08-19-2017, 03:03 PM
08-19-2017, 03:21 PM
ok i found here https://hashcat.net/files_legacy/cudaHashcat-2.00.7z
md5 result cuda from v 2.0 Speed.GPU.#1.: 2900.4 MH/s
vs oc from v 3.6 Speed.Dev.#1.....: 1832.1 MH/s
cuda much faster and more efficiently. I do not understand why you refused new versions.
md5 result cuda from v 2.0 Speed.GPU.#1.: 2900.4 MH/s
vs oc from v 3.6 Speed.Dev.#1.....: 1832.1 MH/s
cuda much faster and more efficiently. I do not understand why you refused new versions.
08-21-2017, 04:42 PM
cuda is faster with old gpus
opencl is faster with new gpus
opencl is faster with new gpus
08-26-2017, 07:11 AM
I do not consider my card to be old. It is still quite modern and draws modern games
ps gtx 760
ps gtx 760
08-26-2017, 04:51 PM
It doesn't matter what you think, because this is not a subjective topic.
Your GPU is over 4 years old and runs on the Kepler architecture, which is bad with OpenCL.
To be exact, everything before Maxwell is bad for OpenCL and CUDA has an advantage. After Maxwell it's the opposite.
Your GPU is over 4 years old and runs on the Kepler architecture, which is bad with OpenCL.
To be exact, everything before Maxwell is bad for OpenCL and CUDA has an advantage. After Maxwell it's the opposite.
08-28-2017, 09:12 AM
I understand you, however how do you explain this? As far as I know opencl from nvidia was first broadcast in cuda and then into machine code. Thus logically it is impossible to achieve high performance. (opencl over cuda)
08-31-2017, 10:06 AM
That's not correct. Both languages compile into PTX first.