AMD or Intel as well as the memory brand - page 19
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
Yes, four2one, I was wrong, I just didn't bother writing. But these aren't yet the Celerons that are replacing the budget E2xxx.
Finally Belford has shown what a real mushy AMD can do.
And another thing: if to reduce its frequency (3.7 GHz) to the rated frequency of E7200 (to 2.53 GHz, i.e. about 1.5 times), who knows what will be faster :) So I'm capable of it, but at the limit of my strength (AMD has big problems with frequencies). I could overclock my own E7200 up to 3.7 GHz (easy and easy: stones from the Valley of Wolves are fast and cool), if only the motherboard would be better.
Another test...
A bit of explanation to the picture.
The load graph from Everest is also shown.
The idea was to see how the task could be solved if one core was loaded and the second one was like a mt working off.
However, it all went wrong and was screwed.
As you can see on the first chart of Everest, I turn off the stress test and immediately the script started working at full power.
Before that, it just hesitated and that was it... it didn't make sense to wait, and I'm sick of it...
;)
The load is about 50% and I suspect it's actually 100% of one core.
The notches visible on the "shelf" are the moments of changing the type of script test...
On mt5, as expected, the speed has increased, but the CPU load, interestingly enough, is the same throughout the script.
(Those scripts are clean because of running different programs on my computer, but I'm really eager to give something to the world)
;)))
Yes! I deliberately left the previous result 27.048 in the MT5 image
I was too quick to wait for the quotes history to be loaded...
After that the result levelled up in my readings after repeated runs.
And I could have overclocked my E7200 to 3.7 GHz myself (easy and easy: Wolf Valley rocks are fast and cool), if I had a better motherboard.
I don't doubt it. And up to 6.16 GHz would it be possible ? Phenom II X3 720 BE overclocked to 6.16GHz
Mathemat wrote >>
So it is capable, but at the limit of its strength (AMD has big problems with frequencies).
>>) Far from it.) Overclocking AMD Phenom II X3 720 Black Edition
I don't doubt it. Could it get up to 6.16GHz ? Phenom II X3 720 BE overclocked to 6.16GHz
Far from marginal )) AMD Phenom II X3 720 Black Edition overclocking
New CPU overclocking record set - Intel Pentium 4 running at 8GHz
CPU-Z overclocking records:
World Frequency Records :
Nº 1 : duck reached 8180.4 MHz with an Intel Pentium 4 631 (65 nm) [Details]
MB : Asus Commando (Intel P965/G965 rev C2) - RAM : Team Group Inc. 2048 MB
Nº 2 : ThuG OC Team Italy reached 8179.89 MHz with an Intel Pentium 4 631 (65nm) [Details]
MB : Asus Commando (Intel P965/G965 rev C2) - RAM : MSC 2048 MB
Nº 3 : blind reached 8116.19 MHz with an Intel Celeron 352 (65nm) [Details]
MB : Asus P5B-Premium (Intel P965/G965 rev C2) - RAM : GEIL 2048 MB
Nº 4 : Nordling reached 8113.22 MHz with an Intel Celeron 347 (65nm) [Details]
MB : Asus Commando (Intel P965/G965 rev C2) - RAM : OCZ 1024 MB
Nº 5 : DeDaL reached 8004.08 MHz with an Intel Celeron 347 (65nm) [Details]
MB : Asus Commando (Intel P965/G965 rev C2) - RAM : Corsair 2048 MB
World Frequency Records (AMD K10 CPU) :
Nº 1 : LimitTeam(Sigh qooitry Ultra40) reached 7127.85 MHz with an AMD Phenom II X4 955 Black Edition (45 nm) [Details]
MB : Asus M4A79T Deluxe (AMD 790FX rev 00) - RAM : Apacer Technology 4096 MB
Nº 2 : SF3D OC Gathering - Powered by AMD reached 7000.4 MHz with an AMD Phenom II X4 955 Black Edition (45nm) [Details]
MB : DFI LP UT 790FXB-M3eH7 (AMD 790FX rev 00) - RAM : A-Data Technology 2048 MB
Nº 3 : Sampsa & elmor reached 6898.55 MHz with an AMD Phenom II X4 955 Black Edition (45nm) [Details]
MB : Gigabyte Technology Co. Ltd. GA-MA790FXT-UD5P (AMD 790FX rev 00) - RAM : Micron Technology 2048 MB
Nº 4 : K|NGP|N reached 6893.34 MHz with an AMD Phenom II X4 955 Black Edition (45nm) [Details]
MB : Asus M4A79T Deluxe (AMD 790FX rev 00) - RAM : Corsair 2048 MB
Nº 5 : DeDaL Mrak reached 6820.36 MHz with an AMD Phenom II X4 955 (45nm) [Details]
MB : Gigabyte Technology Co. Ltd. GA-MA790FXT-UD5P (AMD 790FX rev 00) - RAM : 2046 MB
Source :
Fresh from computerworld: Review: AMD, Intel budget chipsets go head to head
I don't doubt it. Could it go up to 6.16GHz ? Phenom II X3 720 BE overclocked to 6.16 GHz.
