![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
POV 9800GTX Reviewed
- Thread starter lol417
- Start date
heh, not worth the money... 8800gt still is the best bang for the buck card if these benches are gonna hold till the new drivers release
everytime i look at benches its always these games that i look at
lost planet (why cant all the games be like this game which utilizes all four cores???)
crysis
ut3
WIC
And my the gtx is bad on all four of em once aa and af is on
everytime i look at benches its always these games that i look at
lost planet (why cant all the games be like this game which utilizes all four cores???)
crysis
ut3
WIC
And my the gtx is bad on all four of em once aa and af is on
behemoth85
n00b
- Joined
- Oct 29, 2007
- Messages
- 11
Not surprising based on reviews of the GX2 but still disheartening to know NVIDIA really did call this thing a GTX, ahh well, anyone know when the 10000gtx is coming 
Oh and where can I find an explanation of the points system he was using? Im assuming the reviewer was just doing some averaging for us but Id like to know.
Oh and where can I find an explanation of the points system he was using? Im assuming the reviewer was just doing some averaging for us but Id like to know.
Kaldskryke
[H]ard|Gawd
- Joined
- Aug 1, 2004
- Messages
- 1,346
I still want to see a comparison between a 8800GTS and a 9800GTX at the same clock speed, just to see if there's anything more valuable about the 9800GTX aside from the tri-sli connector.
I love when ball-washers say "I'll wait for the [H] review until I make up my mind".
Like all the other review sites lie or something. Like we all didn't know this card isn't going to be all that great. Like we all didn't know the GX2 was gonna suck before the [H] review came out.
Even if they hit well over 800 core regularly, I doubt it would be worth the price premium over G92 GTS's. ~$350 vs $200-220? Hell bloody no there's not really any question whose the better deal there. And with the drastic price drops in the last couple weeks on the GTS, I'd say they're now the bftb kings, being only $20-30 more expensive at most than the GT. It seems the only real reason to get a 9800GTX is if you MUST have tri-SLI, and well look what quad gets us over regular, why would tri-SLI be any different?
Nvidia is getting too damn content being top dog, and we're as consumers paying the price (irrespective of the asking price of this/recent cards) for it. 18 months (well 16-17 I guess) since G80 and this is what they come up with? Come on AMD get your butts in gear, if not a big leap in performance, we at least need a big drop in power draw to make such upgrades worth it.
Nvidia is getting too damn content being top dog, and we're as consumers paying the price (irrespective of the asking price of this/recent cards) for it. 18 months (well 16-17 I guess) since G80 and this is what they come up with? Come on AMD get your butts in gear, if not a big leap in performance, we at least need a big drop in power draw to make such upgrades worth it.
You obviously know nothing about what a review/evaluation methodology is, otherwise you wouldn't write what you did.
[H] and a couple other sites (bit-tech for example), use real-world gameplay to test graphics cards. Real-World gameplay reflects essentially the experience each of us will have, if we buy the card.
Most other sites (including the one the OP linked to) use built-in benchmarks, cut scenes, etc, that do NOT reflect in any way, the gameplay experience a given graphics card will provide.
So yes, other sites "lie" in a way, since pre-defined situations, such as the ones present in cutscenes or built-in benchmarks do NOT represent the true performance a given card will provide. It may blow everything away in these types of benchmarks, but provide a stuttering experience during gaming sessions. Real-World gameplay numbers are much more accurate, so trusting [H] (or others that follow this methodology) is a safe bet to influence purchasing decisions.
And the GX2 sucks ? Sure, I'm not fond of it either, since it depends greatly on SLI scaling and is expensive at this point, but a single card that provides performance equal to a 8800 GTS 512 SLI setup does not suck...
In all honesty, you don't really grasp the definition of benchmarks.
Let me first say that I consider almost every benchmark 'not accurate'. If they get 60fps at whatever settings in whatever game with whatever card, that isn't exactly going to mean i'll have exactly that amount of fps too when i buy the graphics card. Performance is dependent on the entire system. It's even dependent on how clean you keep your windows installation. I crumble in agony when I boot up someone else's pc and it has to load for 5 minutes to fill the bottom tray with 50 icons.
