Intel: Core i7 to Be Up to 52% Faster Compared to Core 2 Quad


Recommended Posts

heh, you are not making any sense here. There's no way to measure this so-called "CPU processing power in 3D gaming" of yours. What's the "CPU processing power in 3D gaming"? is it AI, physics, data loading, data decoding, or what? That basically covers everything out there, and varies greatly from game to game. In one 3D game it may have complicated AI but simple physics, in another 3D game it may have complicated physics but simple AI, and the "CPU processing power in 3D gaming" are about completely different tasks, favoring completely different CPU architectures. The "raw calculations regarding games" is completely undetermined, so talking about a certain percentage of "performance boost" in "raw calculations regarding games" is completely non-sense. :D

Because video encoding is 100% done on the CPU, so we can measure the 40% increase in performance for sure. On the other hand, there's no way to measure the "calculations related to gaming operations" since there's no such fixed set of "calculations related to gaming operations", thus it's unrealistic to talk about "performance boost" percentages regarding "calculations related to gaming operations", and it can't be determined at all. So by your interpretation of that line, Intel must be pulling some random numbers out of their asses :laugh:

Yes it isn't unrealistic for encoding to gain 40%, and it isn't unrealistic for gaming to gain some performance, but it's completely unrealistic and ridiculous to pull out a 52% figure out there, no matter how you look at it. It either can't have that much of increase (52% fps increase with 3D accelerated graphics), or it's absolutely impossible to measure the increase in any determinable manner (your so-called interpreted 52% boost in "calculations related to gaming operations"), or it's just mostly useless (52% fps increase in software rendering mode) :D

Theres specific operations that will be performed by a CPU which will be called more often based on the sort of task at hand. Video encoding will favour certain operations to perform the video encoding and likewise gaming will have specific instructions that are also used more often then others. Clearly what Intel is saying is that based on their research of what operations each application performs on the CPU, those opperations used most often for video encoding has received a 40% boost in processing time and those favoured by games (be it physics calculations, AI calculations whatever) amounts to around a 50% gain.

Yes each game is different, noone is denying that. But so are encoders. Look, maybe YOU can't measure what operations in a game or encoder are being done by the CPU but do you really believe that Intel isn't testing inputs and outputs for the CPU's under certain conditions. Of course they have theoretical and probably practical ways to measure performance under gaming conditions. No, the operations aren't fixed, thats why clearly these figures are "up to 52%" and "up to 40%". It's not a static figure to be seen by every game just as every encoder won't see the same 40% improvements. Exactly why do you feel it is impossible to measure the increases under game but not impossible under an encoder. It's not like every other piece of hardware in the system can't be identical if need be and I'm sure Intel can test to a far far lower level than that. Hell they probably wouldnt even need a game, just a benchmark that spits out operations a game is likely to use and test how long it takes to do a few million of those.

Hell what is a graphics card? Really it's a CPU which has been optimized to perform routines needed by games as well as handling outputs ect.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.