I know this stone is good.
And my E7200, if I had liquid nitrogen and the right equipment, I could probably overclock it. Wolfdale is a very good thing for overclocking in general.
Listen, I've been thinking a bit about the optimisation test. It's a bit complicated, I mean a lot of gestures have to be made to create the same test environment. The main problem is to provide a run on one test histories - it is not solved directly. I.e. it can be solved, but in this case, the instruction must be attached to the test. Everything else is rubbish.
Maybe someone has a solution that I don't see?
Listen, I've been thinking a bit about the optimisation test. It's a bit complicated, I mean a lot of gestures have to be made to create the same test environment. The main problem is to provide a run on one test histories - it is not solved directly. I.e. it can be solved, but in this case, the instruction must be attached to the test. Everything else is rubbish.
Maybe someone has a solution, which I cannot see?
There is a solution. And it is not as difficult to implement as it may seem at first sight (in terms of the identity of the test data). After all, for most people, it is the optimization test in the terminal that is of the most practical interest, as you have repeatedly said. However, it will be of no use to me anyway since the tester will never provide an opportunity to optimize thousands of variables either in 4 or 5. This is only possible by writing a separate program, which is why I insisted on testing on a script.
The solution may be the following. We take a standard Expert Advisor based on MACD. Let us say from the beginning on 18.09.2009 up to one month ago on M5. The difference in the number of bars between testing participants will amount to fractions of a thousandth of a percent and the results will not be affected. We add 10 "empty" external variables to the Expert Advisor, which do not take part in trading functions. We specify the step and range of variables and go!
One fact is remarkable. A bunch of people ran the script, probably looked at it, and not a single one mentioned that the first test with integer calculations, uses double numbers. There is another bug in the code, but it does not affect the result.
There is a solution. And it's not as difficult to implement as it seems at first glance (in terms of the identity of the test data). After all, it is the optimization test in the terminal that is of the most practical interest to most people, as you repeatedly stated. However, it will be of no use to me anyway since the tester will never provide an opportunity to optimize thousands of variables either in 4 or 5. This is only possible by writing a separate program, which is why I insisted on testing on a script.
The solution could be as follows. A standard MACD based Expert Advisor is taken. A test tool and testing period are agreed upon beforehand. Let us say from the beginning of 18.09.2009 to a month ago on М5. The difference in the number of bars between testing participants will amount to fractions of a thousandth of a percent and the results will not be affected. We add 10 "empty" external variables to the Expert Advisor, which do not take part in trading functions. We specify the step and the range of variables and go ahead!
One fact is remarkable. A bunch of people ran the script, probably looked at it, and not a single one mentioned that the first test with integer calculations, uses double numbers. There is one more bug in the code which will not affect the result.
What I meant was the same story for everyone - roughly speaking, the problem is creating your own symbol. Without offline and editing some MT service files, I don't know how to do it.
If you forgo the reference history, there's no problem.
Z.U. You'll laugh, but I haven't looked! Maybe I should also antivirus it? )))))))))))))))
If you give up the benchmark history, there is no problem.
Z.U. You'll laugh, but I haven't looked! Maybe I should also run an antivirus on it? )))))))))))))))
Something tells me that different quote values will have no effect on the result. Everywhere there will be numbers like x.xxxx. That is, the amount of data processed will be the same for everyone, except for minute differences in the number of bars.
About the antivirus... you probably shouldn't joke about that sort of thing. We're all dealing with real dough. ;)