Public discussion of the formula for calculating the cost of resources in the MQL5 Cloud Network - page 14
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 partisans are silent. They're keeping a military secret.
That's because it's hard to answer that easily. There are many factors to consider that are not obvious at first glance, and the factors are different for everyone.
We need to launch a paid cloud as soon as possible, only then we will see how much and what it should cost - the market! (sorry, I don't know the smiley face pointing up with the index finger)
That's because it's hard to answer that easily. There are many factors to take into account that are not obvious at first glance, and the factors are different for everyone.
How much more do you have to consider? Add, divide, subtract - five minutes.
Over the years...
The owners of the cores got bogged down in complicated calculations...
Cores intended for cloud computing, were breaking down because of the difficult task of calculating the cost of renting themselves...
for 2 cents an hour per core?
Practical or theoretical interest?
for 2 cents an hour per core ?
0.02×24×30×34≈ $490 per month
For 2 cents I would only give up free resources for myself (office or whatever).
Own iron + electricity + depreciation (iron, too, wears out) + noise + cooler + not be able to use all their resources at the moment when it's needed = much more expensive.
And I would rent resources at a higher price without thinking twice.
For example, I have to optimize my system. Notebook says : wait 100 hours. I say : no, it's too much )
And I buy 100 cores for 1 hour. I think anything less than $10 ($0.10 per core) would be acceptable - it saves me 4 days of time and nerves, counter is not spinning, hardware is not loaded.
For 2 cents, I would only give up free resources for myself (office or whatever).
Own hardware + electricity + depreciation (hardware also wears out) + cooler noise + inability to use all your resources when you need them = much more expensive.
And I would rent resources at a higher price without thinking twice.
For example, an optimization is imminent. Notebook says : wait 100 hours. I say : no, it's too much )
And I buy 100 cores for 1 hour. I think anything less than $10 ($0.10 per core) would be acceptable - it saves 4 days of time and nerves, counter does not spin, hardware is not loaded.
Oh, there is a figure, thank you )
And on the consumer side.
Vendors who are willing to pay $0.1 per hour per core
Oh, there's a figure, thank you )
And on the consumer side.
Vendors who are willing to give $0.1 an hour per core
Why the consumer? Seller too.
I have 2 servers, 4 cores each. I pay about $100 per month for each (colocation).
If they're idle, I'll "hand them over" to recoup the cost.
Calculation is inspiring: $0.10 * 4 cores * 24 hours * 30 days = $288 per server per month (at ideal 100% utilization).
Even if the capacity is used 50% of the time, the colocation will pay for itself (and it already includes power and internet).
Apparently, I was wrong with this figure, it would be much cheaper in reality.
$0.02/hr ($14.4/month) is not much, but $0.05 ($36.0/month) is decent.
.....
So far all calculations are at the level of proposals. Please give your opinion, correct it or offer your variant.
look at budget vending machines on WebMoney
set up a similar system
to create an exchange for these instruments
Quotation of budget automatons' analogues by type of shares
Why the consumer? The vendor too.
There are 2 servers, 4 cores in each. I pay about $100 a month for each (colocation).
If they're idle, I'll "give them away" to recoup the cost.
Calculation is inspiring: $0.10 * 4 cores * 24 hours * 30 days = $288 per server per month (at ideal 100% utilization).
Even if the capacity is used 50% of the time, the colocation will pay for itself (and it already includes power and internet).
I guess I was wrong with this figure, in reality it will be much cheaper.
$0.02 per kernel per hour ($14.4 per month) is not much, but $0.05 ($36.0 per month) is pretty decent.
There is a category of users that will not turn on their computers to optimize Expert Advisors. This category includes:
- those who trade "manually"
- those who trade using Expert Advisors - those who do not have all their cores loaded
- who do not trade at all
- in the future, they may be added to this category of users of services that charge them for viewing advertisements, reading e-mails, etc.
For this category of users, the cost of $0.002 will be normal. Because it is better to sell resources of your computer for $0.002 through MT5 Cloud Network than not selling them at all
(assuming that load of their computers will not affect performance of their computers, large outgoing internet traffic is good enough for them).
The CPU load affects the power consumption, and quite significantly. Plus significant disk operations, logs are erased, but still written, plus the history itself is stored.
The figure of $0.002 is odd, to say the least - it's like giving people a ride for 10 roubles, all the same "on the way". And it consumes extra petrol, the suspension is not eternal either.