"New Neural" is an Open Source neural network engine project for the MetaTrader 5 platform. - page 57
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
The GPU features in MQL5 will appear soon, so it makes sense to build upon them.
You may find what I say a little strange, but in the vast majority of cases where networks are used, speed is not the bottleneck.
You may find what I am about to say a bit strange, but in the vast majority of cases where networks are used, speed is not the bottleneck.
The bottleneck is not the bottleneck, but any acceleration is highly desirable. It is clear that if you "gap" only the grid, the gain will be small compared to the execution time of the entire program.
High acceleration rates will be given by calculating the whole history runs in the GPU.
PS If the neuronka has impressive size (several thousand weights), paralleling it only will give a tremendous speed gain. If weights are a couple of dozen, then yes, there is no point in fiddling with GPU. But this is a serious project. :)
But this is a serious project. :)
only very slow ((
It's not the project that's slow, it's the participants. :)
But, the Russian man takes a long time to harness, but then he rides fast.
You may find what I say a bit strange, but in the vast majority of cases when using networks, their speed is not the bottleneck.
When used, yes, but will the consumer like a grid that takes months to learn? the question is rhetorical.
That is, the biggest gain in performance from using GPUs is in the learning step.
Andrew TheXpert, if you say A, say B. What do you see as the main bottleneck of the NS?
It's not the project that's slow, it's the participants. :)
But, the Russian man harnesses long, but then rides fast.
Yeah, fast))
For a year already hangs MARKET not working. But it hangs. And wishing in most will still wait.
I mean, if you can do something small but working, and then already hang him on the contrivances, it is better to go this way than years to design a superproject.
-----------------
By the way, today is the birthday of wind ! v 1.01
I mean, if you can make something small, but working, and then already hang the features on it, then it is better to go this way than years to design a superproject.
The evolutionary path of the bicycle will not turn into a spaceship, you understand. Without designing, nothing good will come out anyway.
And you can run to the codebase for a small and working one. :)
A bicycle is not going to evolve into a spaceship, you know.
You can't make a bicycle into a spaceship by evolution, you know. Without engineering, nothing good will come of it anyway.
And you can run to the codabase for a small and working one. :)
I disagree. But it doesn't matter.
Turn on Skype. I'm still not authorized. We need to talk.
???