Just a hunchSeriously why do you think they can not?
https://www.theregister.co.uk/2018/04/18/cray_amd_epyc_cs500/
Just a hunchSeriously why do you think they can not?
I see no IPC advantage for intel any more. They have the frequency advantage only. If an AMD chip with 4.2 ghz freq can compete (can score a little less/a little more) with a 5.2 ghz intel I see that as a win for AMD. Also anandtech benches are worthy of a look if you havent already
I see no IPC advantage for intel any more. They have the frequency advantage only. If an AMD chip with 4.2 ghz freq can compete (can score a little less/a little more) with a 5.2 ghz intel I see that as a win for AMD. Also anandtech benches are worthy of a look if you havent already
I'd take anandtech's benches with a grain of salt, they showed the 2700X matching or outperforming the 8700k on every game they tested. Not that it wouldn't be nice if it could, but being realistic there's something wrong there.
I'd take anandtech's benches with a grain of salt, they showed the 2700X matching or outperforming the 8700k on every game they tested. Not that it wouldn't be nice if it could, but being realistic there's something wrong there.
I lost around 5% on benches after the last lot of micro-code patches on my 7820X. I have since compensated by OC to 4.5 base clock with 4.7 on the best 2 coresThey did say that, rather than relying on prior testing, they made sure that all of the CPUs -- Zen+, original Zen, Kaby Lake & Coffee Lake -- had all Spectre/Meltdown patches (both microcode & motherboard) applied, plus all motherboard BIOS updates & all Windows 10 updates. So it's possible that a combination of those updates (possibly not done on other testing sites) might be the reason for the discrepancy...
That would only skew test if the CPU was getting too hot. Which is only likely if testing included over-clocks. And even then it would only be the over-clocks that was effected. Just because one fan is 90mm, doesn't mean it cools better than the other 70mm. A the three heat pipes you are throwing stones at are cooling from both ends. That is essentially six heat pipes the way they have it designed. And lets not forget the Intel OEM cooler you was conpairing it to doesn't have heat pipes at all.The anandtech benches are screwed-up for the intel CPUs because they used a lousy Silverstone AR10-115XS for their test bed.
Interesting, reviewers claim both those games are very well optimised. In fact we even see a Vega 64 almost matching a 1080ti and that game is DX11 only!Yeah, IF you play the poorly optimized assasin's creed or far cry 5. Remove those from the benchmark and the 2700X is faster.
Interesting, reviewers claim both those games are very well optimised. In fact we even see a Vega 64 almost matching a 1080ti and that game is DX11 only!
Oh wait of course you are claiming games that dont run well on Ryzen are optimised because you are a massive AMD fanboy and everybody knows it!
Do you think a Ryzen 2700 is worth $100 more than a 1700 out of curiosity?
Interesting, reviewers claim both those games are very well optimised. In fact we even see a Vega 64 almost matching a 1080ti and that game is DX11 only!
Oh wait of course you are claiming games that dont run well on Ryzen are optimised because you are a massive AMD fanboy and everybody knows it!
Do you think a Ryzen 2700 is worth $100 more than a 1700 out of curiosity?
Doesn't seem like it's fast enough to justify the purchase. More like Ryzen 1st generation re-issue. AMD is at a disadvantaged with software developers because everything's optimized for Intel and Nvidia chips. It'll take another die shrink to realize any gains.
ACo is definitely not optimized. Take the benchmark run for example. It maxes my 8700k @ 5.0ghz yet the difference between a 7700k @ 5ghz is ~5 fps. +2cores + 4 threads at 100% utilization for 5 frames. "optimized"
Nah
Yes, you are partially correct. The problem though is that ACO maxes out a 6c / 12 th even though it gives no extra fps compared to 4c/8t. That's not really a sign of good optimization.That really depends on how you look at the results - perhaps a 7700K is close to the 8700K because the GPU is the main limiting factor in framerates? Or perhaps 4C/8T or 6C/6T is the point of diminishing returns for this game, and throwing additional threads at it only yields minimal gains.
I say this because in Techspots own testing, there is significant improvement from a 1500X to 1600X (4C/8T vs 6C/12T) and also a 7600K to 7700K (4C/4T vs 4C/8T), however the difference between a 1600X and 1800X is minimal, as is the 7700K to 8700K.
Yes, you are partially correct. The problem though is that ACO maxes out a 6c / 12 th even though it gives no extra fps compared to 4c/8t. That's not really a sign of good optimization.
Yeah, Crysis 3 does, although you can't call it popular any more. Also watchdogs 2 seems to use those extra threads and produce extra frames, unlike ACo. Probably the multiplayer segment of BF1, although I'm not sure how many threads it uses, it seems to be up there.Do you know of any other games that scale much past 6 cores? I can't really think of any off the top of my head. I don't really count AOTS as hardly anyone plays that, I mean like popular AAA titles.
Do you know of any other games that scale much past 6 cores? I can't really think of any off the top of my head. I don't really count AOTS as hardly anyone plays that, I mean like popular AAA titles.
The witcher 3 is pretty multi-threaded. You don't really want games using more than 6 cores on an 8 core system though. Typically you'll want to have headroom for the OS and the game as load and background tasks do vary. You can see that in 2700X vs 8700K streaming benchmarks, where when streaming 12mbps quality video causes the 8700K to drop nearly 70% of it's frames on the viewer's side while the 2700X will only drop 7%.
I'd only expect more games to scale past 6 cores when 10 core becomes mainstream.
My bet is Lisa Su is listening to the comments - and is not happy with everyone yelling WTF - you said it would be better. At least, if she is listening, her street savvy is as great as her technical savvy.Just want to say I am disappointed with VEGA
Know a few people who do after effects for movies and they were pretty excited by the 1800X, to be fair a lot of them do run MAC but they really wanted this change.
I was really disappointed by Vega as well. Raja over promised and we only got a so-so gaming video card. I don't know what he was thinking by adding a high bandwidth cache controller to the GPU that takes up so much space but it hasn't helped AMD in non-gaming segments and it doesn't help gamers at all.