- Mar 3, 2017
- 1,614
- 5,856
- 136
One should also see how limited is the single core clock in a mobile (premium) chip compared to the desktop version. I suspect not much, but not negligible. Again, if these numbers are real and not speculation.
Ah, OK, I misread.That s the improvement relatively to the 7945HX wich is the mobile version of the 7950X and is also displayed in this alleged leak, so mobile vs mobile.
The limitation is certainly not big enough to make up for the RTG claims of 3000 points CB score. While i would want those higher numbers to be true, i am more inclined to believe these lower, less dramatic ones.One should also see how limited is the single core clock in a mobile (premium) chip compared to the desktop version. I suspect not much, but not negligible. Again, if these numbers are real and not speculation.
I have absolutely no doubt RTG's claims are incorrect (his track record is one of the worst, with a lot of backtracking). What I'm more interested is how true are those SPEC leaks.The limitation is certainly not big enough to make up for the RTG claims of 3000 points CB score. While i would want those higher numbers to be true, i am more inclined to believe these lower, less dramatic ones.
I have no preconception, I was merely asking if there was some limitation due to the mobile platform. CB scores in any cases depend more on the FP side (which should have improvement too but no one stated how much).The limitation is certainly not big enough to make up for the RTG claims of 3000 points CB score. While i would want those higher numbers to be true, i am more inclined to believe these lower, less dramatic ones.
The limitation is certainly not big enough to make up for the RTG claims of 3000 points CB score. While i would want those higher numbers to be true, i am more inclined to believe these lower, less dramatic ones.
Doesn't look too good for "40% core-for-core faster than Zen 4 in SPECint 2017" claims, if true.
CPU die using N4X? Not N4P? That's a first for a client part I believe.
Doesn't look too good for "40% core-for-core faster than Zen 4 in SPECint 2017" claims, if true.
Can you explain why? For those of us who realise that not all programs are made the same, and SPEC isn't one of the programs listed in the tweet, and also saw that adroc previously said that cinebench wouldn't show the full increase... and cinebench isn't a measure of IPC, and you have no idea what power envelope and GHz this chip was running at, etc etc etc.Doesn't look too good for "40% core-for-core faster than Zen 4 in SPECint 2017" claims, if true.
SHIIIIII ION EVEN REALLY KNOW LIKE DAT CUHHH ON GOD HOP OFF HIS NUTSCan you explain why? For those of us who realise that not all programs are made the same, and SPEC isn't one of the programs listed in the tweet, and also saw that adroc previously said that cinebench wouldn't show the full increase... and cinebench isn't a measure of IPC, and you have no idea what power envelope and GHz this chip was running at, etc etc etc.
Can you please explain why, in words and numbers, not feelz. Thanks.
He can't.Can you explain why? For those of us who realise that not all programs are made the same, and SPEC isn't one of the programs listed in the tweet, and also saw that adroc previously said that cinebench wouldn't show the full increase... and cinebench isn't a measure of IPC, and you have no idea what power envelope and GHz this chip was running at, etc etc etc.
Can you please explain why, in words and numbers, not feelz. Thanks.
Because that "leaker", if you could even call him that, has y'all gaslighted into believing that Cinebench and SPECint 2017 give vastly different performance per clock gains.Can you explain why? For those of us who realise that not all programs are made the same, and SPEC isn't one of the programs listed in the tweet, and also saw that adroc previously said that cinebench wouldn't show the full increase... and cinebench isn't a measure of IPC, and you have no idea what power envelope and GHz this chip was running at, etc etc etc.
Can you please explain why, in words and numbers, not feelz. Thanks.
I thought it was only Geekbench that correlated well with SPEC.For a new architecture, it seems that Cinebench tracks well with SPEC. Anandtech found a mere 2 percentage point difference
That sounds like an exciting place!So if this rumor from Chiphell forums turns out to be correct, he would have to eat crow.
Cinebench tracks well with SPEC?Because that "leaker", if you could even call him that, has y'all gaslighted into believing that Cinebench and SPECint 2017 give vastly different performance per clock gains.
For a new architecture, it seems that Cinebench tracks well with SPEC. Anandtech found a mere 2 percentage point difference between the two:
AMD Ryzen 5000 and Zen 3 on Nov 5th: +19% IPC, Claims Best Gaming CPU
www.anandtech.com
So if this rumor from Chiphell forums turns out to be correct, he would have to eat crow.
For a new architecture, it seems that Cinebench tracks well with SPEC.
Yeah, since CB tracks so well with SPEC, server vendors should just use CB exclusively since it's free and takes far less time to run, amiright? Who needs industry accepted SPEC tests anyways.Cinebench tracks well with SPEC?
Why are you guys only talking about 16C Zen 5 + 40 CU RDNA3.5 for Strix Halo? That's only the top configuration. There will certainly be cut down parts.
If Strix Halo uses Zen 5 CCDs, it means there is a possibility to have only one CCD; 8C Zen5 + 40 CU RDNA3.5 SKU. Such an SKU would be perfect for a gaming laptop.
There should be efficiencies in APU from BOM, power efficiency, memory efficiency. But this has been covered number of times.Why not just use a discrete GPU for a gaming laptop? That will give you the newer RDNA4 feature set and dedicated VRAM.
A major selling point for an APU is that it makes building a simple low-end gaming PC a lot easier. Since you don't have to assemble the laptop yourself, do you care if it uses an APU or has a separate GPU?
There should be efficiencies in APU from BOM, power efficiency, memory efficiency. But this has been covered number of times.
Yes - for comparing against an older uarch with a uarch that is supposed to be a "grounds-up redesign" like Zen 3 over Zen 2.Cinebench tracks well with SPEC?
Two reasons - first one I've already explained. The other one being the fact that Zen 2 and Zen 3 are on the same node.Any reason you posted the charts for Zen 3 over Zen 2 instead of Zen 4 over Zen 3
Ever heard of correlation =/= causation?Yes - for comparing against an older uarch with a uarch that is supposed to be a "grounds-up redesign" like Zen 3 over Zen 2.
Two reasons - first one I've already explained. The other one being the fact that Zen 2 and Zen 3 are on the same node.
Ever heard of a chief CPU architect getting excited for an unreleased future product being hyped up as revolutionary?Ever heard of correlation =/= causation?
I'd definitely recommend giving the phrase a quick Google search if I were you.
so explain why a FPU based workload , corelates with integer based workloads to the point of it invalidating a consistent rumour of integer performance.Ever heard of a chief CPU architect getting excited for an unreleased future product being hyped up as revolutionary?