New Posts  All Forums:

Posts by sumitlian

This.This is one hundred percent exactly what I have been told too by my little human brain when I tried to care about them.And the same is happening in India as well. Their ego gets hurt and they think you are intruding their safe space If you try to give them advise on how transcending past the religion and utterly stupid absolutely worthless Indian cultural sht, can improve the quality of life, overall environment for future generations and for everybody including...
Emotionally stupid comment in ^here.You ought to, instead, tell that to those who approve $1 Trillion just to take care of a single type of fighter jet. Do you even have a slightest idea on what tiniest amount of money is spent on scientific research and space exploration as compared to unnecessary military equipments, Globally ?
Who else is left to do so when whole congress bow down to Israel ?
Wrong!Infinity fabric will only come into play when CPU core interconnection* happens, other than that each core is fully able to read/write data into memory through its primary path(core#x->L1->L2->L3->RAM). You are going to get full benefit with fast RAM in almost every type of multitasking scenarios. Overall operating system will run smoother. It is games only that need very tight connection between threads that may not get benefit from that fast RAM.Also thats other...
While Russia hacks USA, Israel hacks the world.
MaxxMem or MaxxMEM2_preview was very famous in 2010-2012 I think.http://www.softpedia.com/get/System/Benchmarks/MaxxMEM2.shtml or google for more versions.
Not all caches. Only L3 one. Afaik L1 and L2 speed is fixed or it is may be tied to core speed by some multiplier, I don't know (because we still don't know the exact performance capability of L2/L1 since both Intel and AMD don't publish the internal clocks of it, I think you can know the bus width of L1/L2 on Internet). All you can do is increase core speed to max or wherever it reaches, since core's IPS is increased and now you do some tests and see if there is...
Moral is these tests have got almost nothing to do with real world performance. I can literally make two cache test benchmark and both of them will show very different scores in all CPU. It all depends on what calculations you do and what optimization you choose. For me, all cache benchmarks are 100% worthless as you can never guess or confirm any of the real world application performance based on those cache/bench tests whether it is from AIDA or Sandra.
@chuy409 Also I have a 2012 record of Sandra (probably one from 2011-2012) on PII. It is in line with what you showed (at least in L1/Integrated one), but again, that old Sandra is showing much higher L2 and L3 cache performance as compared to what you and Artikbot showed with current AIDA tests. Phenom II X6 1055T @ 4.0 GHz (286 x 14, NB = 286 x 10) P.S. I don't know what I was thinking back then, I only cropped the main score side from whole window and kept all the...
I get what you are saying. The confilcting scores between those versions is a proof that we must never conclude the ultimate benchmark scores of a certain CPU by any of the AIDA64's version. Whatever size of data they are using with AIDA64 v5.80 is only showing that it is fitting(cache hit) on Phenom II's cache architecture much better than it was with older versions. There are countless methods with countless data size you can test a CPU with to test cache/mem bandwidth...
New Posts  All Forums: