AMD or Intel as well as the memory brand - page 27
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
Oh, come on. Dream on, that's enough. :-)
For me, stock exchange is a game, not a professional occupation. If I did it as a business, then you could really think about how to acquire such "power".
Honestly, I don't even know exactly how to use such power in a way that pays for itself. With my current level of "technical" development I definitely can't use it to my advantage.
The difference between the computer from Svinozavr on Celeron for 16800 and this "beast" for more than 100.000 R. 27/52 = 0.519 - less than twice. In my opinion, the twofold lag in our case (a single trader) is not pripical. Suppose I optimize my system for two days, and my friend optimizes it for twenty-four hours. Of course, it is more convenient. But not so much to collect the last money, go into debt and buy a "beast" for more than 100 grand.
In general, the emphasis should be placed on algorithm optimization, refactoring, etc. rather than on building more powerful hardware.
The factor performance * stone frequency (sec*Hz or sec/sec - no dimensionality). Let's call it m-m-m ff - frequency factor. The lower its value, the more efficient the stone is being used.
Mine for script ff=52 sec * 2.2GHz = 114.4; for Expert ff=206 sec * 2.2GHz = 453.2
Mathemat respectively 46 sec * 2.53GHz = 116.4 and 169 sec * 2.53GHz = 427.6
From the above figures, it is clear that my stone is more efficient on small code tasks, while Mathemat is more efficient on larger tasks, where larger cache and RAM speeds start to play a role.
Such is the case... (Kurt Vonnegut)
Oh, come on. Dream on, that's enough. :-)
For me, stock exchange is a game, not a professional occupation. If I did it as a business, then you could really think about how to acquire such "power".
To be honest, I don't even know exactly how to use such power in a way that pays for itself. With my current level of "technical" development I definitely can't use it to my advantage.
The difference between the computer from Svinozavr on Celeron for 16800 and this "beast" for more than 100.000 R. 27/52 = 0.519 - less than twice as much. In my opinion, the twofold lag in our case (a single trader) is not pripical. Suppose I optimize my system for two days, and my friend optimizes it for two times faster - 24 hours. Of course, it is more convenient. But not so much to collect the last money, go into debt and buy a "beast" for more than 100 grand.
In general, the emphasis should be placed on optimizing the algorithm, refactoring and so on, not on developing more powerful hardware.
))) I agree, my colleague.
Although for me stock exchange isn't a game, but a professional activity (since 2004 I have no other sources for living), but to look for the reason of my failures not in myself, but in something else - for example, in weak hardware - means to deceive yourself.
It's as if an artist's skill is determined by the quality of brushes, canvas etc. they use. Or - this is just close to me - if I was paid like Alexey, I would write like Lev. (referring to Alexei Tolstoy and Lev Nikolaevich.) That's not the point, in short.
Let us remember, comrades, the great French philosopher and scientist Blaise Pascal! "Thinking well - that's the basic principle of morality." Or our poet-obelieutist Zabolotsky: "Don't let your soul get lazy." Which in this context means work on yourself and healthy self-criticism. So that, you know, it won't be like other poet Kharms: "The theatre is closing. We're all sick."
Oh, man. That's the end of the sermon - an address to the traders' professional cell. Dixi. Amen.
Factor performance * stone frequency (sec*Hz or sec/sec - no dimensionality). I'll call it m-m-m ff - frequency factor. The lower its value, the more efficient the stone is used.
Mine for script ff=52 sec * 2.2GHz = 114.4; for Expert ff=206 sec * 2.2GHz = 453.2
Mathemat respectively 46 sec * 2.53GHz = 116.4 and 169 sec * 2.53GHz = 427.6
From the above figures, it is clear that in small code tasks my stone is more efficient, and in larger tasks, where larger cache and RAM speeds start to play a role, Mathemat's is more efficient.
Such is the case... (Kurt Vonnegut)
1. Forgot about me? :)
2. The price of the solution should be added to the formula.
1. Did you forget about me? :)
2. It is necessary to put into formula the price of solution.
1. No, I didn't!!! ))) What am I, an arithmomaniac? Like an IBM from the beginning of last century or "raine-fucking-metal-nah-fuck"? (Don't beat me - it's not mat - it's a quote from Dovlatov. Mathemat will confirm it, if anything... True, it's about a typewriter, but this brand of arithmometers was there too.))
I just gave this as an example. I'm waiting for you to do the rest (arithmetic). Then we'll make a table, because everything is scattered around the branch - it's difficult to navigate. And here's something else: where is the top starter? What's he skimping on? Why, you know, started a theme, and himself with a quiet smile of a teenager watching us here trying. It's like army school. Arbeiten.
2. It's more complicated with the price. We'd better decide on the stone's efficiency. And then we will recalculate the price. In addition to the stone, you will have to take into account the cost of the motherboard and the memory, as a minimum.
What are we waiting for?
;)))
In principle an answer has been received, and I didn't expect such a lively interest in the topic.