What benchmarks are for, is COMPARING hardware to each other. How much better is gfx card A than card B? Take this 9800GTX review as an example: what card in the list would NOT give you an acceptable gameplay? Noone, and I really mean NOONE, is going to alt+f4 the game when it doesn't run steady. You hit the escape button and lower the graphics settings, and I myself usually do that with things like physics and shadow detail. You probably won't even notice a difference at all. Neither do you see AA really. If you are focusing on the game itself, you won't ever notice whether you've got 2xAA or 8xAA on. I certainly don't at least, which is quite a blessing for my bank account.
When people buy hardware, they want to get the best price/performance most of the time. That's what benchmarks are for, and that's why I check them all out. The HardOCP reviews compare the cards against each other at certain 'settings'. I honestly have to say it's quite useless to me, because I know I would most likely not choose those settings. I don't buy a graphics card with in the back of my head knowing exactly which settings I'll be playing. I usually go the route of 'everything high and shadows low', which is what [H] doesn't anyway. Maybe I'll have a slowass old hdd and a slower CPU and won't be able to run the settings [H] recommends anyway. And you know what? I can't compare the performance of cards against each other at all in a [H] review. They test like 3-4 different cards and use different settings on all of them. How the hell am I going to know that card A is 20% faster than card B but actually costs 50% more? These are the questions you need to answer yourself when you go buy hardware, how much are you willing to spend more for how much performance, and do you think you will need the performance? I cannot answer any of this at all with a [H] review, but it does give me a little insight into how far the card can go in modern games. But truth be told, if traditional benchmarks like tomshardware charts weren't around, [H]'s benchmarking method would be useless. It just doesn't give you an idea in % how big performance differences are. The only reason you find them quite comforting to read, is because you get to know somewhere else that a 9600GT is, on average, 13% slower than an 8800GT. Whether that means that you have to put the Texture detail from High to Medium in game X, doesn't matter a damn in my opinion. The only thing I'm interested in is getting as much 'bang for the buck' as possible in a mid-highend segment of the available options.
And yeah, I do read HardOCP reviews too, and I value the (little) information I get from it. The forum gives me a lot more information though. ^^
firas
2[H]4U
- Joined
- Oct 29, 2006
- Messages
- 2,458
wow really wow
and I was thinking of dropping the GX2 for the GTX
well now I can say....GX2 FTW
and I was thinking of dropping the GX2 for the GTX
well now I can say....GX2 FTW
NKDietrich
Supreme [H]ardness
- Joined
- Jul 25, 2004
- Messages
- 5,442
I want to see some benchmarks with the 20%+ OC this thing is capable of. [H] please take note
Agreed. In the reviews I have seen the benchmarks for the comparison boards (8800 GT OC,
9800 GX2, etc.) are not close to the benchmarks in reviews that [H] has put out.
I can't get a clear picture of how the 9800 GTX compares because the numbers are all over the place from site to site.
snaggletooth
Supreme [H]ardness
- Joined
- Aug 16, 2005
- Messages
- 4,767
in all honesty, you're all over the place with your three paragraph synopsis of why silus doesn't grasp the definition of benchmarks and you never really explain why you think that is. it's hard to read exactly what you're trying to say. i took your essay as "i don't like to read."
what you do outline, in detail within your final paragraph, is your complete lack of understanding with regards to how [H] evaluates game-play experiences using new hardware.
i really can't understand why you quoted silus in order to tell the world that, but there's alot of other things i don't understand in life too.
ball-washers.
Well was gonna have some meatballs for dinner, but that aint happening now
bigdogchris
Fully [H]
- Joined
- Feb 19, 2008
- Messages
- 18,737
I only care about how much better the 9800gtx sli is compared to 8800gts SLI with the 9 series SLI improvements in mind. I have a step up waiting to happen to 9800gts if it's worth it.
If Lewis Black has his own personal ball washer, then he must be doing something right.
The 9800GTX in this review is very depressing. I guess its true colors will come out in tri-SLI. I just pulled the trigger on the 8800GTX from NewEgg and if the 9800GTX knocks some socks, I can step-up. I just couldn't pass on the $290 price from NewEgg - such a great deal.
