Thank you all for your results..this is a nice start..and i'am happy to see that nobody looks to have problem with my explenation.
If you know the program,it takes 2 seconds to start the test.
With this chess program i want to show people how powerfull these cores are.
I see on many forum (for sure gamers) what is the use to have many cores,but it's of course that most programs are behind in programming to use these multi-core cpu's.
Chess is always ahead in these things,we where years waiting for multi-core cpu's
When the first dual 2 core come out we have already a chess engine that can handle 2048 cores!!
But the programmers see it goes to slow..we are now only at 8-16cores (i know they are testing 64 an so on) but is still far away from the 2048 cores.
So they find a next way..make a cluster!
Take begin 2010 and we see cluster chess engines ,means we can connect different computers to each other and the chess engine will use all the cores from these connected computers! Then it will come interesting!!
This engine is now already playing tournaments on a cluster from 56cores!
11 results in good one day is very nice..also results with HT OFF are welcome like some people already have done,just to compare them with the normal quads!
When the list growing it comes more and more interesting!
Hwy JP, time to take the i750 up a notch, still at stock speeds. It seems that setting 8 threads on 4 core no HT CPU is not good, so here it is with 4 threads
__________________
Suma prostiei ramane constanta. Doar prostii se schimba...
Nu vrea. Mai sus de 450 FSB n-am reusit sa-l duc. Si atunci destul de instabil.
Si in mare parte cred ca de vina sunt temperaturile pe NB...ori nu sta bine heatsinkul ori e pasta foarte proasta...sau eu stiu...e totusi cald/fierbinte la atingere deci ceva caldura tot disipa...
As you can see, cache sizes make almost no difference in this test, so I assume the data it has to work on is small, therefore limiting the number of cache misses.
poparamiro's last result using his Celeron E1200 is almost identical to my result using a mildly overclocked E8200, even with a 10MHz FSB difference in my favor.
I will test this because I think I'm right, my cpu giving same results @ the same frequency but with a different FSB.
I will give this test a go, again @ 6x400 MHz and also @ 8x300 MHz to see what the results would be.
Later edit : I also think that memory bandwidth doesn't make a huge difference either.
Viziteaza-ne pe Facebook
Urmareste-ne pe Twitter
Vezi ce facem pe YouTube