Ok where the heck are the GeforceFX reviews?

[martrox]

I am one who upgraded to the 9700 from a GF4 TI4600 - of course it only cost me NOTHING.... well, $25 for the replacement warrnty, so I can do it again - and there is a huge difference IF you want excellent IQ. Even at a frame killing 4Xs FSAA, the 4600 just doesn't look very good, and lets not even talk about AF. I run my 9700 at 4X FSAA & 16X AF without even a thought. And I have had not a problem with ATI's drivers. DX9? Well, that's a bonus, as I will very soon be able to play 3DMark2003

EDIT: And I haven't had to resort to any hack/tweeker to get all the features, either.

 

[RussSchultz]

Not that it means too much, but I beleive the GF2GTS was a 'suprise'. The NDA didn't expire until (essentially) boards were available in the stores.

Or maybe it wasn't the GF2GTS, but the GF3...

So, its not unheard of, but they sure seem to be taking their sweet time, don't they?

 

[Nagorak]

To all those that said that the NV30 was just around the corner when the R9700 was launched and managed to hold off until now:

The R400 is just around the corner, and the R350 is even closer. Why don't you just wait a few more months?





With regards and surprisingly inobvious irony

Kjetil

Well waiting is always best. I am in need of a new graphics card because my aging 8500 is no longer cutting it. Still, I plan to hold out for another month or two to allow prices to drop more. Of course there aren't any major games I'm interest coming out for at least another couple months.

I really have no interest in the GFX at all because I know it will be overpriced since it is new. The only reason I want to see it come out is so the R9700 gets cheaper!

 

[Randell]

Not that it means too much, but I beleive the GF2GTS was a 'suprise'. The NDA didn't expire until (essentially) boards were available in the stores.

Or maybe it wasn't the GF2GTS, but the GF3...

So, its not unheard of, but they sure seem to be taking their sweet time, don't they?

I believe it was the GTS

 

[Joe DeFuria]

Russ,

You are correct about the GTS. (Reviews started appearing only at the point where shipping was imminent.)

However, that situation was very different: nVidia already had the fastest card available on the shelf (GeFore DDR) when the GTS NDA lifted. GTS reviews started appearing (probably not coincidentally) just before Voodoo5 boards hit the shelves. It was not in nVidia's advantage to lift their NDA any sooner...even if the GTS was ready earlier.

In today's case. it's obviously in nVidia's best interest to get benchmarks out ASAP to combat the mindshare that ATI has. This is assuming of course that at least a good case can be present showing the GeForceFX to be the "superior" product.

What does this all mean?

I see 1 of 3 things are happening:

1) GeForceFX can be clearly shown to be the superior product, though it's just taking this long to finalize PCB design and drivers for a reviewable product. We will see reviews pop up on day, and shipping start 2 weeks to a month later, as per usual.

2) In a "preview" state, GeForceFX can not be "clearly" shown to be the superior product, but in a "final" state, it might. So nVidia is waiting as long as possible to give driver writers a chance to wring out as much performance as possible. That means, the first "hands on p/review" cards would actually be with extremely close to shipping, if not the shipping, drivers. nVidia would not want to have "disappointing previews" of cards knowing that the shipping product will have some significant speed boost from drivers that are still in development. Reviews and actual shipping will be very close together.

3) In a "final" state, GeForceFX can not be clearly shown to be the superior product. We may not actually see the NDA lifted until after the first shipments begin, ensuring that the p/reviewed product is the shipping product. This happened with the GeForceSDR launch, IIRC. Some markets in the far east saw products on the shelf before benchmark NDAs were lifted.)

 

[Randell]

Massive speculation mode on

Maybe pricing has something to do with it, i.e the GeforceFx is the better product - but not to the degree nVidia can be sure the reviews will say it is worth the price premium. If the Fx was say $50-$75 more than the 9700Pro - then the nVidia name, driver rep and strong consumer loyalty will accept that as a fair premium for a circa 20-30% performance advantage in UT2003 as opposed to $150/£150 being touted at the moment.

Maybe..

 

[LeStoffer]

it's obviously in nVidia's best interest to get benchmarks out ASAP to combat the mindshare that ATI has. This is assuming of course that at least a good case can be present showing the GeForceFX to be the "superior" product.

Given the high corespeed it shouldn't be a problem a find a couple of those cases.

I just think that they have realized that 3 or 4 weeks more doesn't make or break the GeForceFX given the already big delay. If you release benchmarks prematurely with too low numbers you just risk that the people that have waited for several months will finally give in to ATI. Nah, better hold them until February and give them the whole polished picture.

 

[Sabastian]

Given the high corespeed it shouldn't be a problem a find a couple of those cases.

