Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: wanna_bmw who wrote (56766)	10/1/2001 3:15:26 PM
From: pgerassi	Respond to of 275872

Dear Thread: The best games for CPU testing are those that use lots of CPU resources yet small amounts of graphics resources. The best games for this are the simulation types. They have to do a large quantity of floating point mathematics for each object (the better for realism) and the number of objects may be very large. Thus, most simulations have yet to exceed 30fps even in low resolutions. To further stress a CPU, one could use software based rendering. That will show which CPU does the best and has the highest performance. Most 1st person shooters take quite a few shortcuts that use little CPU power to get a somewhat more realistic effect on the screen. Those that do not do the shortcuts, look and feel much more realistic, but a very high CPU power cost. Movies are shot at 24fps, TV displays 29.97fps but redraws every other line which doubles the apparent frame rate and computer monitors draw 75fps to 100fps for that "flicker free" effect. The reason that movies and TV can get away with the lower rates is that they capture a scene over a short period averaging what happens. To simulate this digitally requires the use of many subsamples added together for each frame. This shoots the needed frame rates much higher as found when antialiasing is done. For "natural real world" effects, frame rates would probably need to be in the 1000s per second and the polygon counts somewhere in the 10^9 range. Studies have shown that a tree with 1000 leaves simulates quite well. Most games use a static image of two or three views of a tree and set them at perpendicular angles to each other. This looks ok at high speeds, but terrible at low speeds. A forest could be simulated by 1000 trees and close objects by 10K to 100K polygons. Adding wind, rain, sunlight and all of the other environmental parameters, yields a quite realistic scene. Now think of doing all that for each of two eyes per person playing the game. We are no where near the CPU and graphics power required to do this. So what we have may be good enough for 1st person shooters of low quality, but it is not where we need to be for those games as yet beyond our top of the line cutting edge hardware and software. And we have not even started on opponent and friend AI. Pete