PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
11) Message boards : Aggie The Pew message board (Message 146906)
Posted 65 days ago by Profile j.sheridanProject donor
Before I throw this out to the masses, hopefully someone here can explain.

I have 2 2080s in 1 pc.

device 0 = 46 computeUnits @ 1830MHz
device 1 = 46 computeUnits @ 1710MHz

but what I cannot work out is why device 0 GFN run times are 10% slower than device 1.

Surely it should be the other way around?

I've double checked which is which by pausing my tasks and confirming the higher clocked card in afterburner is device 0.




Many possibilities...

1) Clocks are lower than stated due to heat
2) Device 0 slot is throttling PCIe speeds for some reason (e.g., sharing the lane with another device like a storage device)
3) device 0 is running the monitor(s)



The cards are water cooled and running at 45C so we can rule that one out :-)

In theory when you plug in a 2nd card it's meant to run at PCIe4: x16x4 - ie the main slot doesn't change speed. I've still got some old pending tasks and when device0 was there by itself it was returning WUs faster than device1 is now.

I've just changed the bios from "auto" to the "x8x4" option (strangely there is no x16x4 option) and it doesn't seem to have changed anything.

Just one more thing to hate about this crappy asrock board.



12) Message boards : Aggie The Pew message board (Message 146901)
Posted 65 days ago by Profile j.sheridanProject donor
Before I throw this out to the masses, hopefully someone here can explain.

I have 2 2080s in 1 pc.

device 0 = 46 computeUnits @ 1830MHz
device 1 = 46 computeUnits @ 1710MHz

but what I cannot work out is why device 0 GFN run times are 10% slower than device 1.

Surely it should be the other way around?

I've double checked which is which by pausing my tasks and confirming the higher clocked card in afterburner is device 0.

13) Message boards : Aggie The Pew message board (Message 146895)
Posted 65 days ago by Profile j.sheridanProject donor
GFN 18 runs in 450s on my stock 3080. .

Sweet !


That was testing earlier, now running GFN 20:

3080: 4000s
2080: 6500s

14) Message boards : Aggie The Pew message board (Message 146893)
Posted 65 days ago by Profile j.sheridanProject donor
How can it be ?
Home farm, in the UK BOINC Team, did 3 tasks almost immediately ?

I saw that... and I can't see his systems so no idea what he is running.

GFN 18 runs in 450s on my stock 3080.


As mentioned above, GFN20 is the better bet as they give 10% more credit, you just need to remember to shift down to 19 then 18 as time runs out.
15) Message boards : Number crunching : Threading (Message 146809)
Posted 68 days ago by Profile j.sheridanProject donor
Today I found my wingman using a 3950x who's running 16 SoB tasks concurrently.
This is the result
Around 110000s/task.

around 3x slower then the 35-40ks/task I get on my 3950x when running 2 at a time with 8 threads each.

3x slower, but he's running 8x more tasks. So, I'm not sure, but doesn't that mean his throuput is actually higher than yours? He does 16 tasks in 110000 s, or on average one task every 6875 s.

You only finish one task per 17000-20000 s. This might have made sense when being 1st meant something, but now it just means less chance of finding a prime.

That difference seems a bit extreme, so maybe I got the numbers you posted wrong?

Your maths is flawed.
He's doing 16 tasks in 1700000 seconds or 110000s/task.
I do 2 tasks in 70000 seconds or 35000s/task.

Prime95 will verify this difference if you run the benchmarks. 1 task will use the entire cache on a ryzen ccd so 16 tasks will be running mainly in ram.
16) Message boards : Number crunching : Threading (Message 146736)
Posted 69 days ago by Profile j.sheridanProject donor
That people don't bother with the message boards is exactly why I'm suggesting a notice & link on the task selection screen would be a good idea.
17) Message boards : Number crunching : Threading (Message 146734)
Posted 69 days ago by Profile j.sheridanProject donor
This was raised by Peter Hucker in the thread about splitting work units across GPUs. As that's now locked, here we are.



So am I running my 24 core work units slower than I should on my dual 12 core xeons? Would it even be clever enough to run two 12 core units on seperate CPUs?


yes, although even 12 core units may not be fastest, you need to test it.
I've seen people running tasks on a threadripper using 24 cores that end up slower than my 3700x running on 8.


Today I found my wingman using a 3950x who's running 16 SoB tasks concurrently.
This is the result
Around 110000s/task.

around 3x slower then the 35-40ks/task I get on my 3950x when running 2 at a time with 8 threads each.

I've PM'd the user to point this out, I'm hoping he understands the message.

There should really be a link in the options on these big tasks to an info page explaining how to check you're getting the best use out of your cpu.
18) Message boards : Number crunching : Surpassing the progress. (Message 146652)
Posted 72 days ago by Profile j.sheridanProject donor
Back when it was all fields around these parts it used to take days to run a single SoB, now I can run 4/day.
19) Message boards : Number crunching : LLR2 installed on all big LLR projects (Message 146527)
Posted 75 days ago by Profile j.sheridanProject donor
New version 9.01 is live on GCW. GCW should be faster now.

fantastic. looks to be around 10% faster.
20) Message boards : Number crunching : Can I run 1 WU across multiple GPUs? (Message 146300)
Posted 79 days ago by Profile j.sheridanProject donor

So am I running my 24 core work units slower than I should on my dual 12 core xeons? Would it even be clever enough to run two 12 core units on seperate CPUs?

yes, although even 12 core units may not be fastest, you need to test it.
I've seen people running tasks on a threadripper using 24 cores that end up slower than my 3700x running on 8.


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2021 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.02, 0.02, 0.00
Generated 24 Feb 2021 | 23:10:14 UTC