I just think that they have realized that 3 or 4 weeks more doesn't make or break the GeForceFX given the already big delay. If you release benchmarks prematurely with too low numbers you just risk that the people that have waited for several months will finally give in to ATI. Nah, better hold them until February and give them the whole polished picture.

I have always concluded Nvidias difficulty with the Radeon 9700 pro vs GeforceFX is that I don't believe that it will be so much faster that everyone is going to drop their Radeon 9700 pro and buy the GeforceFX. Essentially what I am hypothesizing is that for every Radeon 9700 pro sold is one less GeforceFX that will be sold. Sure there will be the few that will, in that case where the GeforceFX outperforms to the slightest degree, upgrade from the Radeon 9700 pro but I am willing to bet the performance disparities will not be enough to make the difference. That is the real problem for nvidia IMHO. That is why they need to have reviews that show the cards superiority ASAP and [sarcasim] also why it needs to be on the market before Christmas.[/sarcasm] Its why they paper launched the product in Nov.(Never mind it won Wired magazines Vaporware Award.) Really I just want to see some god dammed performance numbers and reviews.

EDIT: Hope thats better.

 

[martrox]

That is why they need to have reviews that show the cards superiority ASAP and also why it needs to be on the market before Christmas. Its why they paper launched the product in Nov.(Never mind it won Wired magazines Vaporware Award.) Really I just want to see some god dammed performance numbers and reviews.

shouldn't that read "why it needed to be on the market before Christmas"

 

[Joe DeFuria]

Given the high corespeed it shouldn't be a problem a find a couple of those cases.

No doubt...but that's why I specifically why I said "clearly" outperform.

In other words, nVidia knows pretty much what benchmarks / resolutions / settings are generally being run today. So if nVidia is confident that "overall", running those benchmarks against the 9700 Pro will show a picture clearly favoring the NV30, nVidia would want to have that published ASAP. Finding a few specific cases where NV30 can be clearly demonstrated to be superior isn't enough, if there are other cases where the R-300 is clearly superior. Especially if the cases where R-300 is superior are seen as more important.

For example, (completely made up numbers): NV30 might be clearly better at no AA, no Aniso 1600x1200x32 setting. Running 200 FPS vs. 140 FPS of R-300.

On the other hand R-300 might be better with 4X AA and 8X aniso turned on in the same app. Maybe R-300 gets 70 FPS, and NV30 gets 50.

Of course, what we're likely to see happen, as always, nVidia and its supporters will claim that the areas that NV30 excels at, are the "most important", and ATI and its supporters will claim that whatever R-300 excels at is "most important."

Case for NV30: As games get more and more fill rate demanding, NV30 will be better equipped to handle it in the future.

Case for R-300: You'll get better performance right now with R-300 with all the quality cranked up.

This is where we'll start to see the review sites "take sides...."

 

[Kaizer]

Russ,
2) In a "preview" state, GeForceFX can not be "clearly" shown to be the superior product, but in a "final" state, it might. So nVidia is waiting as long as possible to give driver writers a chance to wring out as much performance as possible. That means, the first "hands on p/review" cards would actually be with extremely close to shipping, if not the shipping, drivers. nVidia would not want to have "disappointing previews" of cards knowing that the shipping product will have some significant speed boost from drivers that are still in development. Reviews and actual shipping will be very close together.

This might also mean that nVidia won't have the usual "Detonator Performance Boost Half A Year After Card Release" (TM) up it's sleeve this time. Since nobody (atleast not ATi) seems to have to catch up to nVidia's performance levels this coming fall (ATi will probably just let thier R400 walk all over the GFFX), nVidia will probably want to max their performance as quickly as possible..

Great for ATi fanboys, not so great for nVindians.


With regards and a slight hope of confusion
Kjetil

(If anybody plans to call me a ATi fanboy, you'd might like to know that I am sticking to my R100 AIW as long as I must, since I am no longer bored enough to play games, and therefore don't need no performance or features)

 

[LeStoffer]

I have always concluded Nvidias difficulty with the Radeon 9700 pro vs GeforceFX is that I don't believe that it will be so much faster that everyone is going to drop their Radeon 9700 pro and buy the GeforceFX.

Hmmm, I don't think that nVidia ever had much hope of people dropping their Radeon 9700 Pro for a GeForceFX. They knew from the very start that the performance difference in most of todays games can't justify a shift but if they could have had the product out much earlier - e.g. October/November - they could have grap some of those people that bought a Radeon 9700 Pro all those months up til Christmas.

Well, a battle was lost for nVidia, but not the war they kinda think.

