CudaLucas Residual

evoflash · 2017-06-07, 21:07

Hello everyone, I'm starting out with CudaLucas, and ran a trial factor test to see if was working ok. So far so good. However when I ran a LL-double check I had residual of 0x0000000000000002 and received an error message at 100% which was confounding.

I put this down to high overclocking and restarted another check with a milder overclock that I use for stable gaming. I'm now getting patches of similar residue (see attachment), is this normal or am I wasting time/gpu watts?

Thanks for looking.

GP2 · 2017-06-07, 21:23

Quote:

Originally Posted by evoflash

Hello everyone, I'm starting out with CudaLucas, and ran a trial factor test to see if was working ok. So far so good. However when I ran a LL-double check I had residual of 0x0000000000000002 and received an error message at 100% which was confounding.

I put this down to high overclocking and restarted another check with a milder overclock that I use for stable gaming. I'm now getting patches of similar residue (see attachment), is this normal or am I wasting time/gpu watts?

Thanks for looking.

A residue of 0x0000000000000002 is going to remain at that value until the end of the test, so there's no point in continuing. You should abort the test.

Except yours didn't, it actually switched to something else before reverting back to 0x0000000000000002 again. (!)

There is a whole lot of wrong happening here. You should get your computer checked out.

kladner · 2017-06-07, 23:39

What hardware are you running, what frequencies for GPU and VRAM?

evoflash · 2017-06-08, 07:00

Yeah, I suspected it wasn't looking good thanks for advice.

I'm running a GTX 970 at ~1490Mhz core clock and 3900Mhz VRAM. I'll kill the test, drop the overclock to base clocks and try again.

VBCurtis · 2017-06-08, 07:26

typical experience around here is the memory clock is more correlated with errors; you may even have to underclock the mem frequency a bit to get GPGPU stability.

Don't be surprised if you find a stable setting that OC's the processor a bit but runs memory at stock or a bit under.

evoflash · 2017-06-08, 08:24

Great yes I'll do that. Thanks for tips. I will set it running tonight with underclock memory and see where it gets to.

kladner · 2017-06-08, 15:27

Quote:

Originally Posted by VBCurtis

typical experience around here is the memory clock is more correlated with errors; you may even have to underclock the mem frequency a bit to get GPGPU stability.

Don't be surprised if you find a stable setting that OC's the processor a bit but runs memory at stock or a bit under.

+1

evoflash · 2017-06-10, 11:53

Quick update to say all looks good. Dropped GPU clock to base setting and bottomed out the memory clock, no errors so far. Thanks for advice.

Mark Rose · 2017-06-10, 13:05

Quote:

Originally Posted by evoflash

Quick update to say all looks good. Dropped GPU clock to base setting and bottomed out the memory clock, no errors so far. Thanks for advice.

You'll want to run your memory clock faster than that, but it's a matter of finding out how high you can go before you get failures.

Dubslow · 2017-06-10, 17:13

Just to emphasize the point, many consumer-grade graphics cards are shipped with defective memory, i.e. memory that will cause CUDALucas errors at stock speed. This is because for *graphics* it doesn't matter if a bit or ten is wrong, it won't noticeably effect the on-screen rendered visuals that these cards are *assumed* to be used for.

CUDALucas, of course, like any LL testing software, is quite the opposite of error tolerant, and so any single bit failure *will* render the test completely useless.

Hence the need to sometimes underclock consumer graphics card memory. If you buy one of the professional cards, which are a lot more expensive, you can and should expect utterly flawless memory since they are designed for computation, unlike consumer graphics cards.

LaurV · 2017-06-11, 04:42

Quote:

Originally Posted by Dubslow

Just to emphasize the point, many consumer-grade graphics cards are shipped with defective memory, i.e. memory that will cause CUDALucas errors at stock speed. This is because for *graphics* it doesn't matter if a bit or ten is wrong, it won't noticeably effect the on-screen rendered visuals that these cards are *assumed* to be used for.

CUDALucas, of course, like any LL testing software, is quite the opposite of error tolerant, and so any single bit failure *will* render the test completely useless.

Hence the need to sometimes underclock consumer graphics card memory. If you buy one of the professional cards, which are a lot more expensive, you can and should expect utterly flawless memory since they are designed for computation, unlike consumer graphics cards.

+1. You just explained very well and in fewer words what I am trying to say for a long time.

2017-06-07, 21:07	#1
evoflash Dec 2012 2×13 Posts	CudaLucas Residual Hello everyone, I'm starting out with CudaLucas, and ran a trial factor test to see if was working ok. So far so good. However when I ran a LL-double check I had residual of 0x0000000000000002 and received an error message at 100% which was confounding. I put this down to high overclocking and restarted another check with a milder overclock that I use for stable gaming. I'm now getting patches of similar residue (see attachment), is this normal or am I wasting time/gpu watts? Thanks for looking. Attached Thumbnails

2017-06-10, 11:53	#8
evoflash Dec 2012 1A₁₆ Posts	Quick update to say all looks good. Dropped GPU clock to base setting and bottomed out the memory clock, no errors so far. Thanks for advice. Attached Thumbnails

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Don't DC/LL them with CudaLucas	LaurV	Data	131	2017-05-02 18:41
Prime95 vs. CUDALucas	storm5510	Software	6	2016-11-28 17:36
CUDALucas gives all-zero residues	fivemack	GPU Computing	4	2016-07-21 15:49
CUDALucas: which binary to use?	Karl M Johnson	GPU Computing	15	2015-10-13 04:44
Primes in residual classes	Unregistered	Information & Answers	6	2008-09-11 12:57

2017-06-07, 23:39	#3
kladner "Kieren" Jul 2011 In My Own Galaxy! 10158₁₀ Posts	What hardware are you running, what frequencies for GPU and VRAM?

2017-06-08, 07:00	#4
evoflash Dec 2012 32₈ Posts	Yeah, I suspected it wasn't looking good thanks for advice. I'm running a GTX 970 at ~1490Mhz core clock and 3900Mhz VRAM. I'll kill the test, drop the overclock to base clocks and try again.

2017-06-08, 07:26	#5
VBCurtis "Curtis" Feb 2005 Riverside, CA 12FD₁₆ Posts	typical experience around here is the memory clock is more correlated with errors; you may even have to underclock the mem frequency a bit to get GPGPU stability. Don't be surprised if you find a stable setting that OC's the processor a bit but runs memory at stock or a bit under.

2017-06-08, 08:24	#6
evoflash Dec 2012 32₈ Posts	Great yes I'll do that. Thanks for tips. I will set it running tonight with underclock memory and see where it gets to.

2017-06-10, 17:13	#10
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 3·29·83 Posts	Just to emphasize the point, many consumer-grade graphics cards are shipped with defective memory, i.e. memory that will cause CUDALucas errors at stock speed. This is because for graphics it doesn't matter if a bit or ten is wrong, it won't noticeably effect the on-screen rendered visuals that these cards are assumed to be used for. CUDALucas, of course, like any LL testing software, is quite the opposite of error tolerant, and so any single bit failure will render the test completely useless. Hence the need to sometimes underclock consumer graphics card memory. If you buy one of the professional cards, which are a lot more expensive, you can and should expect utterly flawless memory since they are designed for computation, unlike consumer graphics cards.