X1800XL Performance numbers - VR Zone
- Thread starter russo121
- Start date
Iron Tiger
Regular
Maybe in HL2 and possibly even KOTOR2, but Chaos Theory and EQ2?Hellbinder said:
78.01 and 5.8 are used.dizietsma said:
TWISI
(The Way I See It)
the best way to check D3D performances is trough 3DMarks, and the only way to check OGL performance is trough only significant OGL game title – D3. We can argue about id “bending†toward ‘vidia, but if you take in account popularity, and the numbers of AAA titles that are to be based on D3 engine, I think that there are enough reasons to “cherry pick†those three scores. In a case of 3Dmark03 600 points is not so big, but in the case of 3Dmark05 600 points represents much bigger gap in performance.Henceforth, Transparency AA will be referred to as TrAA and Temporal AA will be referred to as TeAA. That way, I will stop getting headaches.
EQ2 is a dx8.1 game, splinter cell looks better on a SM3 card , HL2 is cpu limited and star wars doesnt used an advanced engine.Tim said:
How about using more of todays games (dx9) to judge performance.
I'd like to see far cry numbers as well as battlefield 2.
The reason I did not throw in FC results is ‘cos this game is so damn CPU dependant!! Here is example (demo is a mix of indoor and outdoor):
X800GTO 1024x768
On Athlon64 3000+ scores 83.8 fps
On Athlon64 4000+ scores 102.8 fps
On the other side HL2 (Canals demo – indoor demos are more CPU dependant) scores like this:
On Athlon64 3000+ scores 109.6 fps
On Athlon64 4000+ scores 118 fps
DOOM3 is also much better in comparison to FC:
On Athlon64 3000+ scores 80.4 fps
On Athlon64 4000+ scores 85 fps
X800GTO 1024x768
On Athlon64 3000+ scores 83.8 fps
On Athlon64 4000+ scores 102.8 fps
On the other side HL2 (Canals demo – indoor demos are more CPU dependant) scores like this:
On Athlon64 3000+ scores 109.6 fps
On Athlon64 4000+ scores 118 fps
DOOM3 is also much better in comparison to FC:
On Athlon64 3000+ scores 80.4 fps
On Athlon64 4000+ scores 85 fps
Sunday said:78.01 and 5.8 are used.
TWISI
(The Way I See It)
...right...
so in D3D, synthetic benchmark is ok, in OGL, you need real game, whose engine will be used on some titles in future, and using actualy game benchmarks on D3D is out of the question, of course?
Far cry cpu depentant?Sunday said:
It scales well with 7800gtx sli systems
http://techreport.com/reviews/2005q2/geforce-7800gtx/index.x?pg=7
HL2 on the other hand does not.
http://techreport.com/reviews/2005q2/geforce-7800gtx/index.x?pg=8
You assume I have some idea what adaptive AA is. I probably do, but I sure can't think of what it is. I think it's more of "I'm sick of remembering antialiasing acronyms." Oh hey. There's another one. SIRAA. It almost looks legit...BRiT said:And what will Adaptive AA be refered to? AAA, AdAA, ApAA, or UAA (UberAA)
3Dmark5 uses DST for the shadows on the GT which boosts its score ~ 11.5% (at least that is the boost on the 6800 series …elitebastards) So the GT is running a different render path than the X850. 3DMk5 scores would be very close without DST …Sunday said:78.01 and 5.8 are used.
TWISI
(The Way I See It)
7800GT … ~ 4262 …(with no DST)
X850 XT … 4142
In 3DMark2003, AFAIK NV still hasn’t removed their “questionable” optimizations in that bench.
For the game benches, the GT was likely run in only Q mode (shimmering mode). Running the GT in HQ mode will reduce performance by as much as 25% (and probably generally in the 10-15% area).
Similar threads
- Locked
