Maybe it didn't work like that with Z4, but some very loud people complained about how multi-ccd cpus are consuming 30-40W of power in idle, so AMD obliged and modified power-saving algos with this huge latency increase as a side effect. Maybe the reviewers can re-test the latencies (or our resident Zen5 owners like Dom) with JEDEC / XMP enabled and some specific IF power-related settings like DF Cstates off in the CBSIf that was the case, then Zen 4 should show the same behavior, but it’s unique to Zen 5.
Yea, a Zen 4+ which has been in development for 6 years and released just 22 months after Zen 4.they should call it just Zen4+, like Zen1 and Zen1+
Maybe it didn't work like that with Z4, but some very loud people complained about how multi-ccd cpus are consuming 30-40W of power in idle, so AMD obliged and modified power-saving algos with this huge latency increase as a side effect. Maybe the reviewers can re-test the latencies (or our resident Zen5 owners like Dom) with JEDEC / XMP enabled and some specific IF power-related settings like DF Cstates off in the CBS
Amateur hour lolWhelp, we have confirmation the review guide was a mess. They had to message reviewers later and tell them the dual CCD parts need special attention. You can click on the image to open timestamped GN 9950X review.
View attachment 105294
what a pathetic launch by AMD. Seriously, learn some marketing skills AMD.
At least with User Benchmark you know the results before you read.With all of these mixed results, I think I'm going to hold out for the userbenchmark review before making any decision.
And what good did that do them? Everyone who looked at the slides is now confused and disappointed. And they're not selling anything lolMarketing knows what they are doing they fudge the numbers just enough so that can't be taken to court over it. i'm sure legal reviews this stuff before it gets sent out lol.
Is it just me or are the reviews turning out actually worse than expected? I mean, last week CPUs were underwhelming even toward the rather mild expectations, but it was kind of assumed its down to the strict TDP limit. That being no issue with dual-chiplet CPUs i thought they would at very least perform in line with the AMD marketing slides, but its more of 9700x situation anyway?
No new cores counts until IOD resign so Zen 6 this has been known forever.AMD should have added more cores.
AMD's job to ensure that their driver doesn't get installed incorrectly or malfunctions enough to mess with CPU performance.maybe some reviewers not having it installed correctly.
AMD's job to ensure that their driver doesn't get installed incorrectly or malfunctions enough to mess with CPU performance.
So AMD now requires core parking to be on for non-X3D dual chiplet parts. Blows my mind. They are bungling this launch so bad, only Intel is worse by comparison.
I'm having a hard time understanding why the performance is next level in Linux while being underwhelming in windows...
I thought AVX optimisations was mostly handled at the compiler level?As one of the many guys that writes software that everyday folks like you use, I disagree.
I think it's why AMD extended the review embargo. They tried to fix these inconsistencies in performance, failed spectacularly and in the end, had no choice but to lift the embargo as promised. I bet there are people/teams there having a very bad few days/weeks ahead until the problems get sorted out.
I think someone here is jinxing these launches.X-Elite Windows on Arm launch was pretty ugly as well. There seems to be a common thread forming here