VGA Charts Summer 2005 at Tom's Hardware

[PrayForDeath]

I did a seach on it and didn't find it, hopefully it's not a repost.
They're testing 3DMark05, UT2K4, HL2, D3, CoR, Pacific Fighters, and Sims 2, and they added the X850/x800xl/np series to the charts.
Link

 

[xtknight]

This was discussed the other day in this thread:

http://forums.anandtech.com/messageview...atid=31&threadid=1598806&enterthread=y

I didn't reply to say you double-posted, basically I replied to tell you that these charts aren't very reliable at all...just check out that thread and you'll know what I mean.

 

[HeroOfPellinor]

Originally posted by: xtknight
This was discussed the other day in this thread:

http://forums.anandtech.com/messageview...atid=31&threadid=1598806&enterthread=y

I didn't reply to say you double-posted, basically I replied to tell you that these charts aren't very reliable at all...just check out that thread and you'll know what I mean.

Yeah, cause your card came out looking bad, the charts are worthless.

 

[dderidex]

Originally posted by: HeroOfPellinor

Originally posted by: xtknight
This was discussed the other day in this thread:

http://forums.anandtech.com/messageview...atid=31&threadid=1598806&enterthread=y

I didn't reply to say you double-posted, basically I replied to tell you that these charts aren't very reliable at all...just check out that thread and you'll know what I mean.

Yeah, cause your card came out looking bad, the charts are worthless.

No, they are worthless because:
A) Tom's Hardware is one of the biggest ATI fanboy sites on the 'net right now
and
B) The numbers don't make sense. The 6600GT *IS* a slower card than the 6800.

6600GT has a 500mhz core with 8 pipelines = 4.0 gtexels/sec vs the 6800's 325mhz core with 12 pipelines for 3.9 gtexels/sec. SLIGHT difference, yes, but, then, look at the ram.

6600GT has 1000mhz at 128-bit for 16 gb/s while the 6800 has 600mhz at 256-bit for 19.2 gb/s.

So, the 6600GT has a slightly - SLIGHTLY - higher fillrate than the 6800, but a HUGE memory bandwidth deficit. And HOW is it legitimately scoring from 50% better to TWICE the 6800's numbers in so many games?

Obviously, Tom's Hardware screwed up bigtime, there. The fact that they could make such a GLARING error in ALL the tests kinda makes you wonder about the rest of the review, doesn't it?