At least the Parhelia is extremely useful for some people (who would like to run 3 monitors off 1 card, get great image quality and not for gaming).
Originally posted by: merlocka
Wow.
I am quite suprised that such a number of people are so dissapointed with the GeforceFX.
1) It's equal to a product which had the most dramatic performance improvement in a generation since Voodoo2 SLI vs Voodoo Graphics. I'm impressed that nVidia was even able to match the R300.
3) I'm still in shock that so many people are taking the cooling seriously. The jokes and photochops are hysterical, but do that many people really think that the majority of AIB makers will want to sell cards which are 70dB? We know they have the expertise to create performance cooling which is not as loud as a jet engine (ala Abit OATES), I find the whole ordeal to be silly.
But the fact is that RIGHT NOW, FX is so loud that it is REALLY turning potential buyers elsewhere! It's just too loud, period.
Originally posted by: Ilmater
[First off, I'm not just picking on you. There are a lot of people that, like you, are saying this same thing. However, let's look at what was hyped by each company.
nVidia: Same nVidia rhetoric - State of the art graphics, DX9+ compatibility. nVidia never said that they would blow ATI out of the water, and I challenge everyone to prove otherwise. It IS the fastest card out currently. Period. Though not by much.
Matrox: said it would revolutionize gaming and be extremely fast, neither of which it did.
It hyped its bump-mapping and triple-head gaming, and nobody cared.
By the time you turn on all the features that Matrox hyped, you're getting 20fps.
Yes, that Matrox card is useful to some people, but only a very select few. At least the gaming market will open up to nVidia's card.
No gamer is going to buy the Parhelia. And just to make it clear, the Parhelia WAS intended for the gaming market. It's clearly the bigger flop.
Originally posted by: Evan Lieb
But the fact is that RIGHT NOW, FX is so loud that it is REALLY turning potential buyers elsewhere! It's just too loud, period.
No, you just haven't heard an actual (i.e. shipping) GeForceFX card before.![]()
Originally posted by: Nemesis77
Originally posted by: Evan Lieb
But the fact is that RIGHT NOW, FX is so loud that it is REALLY turning potential buyers elsewhere! It's just too loud, period.
No, you just haven't heard an actual (i.e. shipping) GeForceFX card before.![]()
Assuming 9700 is about as loud as GF2 GTS, then FX is too loud. True, I haven't hear a shipping FX yet, and I hope that it's more silent. But _right now_ FX is too loud. If NV can cut down the noise, then I might consider it. But right now, I wouldn't consider it.
Originally posted by: merlocka
LOL. You miss the point. You haven't seen or heard a "production quality" GeforceFX so you don't know if its cooling system produces 7,70,or 170dBa of noise.
Assuming 9700 is about as loud as GF2 GTS, then FX is too loud. True, I haven't hear a shipping FX yet, and I hope that it's more silent. But _right now_ FX is too loud. If NV can cut down the noise, then I might consider it. But right now, I wouldn't consider it.
Originally posted by: Nemesis77
Originally posted by: merlocka
LOL. You miss the point. You haven't seen or heard a "production quality" GeforceFX so you don't know if its cooling system produces 7,70,or 170dBa of noise.
I will put my entire post here, with highlights where needed.
Assuming 9700 is about as loud as GF2 GTS, then FX is too loud. True, I haven't hear a shipping FX yet, and I hope that it's more silent. But _right now_ FX is too loud. If NV can cut down the noise, then I might consider it. But right now, I wouldn't consider it.
I was talking about how RIGHT NOW FX is way too loud. My post is pretty clear on that. And unless they drop the noise, I will not consider it.
Originally posted by: Dulanic
I think your missing Evan's message there... Evan is pretty good at giving hints but not actually saying things![]()
You keep saying RIGHT NOW its too loud... RIGHT NOW you cant buy one
Yes the card nvidia made is too loud... but man... people are WAY overeacting, I could care less how loud a card I CANT BUY will sound
But I'm also not gonna go crazy over a cooling solution that I wont even be on retail cards (except lazy manufacturers).
Originally posted by: Evan Lieb
Why is everyone so surprised? Anand and the rest of the Internet reviewed a non-shipping card from NVIDIA; i.e. not a shipping retail product.
Since I'll be doing individual video card reviews soon, I called up MSI, PNY and Leadtek to see what their FX schedule would be. All of them said FX availability next month, though not in massive quantities...however what was interesting was that they all purposefully claimed that their cards would be much quieter than the competition, especially the FX cards used in reviews over the Internet.
Though 7db is a little hard to believe, I'm sure most FX cards aren't going to be any louder than 9700 Pro cards.
