+- hashcat Forum (https://hashcat.net/forum)
+-- Forum: Misc (https://hashcat.net/forum/forum-15.html)
+--- Forum: Hardware (https://hashcat.net/forum/forum-13.html)
+--- Thread: GTX 970 Benchmark (/thread-3982.html)
Pages:
1
2
RE: GTX 970 Benchmark - TheJoSko - 01-16-2015
Upgrading to 346.22 and 1.32 seems to have resolved the cuStreamSynchronize errors I was seeing in 1.31 / 343.22 when overclocking.
Tuned workload profile and the following overclock:
Code:
nvidia-settings -a [gpu:0]/GPUGraphicsClockOffset[3]=250
nvidia-settings -a [gpu:1]/GPUGraphicsClockOffset[3]=250
nvidia-settings -a [gpu:0]/GPUMemoryTransferRateOffset[3]=400
nvidia-settings -a [gpu:1]/GPUMemoryTransferRateOffset[3]=400/cudaHashcat-1.32/cudaHashcat64.bin -w 3 --remove -o /Hashes.cracked -m 5600 -r /cudaHashcat-1.32/rules/best64.rule /Hashes.txt /mega2.txt
Hash.Type......: NetNTLMv2
Speed.GPU.#1...: 257.9 MH/s
Speed.GPU.#2...: 258.2 MH/s
Speed.GPU.#*...: 516.1 MH/s
RE: GTX 970 Benchmark - epixoip - 01-16-2015
best64.rule is not enough rules to gain full acceleration. try using d3ad0ne.rule or rockyou-30000.rule
RE: GTX 970 Benchmark - TheJoSko - 01-17-2015
(01-16-2015, 08:30 PM)epixoip Wrote: best64.rule is not enough rules to gain full acceleration. try using d3ad0ne.rule or rockyou-30000.rule
Thanks epixoip.
With 150 clock offset and d3ad0ne.rule I am seeing:
Hash.Type......: NetNTLMv2
Speed.GPU.#1...: 294.4 MH/s
Speed.GPU.#2...: 291.9 MH/s
Speed.GPU.#*...: 586.3 MH/s
Bump it up to 200 offset and I get:
Hash.Type......: NetNTLMv2
Speed.GPU.#1...: 302.4 MH/s
Speed.GPU.#2...: 302.4 MH/s
Speed.GPU.#*...: 604.8 MH/s
Bump it up to 250 offset and I get:
Hash.Type......: NetNTLMv2
Speed.GPU.#1...: 309.8 MH/s
Speed.GPU.#2...: 309.6 MH/s
Speed.GPU.#*...: 619.4 MH/s
...
ERROR: cuStreamSynchronize() 714
I guess 200 is the sweet spot for these cards.