AMD or Intel as well as the memory brand - page 15
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
See Source column, source extensions.
See Source column, source extensions.
А... >> I see. You know, when I was making my suggestion, I hadn't read your type on the picture yet. There just wasn't anything else to do. (I dismissed the photoshop right away!!! ))))))
Well, I wasn't that far from the truth about the incense...
Thanks, great news!
And here's the surprise.
Yeah. A surprise indeed. Compared to my first test, more than five times the acceleration! That's already quite a serious breakaway. And just imagine - quite recently (just 3-4 years ago) people were testing their systems on configurations even weaker than mine. Progress is moving at a rapid pace.
I have now answered my own question above. Yeah, you've got to speed up. Otherwise you risk lagging behind the main mass. A five-fold lag is too much lag (although you'll have to test it on a five).
Imagine how happy I was when I replaced a year ago the Stump (still old, Northwood, 1.8 GHz) with E7200. The speed of calculations has grown not less than 2 times, and in some cases - from 4 to infinity (when Stump refused to count at all). But here it was not only the replacement of the stone, but faster memory had its effect.
Besides now I've stopped noticing the antivirus work at all.
Have you tried optimization?
By the way, does anyone know what the situation is with the use of video cards for computing? I've searched through the forum and found this statement by Renat:
"куча других способов нарастить аппаратную производительность" является маркетинговым бредом производителей, пытающихся пристроить свои продукты (речь об CUDA) куда угодно. Аппаратно производительность обычно повышают на десяток (грубо) процентов в год, а вот алгоритмически (софтверно) производительность обычно повышают в разы или десятки раз.
In short, CUDA doesn't rule. But algorithm optimization does.
I completely agree with the latter. Suffice it to look at my tests above - testing the whole script at once and testing it separately - the results are very different. Although I still need to comprehend how this must be taken into account in algorithmic optimization.
But I would like to ask: maybe someone has got results of comparative tests in MQL4 using CUDA and without it.
PS. Mathemat, could you please show me your configuration with this programme? It will be interesting to see. :-)
By the way, does anyone know what the situation is with the use of video cards for computing? I've searched through the forum and found this statement by Renat:
In short, CUDA doesn't rule. But algorithm optimization does.
I completely agree with the latter. Suffice it to look at my tests above - testing the whole script at once and testing it separately - the results are very different. Although I still need to comprehend how this must be taken into account in algorithmic optimization.
I would like to ask who has the results of comparative tests in MQL4 using CUDA and without it.
CUDA is the same parallel computing, only not on CPU cores but on gr.card processors. Forget both the former and the latter for MT. There are none and are not expected to be.
PS. Mathemat, could you please show me your configuration through this program? It will be interesting to see. :-)
There's nothing special there. >> Let's give it a try. There's no graphics card (the graphics are built in). If you need anything else, you tell me.
CUDA is the same parallel computing, only not on CPU cores, but on gr.card processors. Forget both the former and the latter for MT. There are not and will not be.
No, I just didn't phrase my question correctly. I know that you can't work with CUDA directly from MT. It seems that some C-like language is used there.
I wanted to ask whether it is worth bothering with CUDA and the like. Is there a real gain in speed (by "real" I mean at times, not by 10-20%).