Do the new products stand up well against Intel’s newest chips or those in the offing?
The 5800X3D launch date surprised me. As did Zen 4 desktop being announced as 2H2022 vs. 3Q2022. That is no big deal. I expected an announcement in September with an October FCS date. This means that when the 58OOX3D date finally arrives, I’ll need to decide whether to buy one or wait for Zen 4. Right now if I need to test the difference between Zen 2 and Zen 3, I can run a program on my son’s PC. Hmm. Note that for most of what I do, my current system is ridiculously fast. What I need from time to time is to run tests against AMD’s most recent silicon. That discussion about testing MACs shows where I’m coming from.
Back to the 5800X3D. I suspect this is the only 3D part AMD will announce for the desktop or laptops. Servers are a completely different story. A lot of Milan (or earlier) CPU customers will be upgrading or trying to upgrade to Milan X. In a server or HPC system the payback should be in weeks, not years. With (non-hyperscalers) looking at Zen 4 (Genoa) as about a year out, if you need more CPU horsepower, Milan X is a no-brainer. Of course, if you have a 32-core server, and the maximum current load is say 28 cores? Wait for Genoa.
Note that my thinking on this is shaped by geologic/seismic processing. The extra cache will make feasible resolutions that are currently a bit too painful. Getting the resolution you want may require trimming the volume you are looking at. And as always, twice as fast tends to result in four times the demand. (For some reason these companies don’t share much, insist on non-disclosures, and won’t use the cloud. Most have seismic data-crunching machines isolated from the web. Found a new field, and need more CPU? They don’t want anyone guessing at that from cloud usage.
MLID is predicting that the Ryzen X3D parts will not be able to retake the gaming crown from alder lake, and that is why they have been deprioritized. AMD needs to race to 5nm and zen 4. Ryzen X3d will still be a great upgrade CPU for some.
I agree the server market is much more interesting than the client market we hear about at CES.
It looks like the first Intel Alder Lake mobile products will be gaming and workstation laptops with the higher power “H” series, likely coupled with an ARC GPU. Intel mentioned something around 50 design wins so far. Intel ARC doesn’t have to be the highest performing GPU out there, but it needs to be stable, bug free, and with adequate driver software.
Alan
MLID is predicting that the Ryzen X3D parts will not be able to retake the gaming crown from alder lake, and that is why they have been deprioritized. AMD needs to race to 5nm and zen 4. Ryzen X3d will still be a great upgrade CPU for some.
While I have the utmost respect for Morse Law Is Dead, that statement does not make sense to me.
Did you note in the slide where AMD claims the 5800X3D has 15% better gaming performance, the claim was in reference to performance against the 5900X in 1080p gaming. It was not compared to the 5800X.
Here is a comparison between the 5900X and Intel’s top i9-12900 from Tom’s Hardware
https://www.tomshardware.com/news/intel-core-i9-12900k-vs-ry…
The $549 12-core Ryzen 9 5900X is AMD’s fastest gaming chip, but the $589 Core i9-12900K is 8.7% faster in the cumulative 1080p gaming measurement across our entire test suite.
So this data would indicate that the 5800X3D should take the crown back by a 6.3% margin (15%-8.7%).
AMD’s later slide also showed it trading blows with the i9-12900K and on a cumulative basis coming out on top. (I would like to see the slide deck end notes but have not found them yet)
We will have to wait for reviews to be sure but, at this point I feel like there is more evidence pointing to AMD successfully taking back the gaming crown than not.
This in my mind also supports why we are only seeing one Zen3 3D part as well. Just enough to nock Intel off the top of the gaming hill and then onto Zen 4 for the real fight.
I agree we need to wait for reviews. The comparison will also likely depend on the choice of GPU cards.
Note that at CES Intel introduced the 12900KS, which clocks about 5% higher than the “K” part.
Alan
I agree we need to wait for reviews. The comparison will also likely depend on the choice of GPU cards.
Note that at CES Intel introduced the 12900KS, which clocks about 5% higher than the “K” part.
True graphics cards have the biggest impact on gamming performance. Although one of the advantageous of comparing the CPUS at 1080p rather than 4K is you get a better feel for the impact the CPU has. At 4K you are testing graphics cards. At 1080P the graphics cards are nowhere near maxed out so the differences in the CPUs stand out.
The gamesmanship between AMD and Intel is hysterical.
Intel claims that their upcoming i9 will take back the gamming crown.
AMD releases information that their upcoming 3D CPU, the 5800X3D, sees 15% better performance in games [AMD casually leaves out the words “when compared to the 5900X”]
Intel now has a target to stretch the clocks and thermals too. Intel sets the bar and releases the i9-12900K.
The truth about what AMD was really saying leaks out and Intel figures out they are probably 5-6% behind the 5800X3D. So Intel comes up with an s version 12900K which happens to push the clocks enough to eke out another 5% in performance. (possibly S for stupid hot and damn the consequences?)
Meanwhile the world sees the 5800X3D and the i9-12900KS unveiled at CES and we all wonder, which one is really faster.
…all for a gaming crown. LOL
While I have the utmost respect for Morse Law Is Dead…
It’s Moore’s Law, for Gordon Moore. https://en.wikipedia.org/wiki/Moore%27s_law
I think this is the footnote you are looking for:
Based on testing by AMD as of 12/14/2021. Performance evaluated with Watch Dogs Legion, Far Cry 6, Gears 5, Final Fantasy XIV, Shadow of the Tomb Raider and CS:GO. All games test at 1920x1080p resolution with the HIGH in-game quality preset (or equivalent). System configuration: Ryzen 7 5800X3D and AMD Reference Motherboard with 2x8GB DDR4-3600. Core i9-12900K and ROG Maximus Z690 Hero motherboard with BIOS 0702 and 2x16GB DDR5-5200. Both systems configured with GeForce RTX 3080 on driver 472.12, Samsung 980 Pro 1TB, NZXT Kraken X62, Windows 11 28000.282. – R5K-107
Thanks for the footnote. That’s what I was wondering about. The 5800X3D wins while running 16GB of DDR4-3600 memory against the i9-12900K running 32GB of DDR5-5200. Those results look to be solid and I expect them to be confirmed by the reviews.
I believe that is the second time that spell check has tripped me up on Moorse’s Law. Although, as we all know my spelling does not keep up with the pace of my mind so it might have been me as well. Sorry for the error.
I believe that is the second time that spell check has tripped me up on Moorse’s Law.
Would that make #3?
Would that make #3?
LOL. Nope that one was all me. Spell check just gets me when it switches it to morse code.