Regarding the noise non-issue, maybe this will help clear up any "doubts":
Originally posted by: Evan Lieb
Why is everyone so surprised? Anand and the rest of the Internet reviewed a non-shipping card from NVIDIA; i.e. not a shipping retail product.
Since I'll be doing individual video card reviews soon, I called up MSI, PNY and Leadtek to see what their FX schedule would be. All of them said FX availability next month, though not in massive quantities...however what was interesting was that they all purposefully claimed that their cards would be much quieter than the competition, especially the FX cards used in reviews over the Internet.
Though 7db is a little hard to believe, I'm sure most FX cards aren't going to be any louder than 9700 Pro cards.
Yup, looks like early summer 2003 if it's .15-micron or sometime after that if it's .13-micron. However, if it's just a core clock speed increase (as we suspected), then I can't see it defeating the GeForceFX, though it'll be in the ballpark probably.
The Inq gets it dead wrong here. I have great respect for Magee, but this piece is just FUD. Wait for the reviews (coming very soon), and then you can see whether you believe NVIDIA is going the way of 3DFX.
Originally posted by: Pariah
Sorry, but I would absolutely not take his word as gospel.
But instead, you're willing to take the previews of a pre-release reference board as "gospel"? nVidia has said throughout OEMs had full reign over cooling design, yet despite reports to the contrary in a press release by a well-known AIB maker AND someone who works in the industry you still choose to base your opinion on a reference board? That makes sense.![]()
Considering the r300 is already "in the ball park" the r350 would litterally have to be slower than the r300 to be considered slower than the FX. Anyone want to predict that the r350 will be slower than the r300? Who here thinks the r350 even with just a clock boost (which hasn't been confirmed) won't be able to beat the FX in the majority of benchmarks now?
I'm not sure what you're expecting out of R350, but the latest roadmap does indicate it will be nothing more than a clock/memory frequency boost. No die revisions and no enhancements to the architecture were mentioned. The first .13 micron chip is expected to be RV350. If you want to see how a clock/memory speed increase will improve performance, you need to look no further than Tyan's Tachyon, which provides about a 17% boost over a stock Radeon. How well the R350 will scale or overclock is anyone's guess, but considering it will still be on a .15 micron process, headroom may be limited.
OK, I've seen the reviews, and I'm thinking 3dfx more than ever, how did the reviews disprove anything? The FX reminds me of the V5 5500 release more than any other card I can think of, huge, hot, late and only on par performance wise with the competition when it was expected to be so much more. NVidia is not going out of business anytime soon, but if the FX release doesn't bring back memories of the V5 release then you haven't been following the industry.
No, you've seen the PREviews. How do the initial previews of the 9700pro and 9500pro stack up to their release performance? Quite a performance delta IIRC. That piece in the Inquirer was FUD and much of that rhetoric has spilled over here. Yes, its huge, hot, and late, but thats a by-product of innovation made possible by being the industry leader. Considering nVidia was late to market and essentially missed a product cycle, they must be tickled-pink to see they maintained their industry market share. Its easy to blame nVidia for the late launch of the FX, but if you had been following the industry you would remember that NV30's specs were finalized last year. Complications in shifting to a .13 micron process at TMSC held up the FX's production; but I guess acknowledging that depends on what side of the industry you follow.
Chiz
But instead, you're willing to take the previews of a pre-release reference board as "gospel"? nVidia has said throughout OEMs had full reign over cooling design, yet despite reports to the contrary in a press release by a well-known AIB maker AND someone who works in the industry you still choose to base your opinion on a reference board?
I'm not sure what you're expecting out of R350, but the latest roadmap does indicate it will be nothing more than a clock/memory frequency boost.
How well the R350 will scale or overclock is anyone's guess, but considering it will still be on a .15 micron process, headroom may be limited.
No, you've seen the PREviews. How do the initial previews of the 9700pro and 9500pro stack up to their release performance?
Originally posted by: Pariah
I believe very few prerelease boasts from companies as they are inaccurate far more often than they are accurate. Nvidia didn't choose the horrible cooling solution they are using to impress anyone or for their own amusement, but because that's what they felt was necessary to run their hardware properly. Sure OEM's are free to use any cooling solution they want to, if they want to put a dinky chipset passive heatsink on the card they can, but it isn't going to run at all. We'll see what other companies come up with but to expect major improvements is wishful thinking unless Nvidia has really lost their edge.
Now you're just waffling. Is it a prerelease or has it been released? How can you question the truthfulness of their claims when every single assertion you've made about the FX is based upon assumptions? You think a company would risk their reputation and not be able to deliver? I'm not sure what you do for a living, but in the real world, people and firms are held accountable for their assertions and "boasts." Of course no one is held accountable in this safe haven called the internet, where the only stakes are a troll or a flame. You should really look into whats going on over at Zalman, they've been producing quiet passive coolers for some time that hardly resemble a "dinky chipset fan".
