Well THAT's annoying.
Well THAT's annoying.
That is why i quit buying Asus after my Xeon v4 board. They just....i dunno. But i had enough.
I may have to put them on my blacklist as well. I'm already done with MSI ... what's up with these manufactures lies and dirty tricks sheit? I feel like they intentionally crippled the earlier boards so we'll have to go out and buy the new Zenith .. f*ck'em.
Is that a different board you used for your 3970X build? If so, why did you change?Originally Posted by VietOZ
my 3970X build used Asus Prime ... I got this board about the same time. My initial thoughts was to build 2x 3970X, but then AMD announced the 3990X so I waited. I bought all the gears(except cpu) before Christmas so whatever was available at good price then I grabbed it. I still haven't update bios on the Prime yet, it still runs agesa 1.0.0.2. Now i'm actually scare to update it. The Strix was fine with the same bios as the Prime, memory was running at 3600mhz. But it couldn't boot into either windows or linux (with mce=off). So I decided to update the bios and the rest is history.
1) Regarding windows and 128 threads:
https://www.anandtech.com/show/15483...3990x-review/3
Looks like SMT/HT + 128 threads + AVX2 + windows is a problem. See the NAMD graph and comments. And the the summary:
Does Linux have this problem?What’s The Verdict?
From our multithreaded test data, there can only be two conclusions. One is to disable SMT, as it seems to get performance uplifts in most benchmarks, given that most benchmarks don’t understand what processor groups are. However, if you absolutely have to have SMT enabled, then don’t use normal Windows 10 Pro: use Pro for Workstations (or Enterprise) instead. At the end of the day, this is the catch in using hardware that's skirting the line of being enterprise-grade: it also skirts the line with triggering enterprise software licensing. Thankfully, workstation software that is outright licensed per core is still almost non-existent, unlike the server realm.
Ultimately this puts us in a bit of a quandary for our CPU-to-CPU comparisons on the following pages. Normally we run our CPUs on W10 Pro with SMT enabled, but it’s clear from these benchmarks that in every multithreaded scenario, we won’t get the best result. We may have to look at how we test processors >16 cores in the future, and run them on Windows 10 Enterprise. Over the following pages, we’ll include W10 Pro and W10 Enterprise data for completeness.
2) I beginning to think that Windows is too dumb to understand the chiplet architecture used in Zen2. For example, there are 4 chiplets of 8 cores each on a 3970X. Each chiplet contains its own cache. Scenario: (SMT/HT off) running 4 MT tasks, each task using 8 cores, it would be best if each task was contained in a single chiplet. That way cache calls would be fastest. Otherwise the task has to go out to the bus to retrieve the data elsewhere. But the poor task times I am seeing are making me think windows is not doing the smart thing.
Is there a way to make windows keep MT tasks tied to a single chiplet?
Is linux any better at this?
3) https://www.reddit.com/r/linuxmaster..._scheduler_vs/
I am going to have to try linux to see the difference. No VM this time, a separate boot device.
Last edited by zombie67; 02-10-20 at 02:28 AM. Reason: more info found
Seems if you limit MT tasks to 16 threads (8 with SMT off) it should keep them on each of the CCXs. Another possibility of poor times is using too many threads at once. Some threads finish first and sit idle waiting on other threads.
@Z: Supposedly Windows 10 Pro fixed the issue. Make sure you get build 18362.592 or later and it *should* be fine. Pro for Workstations/Enterprise isn't required.
Just so I am clear, are you talking about the issue that anandtech talked about, or the problem that I think is happening with keeping MT tasks with 16 threads (8 with SMT off) or less tied to a CCX?
FWIW, I am running the latest general release of win10 pro (not any of the fast(er) rings). In this case, it is 18363.
Last edited by zombie67; 02-18-20 at 10:09 AM.
It's supposed to fix the issue Anandtech was talking about
Posting Permissions
Reply With Quote
