I want to build a PC to work with MT5, what do you advise and why? - page 49
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
Intel Core i9-13900K BOX - this processor has 24 cores and 32 threads. That's not a lot
Rather, it is interesting to discuss a particular component.
That's what I'm saying. It seems now the first/second generation two-gal epics are becoming an interesting purchase - if for the case.
4 x32 ГБAt such volumes and work with money, ECC is relevant, especially since the procs support it.
And, I would rather look at i7-13700K, the main difference is the number of E-core and hence the cache.With such volumes and money handling, ECC is relevant, especially since the procs are supportive.
Why?
Can you justify it or have you just heard a nice word?
Why?
Can you justify it or did you just hear a nice word?
With ECC memory (even with much older, coarser process memory) in server MB logs there are often messages about memory error correction. This gives an opportunity to replace a module if it has these errors regularly.
With regular memory we don't see these errors, we don't control them. And all the more we do not know how it can manifest itself in money transactions.
For a more or less normal memory test, or more precisely, the memory-pc-mother-power supply (-cooling), it is necessary to run Memtest86+ for several days (one pass, especially the first accelerated one, reveals only obvious defects), and all modules together, and each one separately. There were cases when the hardware glitched, and the joint test of modules did not reveal anything, only one by one. Much more cases when the opposite was true, only together.
what tobacco to load in this hookah? )
Powerful, obviously.)
With ECC memory (even with much older, coarser process memory) in the log of server MBs there are often messages about memory error correction. This gives an opportunity to replace the module if it has these errors regularly.
With regular memory we don't see these errors, we don't control them. And even more so we do not know how it can manifest itself in transactions with money.
For a more or less normal memory test, or more precisely, the memory-pc-mother-power supply (-cooling), it is necessary to run Memtest86+ for several days (one pass, especially the first accelerated one, reveals only obvious defects), and all modules together, and each one separately. There were cases when the hardware glitched, and the joint test of modules did not reveal anything, only one by one. Much more cases when the opposite was true, only together.
in a home numerator, which is also for everyday video with "to play", it's not necessary... memory errors will appear "visually", perceptibly. And which of the cards has failed will be detected by memtest.
Also, by the way. It is somehow not accepted to talk about it, but the reliability of memory on a video card is usually much less. Manufacturers believe that small failures of the picture are not terrible at all.
There was also a task to make water cooling, which I have not encountered before, because of the irresistible desire to get silence and maximum cooling efficiency.
Yura, water cooling has a lot of pluses, but there is one minus, which can spoil the components of the motherboard.
The thing is that you need to cool not only the processor, but also the motherboard. Simple ones warm up slightly and do not need additional
cooling. The case is worse with tricky ones. In short, at first control the elements of the motherboard by hand,
temperature is allowed up to 60-70 degrees, but it is better to keep within 40-50 otherwise will quickly dry up electrolytic capacitors and
as a consequence to fail. Repair of motherboards in most cases is difficult and expensive, it is usually easier to buy a new one.