View Single Post
Old 05-23-03, 10:51 AM   #53
Joe DeFuria
Registered User
Join Date: Oct 2002
Posts: 236

Originally posted by Morrow
After officially knowing now that both nvidia and ATI cheat in 3dmark03 (anyone surprised?), what do we learn from this incident?

We learn of course that we can no longer trust synthetic benchmarks, isn't it obvious?
No obvious at all.

It's obvious that if the make of the synthetic benchmark doesn't actively persue cheaters and take measures to combat it, THEN we could no longer trust it.

What we have learned from this incident, is that FutureMark will NOT TOLERATE cheating. They will expose it, no matter who does it. It does exactly the opposite of what you are saying. This make 3DMark all the more trustworthy, because they will actively take measures against cheating.

Cheating in games is certainly easier to hide but not as easy to implement because the cheats found in 3dmark03 do not work in games where the camerapath is random.
Wrong. As the PDF explained, drivers can detect a TIMEDEMO is running, not just the application as a whole. Then, the drivers can turn off the "cheat" for the real gameplay. The result: the performance in the timedemo does not reflect performance during actual gameplay.

Another thing we learn from this is that FutureMark now has also officially stated that their shader routines are inefficient!
Wrong again.

They stated that nVidia's replacements are efficient, and DIFFERENT than their own shaders. They do different but similar things. The bottom line is, nVidia hardware is NOT doing what is asked of it.

It's amazing to me how people can read **** that just isn't there.
Joe DeFuria is offline