Evan seems to think that R350 will only achieve ballpark performance of the FX, which is kind of silly when you can see the R300 is already beating the FX in a number of benchmarks.
He said that prior to the benchmarks being released based on his knowledge at the time. Maybe he knew something we don't. Maybe he still does. :shrug;
It isn't going to take much of a clock boost to push it past the FX in the majority of benchmarks. ATi has had 6 months since the R300 release, I would hope they have achieved more than a 10MHz speed bump of the GPU by now.
The R300 has proved to be a pretty decent OC'er so there is definitely headroom left. Mine hits 380MHz standard cooling, and it may go higher than that, I just haven't tried.
The fact that a 9700pro is currently a decent OC'er is actually a bad sign for the R350, IMO. ATI could have very well been depending on that extra headroom for an R300 refresh. Fault tolerances are built into every core design, but there's going to be a point where scaling hits a wall. Without a core revision or a die shrinkage, its anyone's guess what they'll be able to squeeze out of the .15 process. There have been rumors of lower power consumption as well, but we've seen that it does not guarantee higher clock speeds (early T-bred A's), just heat dissipation.
And again, there have been reports of driver problems leading to flawed AA and AF results with the beta drivers, making it more difficult to judge the FX's performance or potential. Many reviews noted that the FX ran significantly faster with aggressive AA and AF settings than the 9700pro, but b/c the quality appeared to be worse, they tested with the quality settings. Reports from nVidia cite driver issues and frame buffer screen capture techniques. Although the impact is questionable, it does introduce a difficult paradigm involving subjective testing methods based on preference. When its all said and done, maybe the aggressive and quality settings of the FX sit on either side of the 9700pro's performance, leaving the final judgement to the end user to decide between more frames per second or image quality.
Please post links to back up this claim. There isn't going to be any huge performance increase from the tested boards to the shipping boards in 3 weeks. Every product release no matter the company these claims are made, and they never turn out to be true. Maybe 6 months from now, but not in 3 weeks.
Not sure if this is a case of selective memory or genuine oversight. The Radeon 9500pro was released just over 2 months ago, and previewed 3 months ago.
9500pro preview dated 10/24
9500pro review dated 11/27
You should be able to quickly see there is in fact a significant increase in performance due to final silicon and drivers; enough to leapfrog an entire GPU family. nVidia may very well be able to perform a similar feat considering the quality of their shipping drivers tend to improve as cited by most review sites. As for having to wait 6 months, I highly doubt it considering they release WHQL drivers almost monthly. There's often 1 or 2 beta releases between WHQL versions, so there should be improvement with each driver update and game patch (that optimizes for the FX).
You think a company would risk their reputation and not be able to deliver? I'm not sure what you do for a living, but in the real world, people and firms are held accountable for their assertions and "boasts."
You should really look into whats going on over at Zalman, they've been producing quiet passive coolers for some time that hardly resemble a "dinky chipset fan".
Fault tolerances are built into every core design, but there's going to be a point where scaling hits a wall. Without a core revision or a die shrinkage, its anyone's guess what they'll be able to squeeze out of the .15 process.
Not sure if this is a case of selective memory or genuine oversight. The Radeon 9500pro was released just over 2 months ago, and previewed 3 months ago.
Not sure what ideal world you live in, but practically every day this happens where companies don't achieve what they claimed they were going to do.
That one went right over your head, it was complete sarcasm but you missed it.
Original quote: Sure OEM's are free to use any cooling solution they want to, if they want to put a dinky chipset passive heatsink on the card they can, but it isn't going to run at all.
The difference between the XP2200+ Tbred Rev A and XP2800+Tbred rev B is proof that miracles can happen without a die shrink. Regardless, look how often Nvidia has "refreshed" their cards without major revisions and increased performance quite well.
Is this the best you could come up with? A hack job with a 9500Pro GPU mounted on a 9700Pro PCB with half its memory disabled? The GeforceFX GPU was not mounted on a TI4600 PCB using only 64MB RAM. How about a comparison where the GPU is actually on the right PCB using the same amount of memory?
The majority of the performance differences come from a new driver that is shipping with the card. The driver improves performance across the board on the 9500 Pro and also works on the 9700 Pro. Performance isn't really enhanced on the 9700 Pro however, in some cases it's faster in others it's slower.and from his FX review:
All evidence points to the GeForce FX's drivers holding it back
Maybe you can second guess Anand's conclusions as well???![]()
Chiz