1)
Message boards :
Aggie The Pew message board
(Message 147223)
Posted 886 days ago by  j.sheridan
I slightly miscalculated my last gfn20 task so had to fling the windows open. Just scraped in with the extra boost the gpu got from running at 30C instead of 50C.
|
2)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
WW project is live
(Message 147222)
Posted 886 days ago by j.sheridan
now slow down and give the rest of us a chance :-)
ps great job!
|
3)
Message boards :
Number crunching :
Is turning on hyper threading better if you still assign number of tasks to only pysical cores?
(Message 147221)
Posted 886 days ago by j.sheridan
(From memory, it's been a while since I checked) windows does something weird like using 7/8 "real" cores when running 8 single threaded tasks on a 16 virtual core system.
|
4)
Message boards :
Number crunching :
Is turning on hyper threading better if you still assign number of tasks to only pysical cores?
(Message 147209)
Posted 887 days ago by j.sheridan
My observations on Windows:
If you run single thread LLR tasks, one per core, HT on is ~10% worse than HT off. Windows scheduler is doing something bad here, which I nicknames "core smear" as it moves the threads around and I think it overlaps them at times, hence the performance loss. Can be worked around by manually setting affinity to the boinc process, which gets inherited by tasks it runs subsequently.
Does the program Project Lasso assist in keeping them on the real cpu cores and away from the HT ones?
As I understand it they're the same - with HT on each physical core has 2 virtual cores.
|
5)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
New Version Testing
(Message 147149)
Posted 888 days ago by j.sheridan
Nvidia 3080 on windows, results as above. Some timings:
Workunit complete. Average speed: 225.42 Million primes per second.
Current:
- n/a
Test 2:
New:
- Workunit complete. Average speed: 420.67 Million primes per second.
Current:
- Checkpoint time: 13 seconds. 2119660489 primes tested (163.05 Million primes per second).
- Checkpoint time: 2 seconds. 404384060 primes tested (202.19 Million primes per second).
Test 3:
New:
- Workunit complete. Average speed: 339.36 Million primes per second.
Current:
- Checkpoint time: 13 seconds. 1413982067 primes tested (108.77 Million primes per second).
- Checkpoint time: 3 seconds. 282799572 primes tested (94.27 Million primes per second).
Test 4:
New:
- Workunit complete. Average speed: 247.75 Million primes per second.
Current:
- Checkpoint time: 14 seconds. 1238747320 primes tested (88.48 Million primes per second).
- Checkpoint time: 3 seconds. 247763709 primes tested (82.59 Million primes per second).
|
6)
Message boards :
Aggie The Pew message board
(Message 147092)
Posted 891 days ago by j.sheridan
I always wonder seeing setups like this if they've mentioned to management just how much of the company infrastructure is being used for non-business purposes.
|
7)
Message boards :
Aggie The Pew message board
(Message 146995)
Posted 894 days ago by j.sheridan
Time to get a decent board - I knew should have sent this one back the second I saw they'd made it 1cm shorter than standard atx. Anyone want to buy a crappy asrock motherboard?
They don't call is ass rock for nothing.
Priceless. Gotta remember this one!
I can't believe we are not getting dinged for the message so I will post it again !! That was funny ! Never heard of that before but of course I don't delve into these that much.
Looks like the 16/4 pci config is actually standard on a lot of x570 boards which is worrying for something that's supposed to be the "performance" option.
I've ordered an MSI that shifts to 8/8 as one would expect to replace it and the asrock can go onto ebay for some other sucker lucky person to deal with.
msi installed and the fast card is now 10% faster than before.
A good reminder to always check the specs closely.
If your motherboard says it does pcie4 4x you might think that's the same bandwidth as pcie3 8x BUT if only 4x of your 16x connector is wired and your card is pcie3 then all you're going to get is pcie3 4x.
|
8)
Message boards :
Aggie The Pew message board
(Message 146950)
Posted 895 days ago by j.sheridan
Time to get a decent board - I knew should have sent this one back the second I saw they'd made it 1cm shorter than standard atx. Anyone want to buy a crappy asrock motherboard?
They don't call is ass rock for nothing.
Priceless. Gotta remember this one!
I can't believe we are not getting dinged for the message so I will post it again !! That was funny ! Never heard of that before but of course I don't delve into these that much.
Looks like the 16/4 pci config is actually standard on a lot of x570 boards which is worrying for something that's supposed to be the "performance" option.
I've ordered an MSI that shifts to 8/8 as one would expect to replace it and the asrock can go onto ebay for some other sucker lucky person to deal with.
|
9)
Message boards :
Aggie The Pew message board
(Message 146948)
Posted 895 days ago by j.sheridan
How does having HT available affect things if you're also running GPU tasks and/or using the pc for other purposes?
|
10)
Message boards :
Aggie The Pew message board
(Message 146910)
Posted 896 days ago by j.sheridan
I'd suggest using GPU-Z to look at the stats on both in real time at the same time, and see what differences there are. Should be easy to see the currently running clock, and maybe also have a look at bus loading and other stuff too.
Thanks, just did that and it reports it's running pci3 x4.
Time to get a decent board - I knew should have sent this one back the second I saw they'd made it 1cm shorter than standard atx. Anyone want to buy a crappy asrock motherboard?
|