The 9800GTX in this review is very depressing. I guess its true colors will come out in tri-SLI. I just pulled the trigger on the 8800GTX from NewEgg and if the 9800GTX knocks some socks, I can step-up. I just couldn't pass on the $290 price from NewEgg - such a great deal.
You're not much better than the poster I quoted, since you also do not understand what a review/evaluation methodology is and why is one better than the other. I explained it in my previous post, but I'll explain it again, so that you can "grasp" it and I'll use some of your words to do it.
This is obviously true, but in ANY benchmarking methodology, the bottleneck must be removed from anything other than the hardware being reviewed. And I never said you will get the same fps as in the reviews. I said real-world gameplay numbers are much more accurate than "canned" benchmarks.
This makes no sense, especially in an enthusiast forum. No AA vs even the lowly 2xAA ? Of course you see the difference. Obviously if you have a tight budget, you won't be able to buy the most powerful video card on the planet, but today, a mid-range card allows you to use at least this level of AA and in some games, even more.
Again, you don't understand the premise of real-world gameplay. What [H] does is show us what are the best playable settings with a given graphics card, when nothing else is the bottleneck. You don't need to use the exact same settings. That's NOT the point. You still have to make your own conclusions based on the "gameplay experience" text, that is always attached to each review. That is what is important. And there is nothing subjective between a stuttering experience and a smooth one. That is something everyone identifies as the same.
But the % difference in charts in tomshardware do not reflect the true difference...Seems you are comfortable in basing your opinion on numbers that are not close to reality.. And for someone worried about the best bang for your buck (which is a valid concern), you are contradicting yourself, since you are definitely NOT buying the best for your money, when basing your purchase decision on just numbers taken from a predefined set of actions, found in built-in benchmarks. But it's your money in the end...
Yet another reason to actually READ the review and especially the "gameplay experience" text. Again, you still do not "grasp" what [H] (and other sites with a similar methodology) offers, which is the reason why you consider they don't give you much useful information.
I would like to hear the official nvda response to the obvious consensus that they have been milking the same design for over 18 months. The fact you can buy a 9800 GX2 ANYWHERE and at less than the $600 launch price just a few days ago is very telling. Now the "GTX" is more of the same if not worse. Where's the road map that says technology is advancing? The last year has been nothing but marketing BS and process refinement.
My $0.02...
It's interesting to see which setups produced better frames than others, but tbreak.com gave no comparison of at what frames/second rate that each game looked too choppy to play. And the word "looked" is the most appropriate because they never really "played" the game, right?
Something else that's missing is a "minimum" or "recommended" number of frames/second on each game to compare against (also assuming you use their graphics/sound settings). Getting 120 frames/second on a game is nice to know, and showing it's frames/second against other hardware setups is useful. Where tbreak's article is lacking is lsiting what frames/second that each game "looked" unplayable and "looked" like pure overkill. Taking their frames/second result when compared with actual user comments of game responsiveness of the different hardware setups is better. Factoring the current pricing of hardware against these two results provides even more depth.
Informative article, yes. But it is insufficient by itself when analyzing the bigger picture of a real-life person playing the game on the hardware setup. And also factoring the money spent by the real-life person for the piece of hardware.
It's interesting to see which setups produced better frames than others, but tbreak.com gave no comparison of at what frames/second rate that each game looked too choppy to play. And the word "looked" is the most appropriate because they never really "played" the game, right?
Something else that's missing is a "minimum" or "recommended" number of frames/second on each game to compare against (also assuming you use their graphics/sound settings). Getting 120 frames/second on a game is nice to know, and showing it's frames/second against other hardware setups is useful. Where tbreak's article is lacking is lsiting what frames/second that each game "looked" unplayable and "looked" like pure overkill. Taking their frames/second result when compared with actual user comments of game responsiveness of the different hardware setups is better. Factoring the current pricing of hardware against these two results provides even more depth.
Informative article, yes. But it is insufficient by itself when analyzing the bigger picture of a real-life person playing the game on the hardware setup. And also factoring the money spent by the real-life person for the piece of hardware.
