- Mar 3, 2017
- 1,747
- 6,598
- 136
This one at 4.5GHz is a good comparison between M4 as it clocks at 4,5GHz too. It doesn’t look like it’s an ES sample as well.There are even more.
ASUSTeK COMPUTER INC. ProArt P16 H7606WI_H7606WI_002009218 - Geekbench
Benchmark results for an ASUSTeK COMPUTER INC. ProArt P16 H7606WI_H7606WI_002009218 with an AMD Ryzen AI 9 HX 170 w/ Radeon 880M processor.browser.geekbench.com
20-25% if you remove SME-outlier subtests?Crazy how far ahead Apple is in 1t while also remaining stable.
It’s scores ~3700 as well in GB6.2 which doesn’t include SME.20-25% if you remove SME-outlier subtests?
There are no 6.2 Strix results so not comparable.It’s scores ~3700 as well in GB6.2 which doesn’t include SME.
Does Intel have a Halo competitor?There are no 6.2 Strix results so not comparable.
Either way STX1 is still a relatively cachepoor baby, real whacko mobile 1t race is where parts like STX-halo crawl into the ring, with full 32M L3 and some MALL on top of that.
No.Does Intel have a Halo competitor?
Nvidia too.Looks like AMD and Apple are the only ones
ARL-H 683 with ADM existed once and obviously vanished without trace.I guess Intel would have to get their GPUs to scale properly first.
Just like the mythical M1 Exteme lolBut hey, it was supposed to have 384EUs.
It was a good idea to shelve it. Using a Alchemist as the GFX IP would be a disaster in the making.ARL-H 683 with ADM existed once and obviously vanished without trace.
But hey, it was supposed to have 384EUs.
Ask the same question to the same AI tomorrow and you might surprise by the quite possible different answer.
I love AI, lol.
I am getting just over 3100 ST on a stock 7950X with 6400C30 tuned ram.I think 5.1 GHz Strix could approach 3050 in David Huang's setup.
I got as high as 3030 in GB6 6.2 on a 7950X with DDR5-6000 memory (Windows 11, VBS/core isolation turned off for the benchmark), so that would have the 7950X already beat (or matched), but perhaps not if I was running Linux too.
Assuming +16% IPC for same-clock 9950X, that would just about push up score 3500 points in single-thread in Windows... of course, it may scale differently in reality.
You can run CachyOS x86-64-v4 version, Phoronix found several applications which benefited from AVX512 too, you can compare with Zen 5 later. Application that uses a lot of cryptography benefited a lot from it (including mining )I'll run on some flavor of Linux when I have a chance tomorrow, it's a more fair comparison of hardware to hardware if we want to make estimates for zen 5 and inevitably compare to Apple.
We are still oscillating between 5% and 50%. The only thing worse than Zen 5 leak accuracy is Samsung 3nm yields.So what's the consensus here, is zen 5 good yet?
ES runs looked very mediocre.So what's the consensus here, is zen 5 good yet?
-5%, fixed for ya.
Are you running air cooler or just AIO? I was under impression that properly clocked 7950x can do about 1.4 gflops if thermals can be sustained (one of my friends tested 7900x with custom water loop at release time and he managed to score 1.1 gflops or about that)I recovered my old config. It only improved the rate marginally
Maybe both are true at the same time, some benches are 5% while others are upwards 40% 🤔We are still oscillating between 5% and 50%. The only thing worse than Zen 5 leak accuracy is Samsung 3nm yields.
Ask and you shall receive. It's a 4+6 chop tho.There are no 6.2 Strix results so not comparable.
Either way STX1 is still a relatively cachepoor baby, real whacko mobile 1t race is where parts like STX-halo crawl into the ring, with full 32M L3 and some MALL on top of that.
I would like to have a Zen 5 Ryzen 9 please, with 2x 6 wide decoders, 8 ALUs, 8 FP pipes, 6 Giggle Hertz and please make it 25 watts.Ask and you shall receive
Just take zen4 and double almost everything , it would have be awesome....I would like to have a Zen 5 Ryzen 9 please, with 2x 6 wide decoders, 8 ALUs, 8 FP pipes, 6 Giggle Hertz and please make it 25 watts.
Nah. This approach leads nowhere, as showcased by the Sunny and Golden Coves.Just take zen4 and double almost everything , it would have be awesome....