mfakto: an OpenCL program for Mersenne prefactoring

[rebirther]

Originally Posted by rebirther

This was a card from a user, Iam trying to find some results but no one has attached an Intel ARC host yet, still need some tweaking with the server setup.

no factor for M131174377 from 2^74 to 2^75 [mfakto 0.15pre7-MGW cl_barrett32_76_gs_2]
tf(): total time spent: 24m 10.213s (1737.73 GHz-days / day)

08:56:56 (11896): mfakto.exe exited; CPU time 1443.187500

Result from an Intel ARC 770 on linux on SRBase.

[28add11]

I haven't checked the forum in a few days and it looks like you guys are really getting into the swing of things with ARC support. Since I have an A750 card if anyone would like me to run tests on it or help out I would be more than happy to help.
A small note: I downloaded mfakto from here but by the looks of things there's a more recently updated repo, so if anyone could provide instructions on how to build that it would be much appreciated, Thanks!

[kriesel]

Does there exist an mfaktc build that will work on Google Colab in either its Ubuntu 18.04 or 20.04 VM incarnations currently, which appear unpredictably, usually 18.04? I'm getting Cudart version discrepancies with 10.0, or glibc / libstdc issues with the cuda 12.0 for linux mmfsktc build, and updating 18.04 does not resolve that issue for mmff so likely won't for mfaktc.

[Mark Rose]

Originally Posted by kriesel

Does there exist an mfaktc build that will work on Google Colab in either its Ubuntu 18.04 or 20.04 VM incarnations currently, which appear unpredictably, usually 18.04? I'm getting Cudart version discrepancies with 10.0, or glibc / libstdc issues with the cuda 12.0 for linux mmfsktc build, and updating 18.04 does not resolve that issue for mmff so likely won't for mfaktc.

Wrong thread, but I've always had to recompile mfaktc for different versions of CUDA.

That being said, it's very quick to compile if nvidia-cuda-dev is installed: even on an old two core/two thread machine it takes less than 15 seconds.

[Prescott]

Hello,

The data file said to "Please post the device name "gfx1032 (Advanced Micro Devices, Inc.)"" to this forum. Device name is AMD Radeon RX 6600

Thank you!

[travisjank]

WARNING: Unknown GPU name, assuming GCN. Please post the device name "gfx1011:xnack- (Advanced Micro Devices, Inc.)" to http://www.mersenneforum.org/showthread.php?t=15646 to have it added to mfakto. Set GPUType in mfakto.ini to select a GPU type yourself to avoid this warning.

OpenCL device info
name gfx1011:xnack- (Advanced Micro Devices, Inc.)
device (driver) version OpenCL 2.0 AMD-APP (3417.0) (3417.0 (PAL,LC))
maximum threads per block 1024
maximum threads per grid 1073741824
number of multiprocessors 18 (1152 compute elements)
clock rate 555 MHz

Automatic parameters
threads per grid 0
optimizing kernels for GCN

Compiling kernels.
GPUSievePrimes (adjusted) 81206
GPUsieve minimum exponent 1037054
Started a simple selftest ...
Selftest statistics
number of tests 30
successful tests 30


AMD Radeon Pro V520 Graphics Card

GPU Architecture: RDNA
Compute Units: 36
Peak INT4 Performance: 58.98 TOPs
Peak Half Precision (FP16) Performance: 14.75 TFLOPs
Peak Single Precision (FP32) Performance: 7.4 TFLOPs
Peak INT8 Performance: 24.94 TOPs
Lithography: TSMC 7nm FinFET
Stream Processors: 2304
Peak Engine Clock: 1600 MHz
Peak Double Precision (FP64) Performance: 461 GFLOPs
Peak Single Precision Matrix (FP32) Performance: 7.4 TFLOPs
Dedicated Memory Size: 8 GB
Dedicated Memory Type: HBM2
Memory Interface: 2048-bit
Memory Clock: 1000 GHz
Peak Memory Bandwidth: Up to 512 GB/s
Memory ECC Support: Yes (Full-Chip)
GHz-d/day: 640


GPUZ Validation:
https://www.techpowerup.com/gpuz/details/4393n

CPUZ Validation:
https://valid.x86.fr/mj07te

AMD Product Page:
https://www.amd.com/en/products/serv...adeon-pro-v520

I also added it to the HWBOT.org database:
https://hwbot.org/hardware/videocard/radeon_pro_v520/

[OEIS11221]

The program asked me to post the device name 'gfx90c' to the forum.
Thank you!

[Magellan3s]

Originally Posted by rebirther

Result from an Intel ARC 770 on linux on SRBase.

no factor for M131174377 from 2^74 to 2^75 [mfakto 0.15pre7-MGW cl_barrett32_76_gs_2]
tf(): total time spent: 24m 10.213s (1737.73 GHz-days / day)

08:56:56 (11896): mfakto.exe exited; CPU time 1443.187500

For anyone curious, the RTX 4090 gets 13,800 GHz-days / day on that particular exponent.

Starting trial factoring M131174377 from 2^74 to 2^75 (29.17 GHz-days)
k_min = 72001355612340
k_max = 144002711226740
Using GPU kernel "barrett76_mul32_gs"
Date Time | class Pct | time ETA | GHz-d/day Sieve Wait
May 28 19:15 | 4619 100.0% | 0.190 n.a. | 13816.24 82485 n.a.%
no factor for M131174377 from 2^74 to 2^75 [mfaktc 0.21 barrett76_mul32_gs]
tf(): total time spent: 3m 6.573s

[moebius]

Originally Posted by travisjank

AMD Radeon Pro V520 Graphics Card

I have a request for you. I need a prp benchmark with gpuowl ver. 6 to 7 for this GPU for the Mersenne exponents 77936867 and 116085643. I don't have any concrete values for the Navi 12 architecture in the List yet! You can post the result in the following thread.

https://mersenneforum.org/showpost.p...postcount=2931