SLI with only 2 cards is very consistent in performance. If a 8800GTS is like 5% slower, the 8800GTS SLI will also be 5% slower than the 9800GTX SLI. I don't expect anything far off.
Unfortunately it's the price to pay for no real competition. Any money making company with a successful product in their hands, will try to make the most of it (in terms of profits that is), without increasing their costs.
As you can see, in the market AMD/ATI is actually competing (the mid-range), we've been having very good cards, with excellent prices. But in the high-end, NVIDIA has no reason to do much more than we they are doing. It's unfortunate for us no doubt, but that's just how things work...
I'll stop hijacking this thread, somehow went on a rant.
Personally I was really looking at the 9800GTX performance compared to the 8800GTX, and as I see it there's no big difference at all. In some 'heavy' cases you even see the 8800 take lead over the 9800 because of the difference in memory (bandwidth).
Sounds like a good moment for anyone to pick up a cheap 8800GTX somewhere.
Personally I was really looking at the 9800GTX performance compared to the 8800GTX, and as I see it there's no big difference at all. In some 'heavy' cases you even see the 8800 take lead over the 9800 because of the difference in memory (bandwidth).
Sounds like a good moment for anyone to pick up a cheap 8800GTX somewhere.
Simpson5774
2[H]4U
- Joined
- Feb 1, 2005
- Messages
- 2,937
Heres hoping for NV an april fools out of their asses and say "gotcha" and bring out the 'real' card.
I feel very little enthusiasm for the discrete video card market atm, but the $/performance ratio is certainly improving. I foresee the GX2 dropping a bit in price pretty soon however, I think its a little high when compared to the 3870x2.
I feel very little enthusiasm for the discrete video card market atm, but the $/performance ratio is certainly improving. I foresee the GX2 dropping a bit in price pretty soon however, I think its a little high when compared to the 3870x2.
Hmm the 8800Ultra is the clear winner over the 9800GTX in this review (especially with FSAA and AF turned on, which really matters anyway).
I find it strange that the Ultra is like 20% faster than the 8800GTX in many cases because the Ultra has only 6% faster core and maybe 15-17% faster shaders. The memory is like 20% faster, but the 8800GTX already has much faster memory than 9800GTX.
Guys, there is something out right now that is much better than the 9800GTX--and it can be bought right now off Newegg. It's a G92 GTS but with 1GB of memory at 2.1GHz and 730MHz core.
http://www.newegg.com/Product/Product.aspx?Item=N82E16814261006
http://www.palit.biz/main/vgapro.php?id=517
They're limited edition--don't be surprised if it becomes unavailable for a while during the 9800GTX launch. It would be nice to see this same benchmark review include the 1GB version of the same G92 chip at 730 MHz and see how it compares against the Ultra.
I find it strange that the Ultra is like 20% faster than the 8800GTX in many cases because the Ultra has only 6% faster core and maybe 15-17% faster shaders. The memory is like 20% faster, but the 8800GTX already has much faster memory than 9800GTX.
Guys, there is something out right now that is much better than the 9800GTX--and it can be bought right now off Newegg. It's a G92 GTS but with 1GB of memory at 2.1GHz and 730MHz core.
http://www.newegg.com/Product/Product.aspx?Item=N82E16814261006
http://www.palit.biz/main/vgapro.php?id=517
They're limited edition--don't be surprised if it becomes unavailable for a while during the 9800GTX launch. It would be nice to see this same benchmark review include the 1GB version of the same G92 chip at 730 MHz and see how it compares against the Ultra.
+1
With basically ALL video card releases since the 8800GT, the same applies: "why bother" just a waste of time. With the one exception, maybe, being the 9600GT...
I think it would have been better NOT to release ANY of the 9800 series cards yet, Nvidia is loosing big respect points in my book. Pfft...
Attack of the Kyle Bennett clones. This is what they sound like when he tells them to "execute order 66".
NKDietrich
Supreme [H]ardness
- Joined
- Jul 25, 2004
- Messages
- 5,442
Palit is good stuff. Not the biggest name but they sometimes have features other cards do not have. Adding additional memory, outputs, and usually better-than-reference cooling.