AMD or Intel as well as the memory brand - page 6
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
It has been proven for a long time that AMD does not heat more than Intel, and sometimes even less! There are just adherents of Intel, who have seen Socket A processors last time - Athlon 1000. For them their acquaintance with AMD is over.
By the way, the computing centre and all IAPO computers are AMD-based =)) Before that they checked which one was faster in working with cads of all types. AMD outplayed Intel by all means. And at the same time it is at least 2 times cheaper.
Yes, AMD has long ago got rid of its childhood pains. That includes power voraciousness, fragile unprotected chip, and many others...
Now:
Intel - multimedia.
AMD - math calculations.
Although, AMD's proceses now even include a graphics accelerator with hardware support for video stream encoding.
I need a hardware H.264 decoder!!! Otherwise, how can I see the small time frames and ticks? You can only see them in high-definition!
AMD's math calculations are said to be faster? Which ones and on what (software)? SuperPi is faster on Intel, for example. Same high definition coding/decoding (a lot of maths) is faster on Integer than on AMD with equal frequency. Can I get you some test references?
Neither are you talking about it. You just have to run optimization on comparable stones and see which is better.
Svinozavr писал(а) >>
You just have to run the optimisation on comparable stones and see which is better.
Oh! Exactly!
I should write a script for performance tests.
The algorithm goes something like this:
- remember the time
- execute type1 operation
- compare and remember the execution time for operation type1
- //- for type2,3,4 ...
- terminate, print run times for each type in the alert
No need long, about a minute on an average machine.
The types of operations are purely conditions, we grind the dubles, divide the ints on the second, etc...
Nor is it an idle interest. You just need to run optimization on comparable stones and figure out what is better.
a little correction with your permission... Not on a comparable, but on an equal price. Perhaps the developers will commission such tests from independent test labs. After all, the user of an MT5 product has a right to know on what hardware it will work better. Besides, it's not just idle interest.
ZS. you should not compare CPU frequency, but cost at equal performance, performance, at equal cost.
a little correction with your permission... Not on a comparable, but on an equal price. Perhaps the developers will commission such tests from independent test labs. After all, the user of an MT5 product has a right to know on what hardware it will work better. Besides, it is not of idle interest at all.
I agree - on price. That would be interesting. Again, I stress (for the umpteenth time!))) that it's about optimisation, as trading (manual, robot) can also be done on a 286 20Mhz. (I may have bent it, but maybe just a little bit.)))
I need a hardware H.264 decoder!!!
then I suggest you change your CPU to this stuff!:))))
http://www.arm.com/products/CPUs/ARM1176.html
then I suggest you change your CPU to this thing!:))))
http://www.arm.com/products/CPUs/ARM1176.html
how can you compare harvard to a classic architecture.
Unless android or some other sane thing is available there is no point.
then I suggest you swap your CPU for this thing!:))))
http://www.arm.com/products/CPUs/ARM1176.html
Why would I need a bare RISC CPU? There are ready-made hardware decoders/encoders in the form of expansion cards. I'll fucking decode a minute's worth of history into ticks!!! ))))))))))))))))))))),
Why would I need a bare RISC processor? There are ready-made hardware decoders/encoders in the form of expansion cards. I'll fucking decode minute history into ticks!!! ))))))))))))))))))))),
and the ticks into traders' names and the names of the banks making the trades:)))
To the iron.
If anyone is interested and has time, we can make a test EA and post it here. Whoever wants can run it - start/end date, pair, timeframe should be postulated in it, so that all have the same result. The history is mcsh. Well then post the result (run time) with specifying the comp configuration. It would be interesting to see what depends on what and how.