mfaktc: a CUDA program for Mersenne prefactoring

[BigBrother]

I can saturate (100% GPU) my GTX680 when running two instances of mfaktc.

http://gpuz.techpowerup.com/12/03/27/a9a.png

Note that the GPU core clock is constantly boosted to +-1100MHz and the power consumption hovers around 72% TDP, which could mean that the performance/Watt for this chip is higher than in James' calculations. This power consumption sensor seems to be a new feature on this chip, I've never seen it displayed in GPU-Z before on any other card.

[BigBrother]

It turns out that I plugged my brand new shiny bling-bling GTX680 into a PCI-E 2.0 x8 slot instead of a PCI-E 2.0 x16 slot... I'll change it tonight, and also try to fix a crazy problem that causes my motherboard to refuse more than one memory module, forcing it to use single channel DDR3. I don't expect radically improved CUDA performance, but we'll see.

[msft]

Originally Posted by Prime95

This is somewhat surprising to me. However, I thought mfaktc would use the more numerous CUDA cores to do the 32-bit muls and adds that predominate in TF. Where did I go wrong?

http://forums.nvidia.com/index.php?s...&#entry1387312

* Relative to the throughput of single precision multiply-add, the throughput of integer shifts, integer comparison, and integer multiplication is lower than before.

It is answer?

[nucleon]

Originally Posted by TheJudger

I guess I need to buy a GTX 6[78]0... ;)

I'd be curious if you can weave some more TheJudger magic to get more out of the GTX680. :)

Now with some performance figures out, I'm pretty disappointed. I was hoping to buy some GTX680s to replace some hardware here to reduce my power bill.

It doesn't even surpass what I have on performance per watt metrics.

-- Craig

[James Heinrich]

Originally Posted by axn

You're nearly there. Rather than using the cum.prob., just use the probability for the given bit depth. You should see a rough doubling of the % with every bit.

Thanks. I didn't have my brain screwed on quite straight yesterday, but I think I've fixed it so it makes sense now.
http://mersenne-aries.sili.net/cudalucas.php?model=13

[BigBrother]

Originally Posted by msft

http://forums.nvidia.com/index.php?s...&#entry1387312

It is answer?

Some exact numbers: (Operations per Clock Cycle per Multiprocessor)

                  CC 1.x           CC 2.0         CC 2.1         CC 3.0
                  
32-bit floating
point add,          8                32             48             192
multiply,
multiply-add

64-bit floating
point add,          1                16              4              8
multiply,
multiply-add

32-bit
integer add         10               32             48             168

32-bit integer
multiply,
multiply-add,      Multiple          16             16              32
sum of absolute  instructions
difference

From table 5-1 in the CUDA C Programming Guide Version 4.2

Not much love for 32-bit integer multiply & multiply-add, compared to 32-bit floating point operations.

[axn]

Originally Posted by James Heinrich

Thanks. I didn't have my brain screwed on quite straight yesterday, but I think I've fixed it so it makes sense now.
http://mersenne-aries.sili.net/cudalucas.php?model=13

Much better. Now, if we could just drill down an individual row to 1M granularity...

[kladner]

Originally Posted by James Heinrich

I didn't have my brain screwed on quite straight yesterday

That page has really come a long way in a short time. Another great tool!
Thanks for doing it.

BTW: I wasn't thinking too well, either, when I ran the CuLu benchmarks. Sorry for the incomplete data, James.

[James Heinrich]

Originally Posted by axn

Much better. Now, if we could just drill down an individual row to 1M granularity...

You can if you click the zoom in/out links I just added.

[msft]

Originally Posted by BigBrother

Some exact numbers: (Operations per Clock Cycle per Multiprocessor)

                  CC 1.x           CC 2.0         CC 2.1         CC 3.0
32-bit integer
multiply,
multiply-add,      Multiple          16             16              32
sum of absolute  instructions
difference

GTX-580 have 16 Multiprocessors,GTX-680 have 8.
GTX-680 each Multiprocessor have 192 core,But only 32 32-bit integer multiply exec.
Lots of thread wait exec.

                  CC 1.x           CC 2.0         CC 2.1         CC 3.0
32-bit integer
shift compare     8               16                16                8

[BigBrother]

Originally Posted by BigBrother

It turns out that I plugged my brand new shiny bling-bling GTX680 into a PCI-E 2.0 x8 slot instead of a PCI-E 2.0 x16 slot... I'll change it tonight, and also try to fix a crazy problem that causes my motherboard to refuse more than one memory module, forcing it to use single channel DDR3. I don't expect radically improved CUDA performance, but we'll see.

Well, The Card is now inserted into a PCI-E 2.0 x16 slot, and my brain surgery skills allowed me to fix a bent pin on the CPU socket so my memory is back at dual channel again.

One instance of mfaktc is now taking +-70% GPU instead of the 74% I reported yesterday, and nVidia's Visual Profiler shows transfer rates of 6Gb/s instead of 3 Gb/s, but since the amount of data to transfer is relatively small, there's no earth-shattering improvement. I could run the same benchmark I did yesterday again if James would like me to do that.