Yes, in principle an answer has been received, and in fact I did not expect such a lively interest in the subject.
Answers to the studio, please.
Although, I've already announced the answers myself about 15 pages ago. Intel is traditionally strong in "straight line" computations, such as encoding/decoding (mp3, video content, MT calculations) and AMD is and has been strong in such applications as mathematical calculations, where program logic consists of complicated transitions, engineering calculations.
so, gentlemen traders, stumps and their children for trading and optimization are the best (I mean stumps and their children)
What are we waiting for?
;)))
In principle an answer is received and I didn't expect such a lively interest to the theme.
I can't say anything about the stone, it's up to you to decide.
About memory - the more the better (our experience shows that): Samsung (original) or Kingston (the main manager of a large computer company said, when I was replacing a faulty patriot).
I've been messing around a bit with the spoofs here. The blue highlighted is where both tests are already. I might have got the numbers wrong - correct me.
It's up to you to draw your own conclusions. Not everything that seems fast in the light of Svinozavr's criterion turns out to be fast. And vice versa. By the way, BLACK_BOX's result for the script seems surprising. But maybe I was mistaken in frequency.
And one more conclusion: old Intel stones fail in comparison with modern AMD ones. With newer ones the situation is quite different.
To complete the picture, of course, it would be desirable, that those which have shown only the first test, would show the second.
Athlon X2 3800 (2000 MHz), cache 1 MB / DDR2 PC-5360 2GB RAM (joo): script ff=82.07 sec*2GHz (?) = 164.14; expert ff= 310 sec*2GHz = 620
core 2 duo 1800 MHz laptop, cache 1 MB (?) / 2 Gb (DDR2-667)(keekkenen): script ff=75.77 sec*1.8GHz = 136.39;
Core 2 Duo E7200@2.53, cache 3 MB / 4GB RAM (Mathemat): script ff=46.27 sec*2.53GHz = 117.06; expert ff= 169 sec*2.53GHz = 427.57
Celeron 900 @ 2.20 GHz, cache 1 MB / DDR2 PC-6400 2 GB (Svinozavr): script ff=52.18 sec*2.2GHz = 114.8; expert ff= 206 sec*2.2GHz = 453.2
Celeron 325 @ 2.53 GHz, cache 256 K / DDR2 PC-6400 1 GB (benik): script ff=103.3 sec*2.53GHz = 261.35; expert ff= 438 sec*2.53GHz = 1108.14
Pentium 4 670 @ 3.8 GHz, cache 2 MB / DDR2 PC-4266 2 GB (begemot61): script ff=78.57 sec*3.8GHz = 298.57; expert ff= 169 sec*3.8GHz = 642.2
Celeron 430 1.8GHz, cache 0.5 MB / DDR2 PC-5333 2 GB(kombat): script ff=68.53 sec*1.8GHz = 123.354;
Athlon 64 X2 4200+ @ 1 GB (?), cache 1 MB / DDR2 PC-3200 (?) 3 GB(BLACK_BOX): script ff=77.84 sec*1.0GHz (?) = 77.84;
Core 2 Duo Q8200 @ 2.33 GHz, cache 4 MB (?) / RAM 4 GB PC-6400 (forex-k): script ff=46.84 sec*2.33GHz = 109.14; expert ff= 189 sec*2.33GHz = 440.37
Phenom II X3 720 @ 3.72 GHz, cache 1.5 MB L2 + 6 MB L3 / RAM ? (Belford): ff=37.91 sec*3.72GHz = 141.03; expert ff= 113 sec*3.72GHz = 420.36
AMD 5050e @ 2.6GHz, cache 1 MB / RAM 8 GB PC-5970 (four2one): script ff=60 sec*2.6GHz = 156; expert ff= 134 sec*2.6GHz = 348.4
Athlon 64 X2 4000+ @ 2.1 GHz, cache 1 MB / RAM DDR2 PC-5625 2 GB(skv.): script ff=80.17 sec*2.1GHz = 168.36
Athlon 64 X2 6000+ @ 3.01 GHz, cache 2 MB / RAM 3 GB PC-6030(lea): script ff=52.85 sec*3.01GHz = 159.08;
Pentium Dual CPU E2180 @ 2 GHz, czche 1 MB L2 / RAM 2 GB(kombat): script ff=55.94 sec*2.0GHz = 111.88;
Opteron 2439 SE @ 2.8 GHz, cache 3 MB L2 + 6 MB L3 / RAM DDR2 4 GB PC-5333(begemot61): ff=42.33 sec*2.8GHz = 118.52
Xeon W5590 @ 3.47 GHz, cache 1 L2 + 8 L3 / RAM DDR3 PC-10670 12 GB(begemot61): script ff=27.53 sec*3.47GHz = 95.53;
Pentium 4 @ 3 GHz, cache 512 MB / RAM 1.15 GB DDR2 PC-2570 (?)(Dmido): script ff=64.49 sec*3 GHz = 193.47;