Aktueller Inhalt von Hans de Vries

It seems that Intel designed this Excel Big Number Crunch test program with macros to use instructions not available on AMD processors before Bulldozer (SSE4?) If you have the time to figure out how Intel managed to skew the results so much you could start to Google for: "Intel provided the...

Thanks! The whole idea (in my opinion) behind these FMA's is that they want to improve Floating Point IPC even while going to a speed racer design. So the latency in cycles has to grow because of the higher clockspeed from 4 to 6 but the ability to fuse multiply and adds would reduces the...

OK, man kann die 44 sekunden im 720p Video erkennen. Der score ist dan eher 19641, die 7 und die 9 sind sehr schwer zu unterscheiden im Video. http://www.tweaktown.com/reviews/2779/intel_core_i7_975_3_33ghz_processor_tested/index6.html Grüße, Hans

Von der video mit 720p: Maxon Cinebench 10 (64 bit) System XXXX: time: 71s, score 11361 Sandy Bridge: time: 52s, score 17641 Grüße, Hans . EDIT : . Der SB hier ist genau so schnell wie ein Core i7 940 mit 2.93 GHz http://www.legitreviews.com/article/986/8/ Grüße, Hans

httx://wwx.electronics-eetimes.com/en/globalfoundries-to-spend-363-billion-on-expansion.html?cmp_id=7&news_id=222902177&vID=44 (change x to w to follow link) Grüße, Hans

Richtig. Mit der bessere auflosung ist es zu sehn das der GPU/IF vollig gleich ist an der vier CPU interfaces. Grüßen, Hans

Mal probieren ob das geht: httx://lh6.ggpht.com/_zp7uCOVms-w/SluHAjkwCqI/AAAAAAAAApQ/2WatRyr-8jA/Final%20-%20Compressed.jpg (oder mit %20 --> " ") httx://lh6.ggpht.com/_zp7uCOVms-w/SluHAjkwCqI/AAAAAAAAApQ/2WatRyr-8jA/Final - Compressed.jpg OK :) Die fünfte ringbus interface ist...

The GPU uses 2MB of the 8MB L3 on Sandy Bridge's die. There are 5 identical interfaces to the L3 ring-bus visible. They all look like the logic part between the two halves of the L3 cache associated with each core. The 5th ringbus interface is located between the GPU and the L3 cache...