OpenCL: internal implementation tests in MQL5 - page 35
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Here is some more information for Renat. When uninstalling the SDK, I got a warning message like this:
That is, although the terminal does not recognise the stone as an OpenCL device, but something in the SDK is still somehow trying to do something. Maybe it is a terminal problem?
There's a problem: I decided to uninstall AMD APP SDK (it worked) and put Intel OpenCL SDK. But it didn't work.
...
What should I do to make this thing work? I don't have video card but still want to run code on GPU integrated processor.
You don't need it, you have GeForce 9600M GT. What kind of stone do you use?
Just for the sake of experimentation. ))
Well, I have a bare stone with Intel HD Graphics inside.
If your AMD card is AMD then nothing better. And this "alien" SDK has showed itself very well on Intel HD Graphics. And it got up immediately unlike the Intel one.
But Intel made its SDK for a reason, especially for its own stones.
At the very least, install AMD drivers and feel the difference.
I also don't have any AMD hardware at all and yet everything runs smoothly.
The procedure is simple: install native drivers for all devices. Then we take AMD OpenCL driver software and run it. The utility will dig into hardware and will set up everything as it should be. Reboot your computer and enjoy your life.
You can also install AMD SDK, if you have a desire to program something. In this case, the drivers will be set automatically.
2012.03.05 01:42:20 Terminal GPU: NVIDIA Corporation GeForce 8500 GT with OpenCL 1.0 (2 units, 918 MHz, 256 Mb, version 285.62)
2012.03.05 01:44:56 ParallelTester_00-01x (GBPJPY,H1) CpuTime/GpuTime = 9.809641532756489
2012.03.05 01:44:56 ParallelTester_00-01x (GBPJPY,H1) Result on Cpu МахResult==1.79346 at 725 pass
2012.03.05 01:44:56 ParallelTester_00-01x (GBPJPY,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.05 01:44:56 ParallelTester_00-01x (GBPJPY,H1) CPU time = 55552 ms
2012.03.05 01:44:00 ParallelTester_00-01x (GBPJPY,H1) Result on Gpu МахResult==1.79347 at 725 pass
2012.03.05 01:44:00 ParallelTester_00-01x (GBPJPY,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.05 01:44:00 ParallelTester_00-01x (GBPJPY,H1) GPU time = 5663 ms
2012.03.05 01:43:55 ParallelTester_00-01x (GBPJPY,H1) OpenCL init OK!
2012.03.16 12:18:09 Terminal CPU: GenuineIntel Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz with OpenCL 1.1 (4 units, 3311 MHz, 8174 Mb, version 2.0)
2012.03.16 12:18:09 Terminal GPU: NVIDIA Corporation GeForce GTX 570 with OpenCL 1.1 (15 units, 1464 MHz, 1280 Mb, version 296.10)
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) CpuTime/GpuTime = 120.8
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) Result on Cpu МахResult==1.37999 at 297 pass
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) CPU time = 15100 ms
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) Result on Gpu МахResult==1.37999 at 297 pass
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) GPU time = 125 ms
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) OpenCL init OK!
2012.03.16 12:18:09 Terminal CPU: GenuineIntel Intel(R) Core(TM) i5-2500 CPU @ 3.30GHz with OpenCL 1.1 (4 units, 3311 MHz, 8174 Mb, version 2.0)
2012.03.16 12:18:09 Terminal GPU: NVIDIA Corporation GeForce GTX 570 with OpenCL 1.1 (15 units, 1464 MHz, 1280 Mb, version 296.10)
Traitor!!!
--
;)
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) CpuTime/GpuTime = 120.8
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) Result on Cpu МахResult==1.37999 at 297 pass
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.16 12:34:08 ParallelTester_00-01x (EURUSD,H1) CPU time = 15100 ms
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) Result on Gpu МахResult==1.37999 at 297 pass
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) Соunt inticators = 16; Count history bars = 144000; Count pass = 1024
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) GPU time = 125 ms
2012.03.16 12:33:53 ParallelTester_00-01x (EURUSD,H1) OpenCL init OK!
Put it in sequential order, please:
4096
8192
16348
32768
65536
Let's see if you're worth being afraid of... ;)
//---
I can't work out a satisfactory theory why my system's time isn't growing up to 12800 concurrent threads and only after that it starts behaving as if from that moment the "out of place" processes are queued. The card's spec says 1280 "bees", i.e. 10 times less. In short, I must have been carelessly pushed 10 times more cores. Or maybe I got into some holiday action... :)
At CountPass 262144 the video card driver crashes unconscious. True, it immediately comes back up.
PS my card has 480 flies in total.
PPS It can be seen that on lower CountPass values my card is faster. But from 65536 onwards it starts lagging behind yours.
PPPS Your card is cheaper, but shows results comparable to mine.
PPPS Your card is cheaper but shows results comparable to mine.
... Which you could have read about beforehand on overclockers.ru :)
By the way, how much did you pay?
// I bought a Radeon 6930 for 5650p