21–30 of 317 posts
Earlier quoted context omitted.
This is an undocumented feature, from what I gather. So it's missing the "bait" part of "bait and switch".
It's not even undocumented, but explicitly a documented as a feature those CPUs don't have. Originally it was supposed to be fused off physically, I wonder why it wasn't.
Also see Linus’s rants in this thread: https://www.realworldtech.com/forum/?threadid=193189&curpost...
And the discussion: https://news.ycombinator.com/item?id=23809335
Today only benchmarks and HPC workloads use AVX-512. I’m sure Intel is happy to force HPC customers to pony up for data center CPUs.
How much controls do users have over microcode updates? In other words, how hard is it to avoid or roll back an user-hostile update like this?
You can apply them or don't at your own discretion.
The problem is knowing which particular update includes the change.
Earlier quoted context omitted.
This is an undocumented feature, from what I gather. So it's missing the "bait" part of "bait and switch".
It's not even undocumented, but explicitly a documented as a feature those CPUs don't have. Originally it was supposed to be fused off physically, I wonder why it wasn't.
Earlier quoted context omitted.
This is an undocumented feature, from what I gather. So it's missing the "bait" part of "bait and switch".
I'm not sure if benchmarks that are out have the AVX-512 enabled or not. If they do then this is indeed fraud. They can always claim that those were engineering samples and not meant to be tested, but, I think they knew. But: that's assuming a lot. It's possible that the ES chips sent to reviewers also had AVX-512 disabled, and that benchmarks did not make use of the instruction anyway.
No, because half the cores on those don't even have AVX-512 in the first place. (to enable AVX-512, you have to disable all the eCores)
As such, running those systems with AVX-512 turned on was academic, but not practically used.
Since Linux loads the microcode on it's own, can this be used to apply an old microcode or is there a prevention against downgrading the version?
Earlier quoted context omitted.
> now Intel is fixing that That's not a fix. That's a disabling. A fix repairs something.
Bringing a product in line with specification is very much fixing it. Even if the fix happens to be disabling something.
Besides the obvious benefit of first having benchmarks out there claiming these chips are better than they really are. So this change just benefits Intel, and nobody else. If it would be a fix then it would be that something that was advertised did not work, and now it does.
How much controls do users have over microcode updates? In other words, how hard is it to avoid or roll back an user-hostile update like this?
Depending on the OS, microcode updates are under full user control. You can apply them or don't at your own discretion. The problem is knowing which particular update includes the change.
I’m not surprised. A lot of people were burned last year when their entire CPU downclocked as soon as one application started using AVX-512. That killed all interest even before Alder Lake. Also see Linus’s rants in this thread: https://www.realworldtech.com/forum/?threadid=193189&curpost... And the discussion: https://news.ycombinator.com/item?id=23809335 Today only benchmarks and HPC workloads use AVX-512. I’m sure…
From what I've heard, there's also a PS3 emulator that benefits greatly from it and its users are quite annoyed by this [0].
According to the sources, users would need to disable E-cores and preferably HT as well to get the best results, though. Pretty niche and not helpful in general (since disabling up to 50% of the cores/threads doesn't seem great for other use cases, but hey - 10% more performance in PS3 emulation).
[0] https://www.tomshardware.com/news/ps3-emulation-i9-12900k-vs...
Stuff like this is why I plan to never buy Intel ever again. I always disliked Intel's strategy for market segmentation by disabling specific instructions, I much prefer AMD's strategy of segmenting just by speed and number of cores. It is really annoying that with Intel you can't run the same program on the server and on your desktop, it makes development and testing a huge pain.
Nvidia do the same with disabling passthrough on consumer GPUs (gotta upsell Quatro). However I would say avoiding Intel completely is unnecessary. They're a good FOSS citizen when it boils to things like WLAN and GPU. Its just that with AMD, you can hardly go wrong on CPU and GPU side. Especially on Linux.
And they - rightly - get plenty of flak for it.