Discussion in 'Graphics Cards' started by Jelle Mees, Jan 29, 2008.
I love how they never use the same settings for all the test data......
hmm memory bandwidth limited, be better to see how it does against another 256-bit cards, or better yet g92 8800gts sli.
Thats why I dont like or read there reviews, there really biased too.
Don't you understand the words "Highest playable".
They start on the same resolution with the same settings and then they go up and up on every card to find the highest playable settings for the card...
This "new way" of HardOCP testing is a bunch of crap. How do you define a playable frame rate? Additonally, how does a card with so much more processing power (3870x2) do so poorly against a 8800 gtx when every single other site out there has the 3870x2 pwning it?
I have not seen a single favorable review for ATI from HardOCP in the past year, especially with this testing method which is highly subjective.
I think that I am smart enough to determine what a good frame rate is for my needs. I don't need a review site determining how I should play my games.
playable can mean different things to different people. Just run the game on defaults or max and give the results. No need to turn this setting up on one card and not on the other then show a score that is only a few FPS off.
I just don't agree with the way they do it, never have.
+1, I believe video card reviews should be done on the fastest system possible as well..
That is not true!
Quote from their 3870/3850 review:
And back in the days with the ATI X19xx releases HardOCP was always very postive about ATI hardware.
But the HD2900 cards were a joke...
In all honesty, ATi really didn't put out anything to warrant a favorable review in the last year, up untill the HD3XXX series.
I didn't see that review. Their reviewing methods are still very sub-par though.
Ati's 2900xt still did much better than what HardOCP led people to believe however.
Just like most other sites. Not too many had decent drivers for the reviews. While I really don't like thier review method, I really don't think they are biased. C'mon, when the G80 came out, it was the next best thing to breathing air. Bad reviews usually don't come by bias, they come by dissapointment, which is what ATi/AMD have been lately.
That's not true. Only the first ATI 2900XT were very positive, and the very high 3DMark score was a big "wauw" in the beginning.
TomsHardware charts clearly show that the 8800GTS 320MB is slightly faster then the 2900XT:
This is the overall result of 8 games tested with many different resolutions and settings.
The above was tested in DirectX 9, now take a look at DirectX 10 results:
Now, are you gonna say that Tomshardware is biased aswell?
The 2900XT was never better then the 8800GTS while many people beleaved it was equal to the GTS 640...some eaven thaught it was faster.
Unfortunately, I believe that Tom's hardware is biased as well. Tweaktown recently conducted a review of the 2900pro and even it beat the 8800gts 320 mb. The 2900xt smoked the 8800gts.
4 months ago = recently?
The 2900Pro faster then GTS 320MB? LMAO!
Tomshardware biased? LMAO times two!
Haven't you noticed how many ATI Analysis tweaktown has? They clearly prefer ATI and whats keeping them from messing around with the scores...
They reviewed so many ATI drivers but never Nvidia drivers, doesn't that tell you anything? Think about it...
Have you noticed that Tom's Hardware never reviews new ATI drivers? I don't see what's funny about the 2900pro being faster than the 8800gts 320. The 2900pro IS physically a faster card. Look at the specs.
Anyway, the pro really shines when you speed it up to over XT speeds. With good air cooling, you can hit 1 ghz on the core and easily pass the 8800gtx.
Well i'm not sure what all the hoopla is about but i can tell you this little bit from what i've noticed. In my immediate friends base we all have q6600 cpus, the same type of ram, and similar motherboards. My 3870 easily bests my friends 8800gts 320 in virtually everything. My other friend has a 8800gt and it seems to be slightly faster. To my knowledge the benchmarks for the hd2900xt are about similar to the 3870 so i'd have to say that a hd2900xt should give a 8800gts a huge beating.
edit- the new 8800gts 512 on the other hand is a different story of course
Look, I personally owned a 8800GTS, a 8800GT and the 3870.
Evidence that I owned 8800GTS and 8800GT:
Evidence that I owned ATI 3870:
2900XT > 3870 = 8800GTS 640 < 8800GT
8800gtx being better value than the 3870x2? Isnt the 8800gtx $450 as well? What complete crap. Most review sites have shown that the card can beat the 8800ultra in half of the games they test it on. Where the new card performs badly is in nvidia's domain i.e. Quake wars etc.
ill have the hd 3870 x2 2morrow will give u guys a better benchmark , honestly all these nvidia fanboy sites are just overly crazy in thier effort to debunk AMD/ATi ............... i mean i like both companies and its good there are 2 of them out there ..................... i have been riding with a 8800 gts g92 since it gave me better performance than the hd 3870 and now i switching back to ati , because they have something BETTER than the green right now.
anywho u guys will get a kick out of reviews on newegg. nvidia fanboys spewing false comments on there under the hd 3870 x2 , to quote one " an overclocked 8800 gt can beat this dual hd 3870 hands down " LOL ..............
Who actually respects hardocp video card reviews? Kyle on the thread for this benchmark, admitted that they do not try to do the same things for each test. Instead, they just "play the game". If that isnt easily manipulated and subject to high variance, i dont know what is.
And where did you see anyone posting that?
8800gtx accordng to hardocp
If anyone believes this has some remblance of accuracy, they are deluded. The 8800gtx is 30 fps slower at 1 second and then 30fps better the next.
Look at some of the differences at ~420secs and ~470. If the scene composition is similar in both tests, i'll eat my hand.
Eeuhm...so? I see that every day. I play America's Army on 2048x1536 and my framerate can drop from 85 to 65 in less then a second. What's so strange about that?
Separate names with a comma.