They already started the random freeze/pause bug in the scheduler on windows 10 to push people off of it.
Amazing how that enters every EOL software from big companies.
Amazing how that enters every EOL software from big companies.
And the reason why I'm trying to learn Linux in a VM atm...They already started the random freeze/pause bug in the scheduler on windows 10 to push people off of it.
Amazing how that enters every EOL software from big companies.
IDK, did desktop high maximum also.Maybe the 720p setting doesn't create enough workload for the GPU drivers to max out all cores.
10508 at 1440p with 4060:Now I would.like to see some 2700k results personally.
Mine was 3440x1440p10508 at 1440p with 4060:
Unless you game a lot, get the 9950xI'm planning a zen 5 build in the coming weeks:
-9950x or 9800x3D
-5090
My main use cases are:
-running deep learning training experiments
-general application performance (e.g. google earth web, had a 13900k that used to drop frames)
I often use my PC while training too.
I'm leaning on the 9950x but I have a few concerns.
-What's the general application performance penalty of NOT using core parking drivers? I really do not want to micro-manage cpu behavior or rely on unnecessary software...
As for the 9800x3D, the major pros are that it's a single CCD, and that I believe that it will have better general application performance, such as more stable frame rates in google earth.
However, I'm concerned that 8 cores aren't enough for feeding a GPU during training. I know it's incredibly vague, but I'd appreciate some insight if anyone has to share?
I used to have a 13900k+4090 where I found some performance improvement when using 8p+8e over 8p+0e for the Dataloader when training smaller models. Hence I'm worried that a 9800x3D would leave performance on the table, especially when using a 5090.
Put my cold weather overclock on, just ran another one, here is the text file.How did you get the score to improve so significantly?
Just remembered I can remove the boost from the over clock. Will try in a bitHow did you get the score to improve so significantly?
You could hold out for the 9950X3D.I'm planning a zen 5 build in the coming weeks:
-9950x or 9800x3D
-5090
got it.How did you get the score to improve so significantly?
Impressive!got it.
Unfortunately I won’t have the funds for it until next year as I haven’t worked in quite a while. I should have income again next year at some point hopefully around March or April.The X670E version is $1200. I think the more you wait, the more the chance that it will go out of stock or there will be a price hike.
I don't think it will have dual V-cache, so reliance on core parking/drivers will be even worse unfortunately. Would be ideal otherwise.You could hold out for the 9950X3D.
The issue with core parking/drivers. Where is this a problem ? I have no issues at all. Including the 8 7950x's I have.I don't think it will have dual V-cache, so reliance on core parking/drivers will be even worse unfortunately. Would be ideal otherwise.
The 7950X doesn't use core parking. Only the 7950X3D, 9900X, and 9950X.The issue with core parking/drivers. Where is this a problem ? I have no issues at all. Including the 8 7950x's I have.
The 9950X does rely on core parking and thread priority drivers as opposed to the 7950X, even though both have the same cores on both CCD's (though one is usually binned lower). So whether the 9950X3D will have identical cores on both CCD's or not, it will almost surely rely on the same core parking and thread priority that AMD actively opted for on the 9950X.I don't think it will have dual V-cache, so reliance on core parking/drivers will be even worse unfortunately. Would be ideal otherwise.
On dual-CCD Ryzens, as well as on Threadrippers and EPYCs, we who are active in citizen science/ scientific computing occasionally use something more targeted than "core parking": When we have multithreaded applications which heavily share data between threads, we assign CPU affinity to thread groups (usually to processes a.k.a. tasks) such that the heavy sharing is confined to a CCX ( = to a last level cache domain) which is common to this thread group. On EPYCs or/and on old Zen 2 CPUs, we also sometimes arrange for such sharing across two CCXs if we want more cores per thread group. Maybe that's what you would call micro-management, but in our case, this is set-and-forget. IOW it's trivial to set up and operate once the performance characteristics of the application are known.My main use cases are:
-running deep learning training experiments
-general application performance (e.g. google earth web, had a 13900k that used to drop frames)
I often use my PC while training too.
I'm leaning on the 9950x but I have a few concerns.
-What's the general application performance penalty of NOT using core parking drivers? I really do not want to micro-manage cpu behavior or rely on unnecessary software...
I'm planning a zen 5 build in the coming weeks:
-9950x or 9800x3D
Are you sure? The Prime series is kinda their boring line of mobos. I prefer their CSM, TUF and Creator series.Think i'm going to pickup the ASUS Prime X670E-PRO WIFI
The prime and tuff are pretty much the same board.Are you sure? The Prime series is kinda their boring line of mobos. I prefer their CSM, TUF and Creator series.