That's strong considering my 3960x is like 33k MT (haven't ran ST) with 24 cores (running stock without pbo atm - Only had it a few days).
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
i9-12900K (QS) > Ryzen 9 5950X in ST, MT
- Thread starter Dave2150
- Start date
That's strong considering my 3960x is like 33k MT (haven't ran ST) with 24 cores (running stock without pbo atm - Only had it a few days).
The IPC on Zen 3 is a lot higher, they are powerful CPU's, gerardfraser scored 16,553 on his at 4.825Ghz, LtMatt score 32,400 MT, 1,715 ST on his 5950X, i can't get mine close to that...
A slight exaggeration of scores there.
5800X, PBO, Negative Curve 15, +100Mhz.
MT 15,936
ST 1,631
Aye without even any tweeking at all from my TR it does more than enough. Looking forward to 5000 series TR and will probably slot one in!
Associate
- Joined
- 29 Jun 2016
- Posts
- 543
My 2p is that the ST score is a bit too high for this to be believable. The reasoning is that R20 isn't very sensitive to memory latency or data rate. If these ST claims were made in a gaming workload, I'd be more accepting, where DDR5 could really improve performance there.
So unless Intel have widened the architecture and the branch prediction is _really_ good, it just seems far-fetched.
Would absolutely LOVE to be wrong, but will wait and see.
So unless Intel have widened the architecture and the branch prediction is _really_ good, it just seems far-fetched.
Would absolutely LOVE to be wrong, but will wait and see.
I'm leaning toward it being accurate, perhaps a little on the high side, i'm constantly reminded about a decade + of "leaked Intel benchmarks" having +15% performance over previous gen that then actually turned out to be 3% in reality.
However, they are using Cinebench and comparing it to Zen 3, that's quite different to the past, so for that reason i think it would be foolish to stick with my cynicism and with that i'm going to air on the side of true, so, its 25% per core faster than Zen 3, if it turns out anything less than +15% over Zen 3 in independent reviews, like 12% i think my trust in anything predictions or leaks related to Intel will die completely and i honestly think at this point anyone who has been watching this space for half as long as i have also really need to take off their Intel tinted shades.
Edit: What worries me is that this is some sort of AVX-512 Cinebench hackery that's going on here. In which case this is the only place you will see this level of performance. Cinebench once again becomes the marketing poster child for Intel. At least until Zen 4.
However, they are using Cinebench and comparing it to Zen 3, that's quite different to the past, so for that reason i think it would be foolish to stick with my cynicism and with that i'm going to air on the side of true, so, its 25% per core faster than Zen 3, if it turns out anything less than +15% over Zen 3 in independent reviews, like 12% i think my trust in anything predictions or leaks related to Intel will die completely and i honestly think at this point anyone who has been watching this space for half as long as i have also really need to take off their Intel tinted shades.
Edit: What worries me is that this is some sort of AVX-512 Cinebench hackery that's going on here. In which case this is the only place you will see this level of performance. Cinebench once again becomes the marketing poster child for Intel. At least until Zen 4.
Last edited: