Optimal Driver for Enterprise (ODE) Most users select this choice for optimal stability and performance. It also offers ISV certification and long lifecycle support Quadro New Feature Driver (QNF...
-Aleks-: Are there any limitations for calculations on CPU and GPU?
As a rule, the limitation is memory, both for GPU and CPU.
But one thing is clear - the code for the CPU is less demanding to optimize than for the GPU, it often happens that on the CPU smoothly works that on the GPU can not even run or quickly terminates with an error or crash of the video driver.
As a rule, the limitation is memory, both for GPU and CPU.
But one thing is clear - the code for the CPU is less demanding to optimize than for the GPU, it often happens that on the CPU smoothly works that on the GPU can not even run or quickly ends with an error or crash of the video-driver.
If this is the case, is it possible to use the GPU to calculate indicators and run the Expert Advisor on the CPU?
OpenCL is very good, you can train neural networks very quickly with this technology, hundreds (thousands) of times faster than on an ordinary processor.
This can be corrected as follows: find these lines in the script -
It was clear to me, as a non-programmer, from the comments that this was not for the masses. But it was still interesting to read. Tell me, should I expect a wave of new ideas in the form of indicators and experts? Or is it just a toy for the elite?
When trying to run the script, at first it went like this:
MG 0 21:04:59.052 Terminal OpenCL not found, install the latest OpenCL driver for better performance
PE 0 21:04:59.052 Scripts script Seascape (EURUSD,H1) removed
and then from http://www.nvidia.ru/Download/index.aspx?lang=ru
the latest card drivers were downloaded.
and the log messages are the same as above, when trying to run the script.
then Intel® SDK for OpenCL™ Applications for Windows* 2016 was downloaded from the website
http://registrationcenter-download.intel.com/akdlm/irc_nas/vcp/11058/intel_sdk_for_opencl_setup_6.3.0.1904.exe
at next new attempt to run the script, the following message will appear in the log:
CQ 0 17:14:22.991 OpenCL Device #0: CPU Intel(R) Corporation Intel(R) Core(TM)2 Duo CPU E6550 @ 2.33 GHz with OpenCL 2.1 (2 units, 2330 MHz, 4094 Mb, version 6.3.0.1904)
EP 0 17:14:23.661 Scripts script Seascape (EURUSD,H1) removed
well, who knows what's wrong?
it can't get this openCL* to work
________________________
openchlor* - OpenCL
Change the script code, instead of
try the following line
You have, in the list of OpenCL devices, no video adapters!
Change the script code, instead of
try using the following line
didn't work (
LD 0 18:10:50.116 OpenCL Device #0: CPU Intel(R) Corporation Intel(R) Core(TM)2 Duo CPU E6550 @ 2.33 GHz with OpenCL 2.1 (2 units, 2330 MHz, 4094 Mb, version 6.3.0.1904)
RM 0 18:10:50.223 Scripts script Seascape (EURUSD,H1) removed
didn't help (
Are there any limitations for calculations on CPU and GPU?
As a rule, the limitation is memory, both for GPU and CPU.
But one thing is clear - the code for the CPU is less demanding to optimize than for the GPU, it often happens that on the CPU smoothly works that on the GPU can not even run or quickly terminates with an error or crash of the video driver.
Compile the script
As a rule, the limitation is memory, both for GPU and CPU.
But one thing is clear - the code for the CPU is less demanding to optimize than for the GPU, it often happens that on the CPU smoothly works that on the GPU can not even run or quickly ends with an error or crash of the video-driver.
OpenCL is very good, you can train neural networks very quickly with this technology, hundreds (thousands) of times faster than on an ordinary processor.
This can be corrected as follows: find these lines in the script -
" uint g=(uint)(res[1]*255);\r\n"
" uint r=(uint)(res[0]*255);\r\n"
and replace them with -
" uint g=(uint)(res.y*255);\r\n"
" uint r=(uint)(res.x*255);\r\n"
Thank you!
It did work, but the "sky" has ripples in it.