 

[Sabastian]

Hmmm, I don't think that nVidia ever had much hope of people dropping their Radeon 9700 Pro for a GeForceFX. They knew from the very start that the performance difference in most of todays games can't justify a shift but if they could have had the product out much earlier - e.g. October/November - they could have grap some of those people that bought a Radeon 9700 Pro all those months up til Christmas.

Well, a battle was lost for nVidia, but not the war they kinda think.

Hrm I do think that the problem is that the NV30 won't be contributing much to nvidias bottom line then. Consider the "enthusiast" market accounting for nearly 10% of nvidias revenues this could only equate to a considerably smaller percent in terms of the over all market. Really the ultra high end market is quite finite. Basically in terms of sales the GeforceFX card *should* be a total flop comparitively speaking. In terms of performance well that is another matter and it is likely to give the Radeon 9700 a fat lip. (On a number of benches.) I am just simply tired of waiting for the card .... totally. :?

In terms of the overall market (or war if you like) ATi is very much the underdog yet. If they can hold the high end with the R350 they ought to be able to manage some real brand name recognition and market share.

 

[antlers]

9700 NDA's (for complete reviews) expired just before limited retail availability, as I recall (there were some benchmarkless previews before then). So, probably, NVidia is just timing NDA expiration with limited retail availability, which still might be the end of January--with broader retail availability a month or two later.

Also, the 9700 experience has shown what a valuable asset is NVidia's reputation for driver quality. The NV30 will require the biggest change to their drivers since the GeForce 3 arrived, and I imagine that they would be doing their best to preserve the reputation of their drivers even in the face of this challenge.

Speculation Follows! A scenario much less likely than the one above!

Of course, while we're speculating, I might suggest the "GeForce FX is a total dog" scenario. In this scenario, they can't get it to run bug free at Ultra speeds on at least some common applications in any quantity at all (or OEM's refuse to make Ultra's in quantity because of the price point and worries about compatibility). The only FX that could make it to the shelves in quantity is the 400/900 Mhz version, which can't be sold profitably for <$400, even though it is beaten by the 9700 in some common benchmarks.

In this scenario, NVidia holds back on making any NV30 devices publically available until the follow-on NV3x mainstream products are ready. The NV30 becomes a nearly-totally-vapor product that nVidia forgets about as quickly as possible--a few heavily subsidized cards do show up at retail and at review sites to build the GeForce FX hype for the mainstream cards.

 

[jjayb]

In terms of the overall market (or war if you like) ATi is very much the underdog yet. If they can hold the high end with the R350 they ought to be able to manage some real brand name recognition and market share.

Ati is still the underdog but they are making great strides with the 9700. I've seen quite a few people who swore off Ati forever actually give in and buy the 9700 due to the gforcefx being so late. Many people's minds have been changed about ATI due to the 9700. I can't count the number of people who warily bought the 9700 worried about the drivers because of all they heard about on the net. Only to install it and have it work like a champ. Alot of Ati's "driver problems" have been propagated by people who have never even used an Ati board who only pass along what they hear on the net. The gforcefx's delay has caused many people to buy 9700's who would have originally bought a gforcefx had it not taken so long to come out. Those are just more people to dispell the myth about ATI's "crappy drivers". Aside from the die hard Nvidia supporters who will never admit when Ati does something right. (you know who you are) The Nvidia delays have really, really helped Ati's image in many ways. Every day the nv30 is delayed is another day a former Nvidia user will buy a 9700 and see that Ati really does make a fine card. They will no longer be wary of buying something other than an Nvidia card. This will also help future sales of Ati cards. Nvidia really needs to get some reviews out if the card is truly superior to the 9700. Every happy 9700 owner is a potential r400 owner and this can also hurt nv40 sales.

 

[Bambers]

I agree with that, looking at forums like madonion, (err futuremark even) etc people are getting fed up with the wait. Before xmas most seemed happy to wait but with the big lack of reviews or even previews around in the new year and the falling price of the 9700s with the high expected price of FX quite a few have got bored waiting. Especially when a couple of people with rev 3 chips say their pros go to 400 core out of the box...

 

[sas_simon]

I am one of those that have been waiting and waiting and waiting and waiting and waiting....well anyway, I am getting a little sick of waiting and hearing nothing about the card. I would of expected reviews by now.

If ati can bring out the r350 at a good price at the same time as the geforcefx ultra, I will consider buying it.


A lot of people are getting sick and tired of waiting as well and all they will do is just give up and buy a 9700pro or a 9500/pro which are just over £100 now.



But isn't there all this talk about the geforcefx 5800 and ultra version being released at the end of january? If this was correct, nvidia must be waiting till the last minute to actually allow the media to review the cards.

 

[Doomtrooper]

I would like to know what people would wait 6 months for that would make it that much better, especially the Madonion crowd who seems to have endless pockets for water cooling, liquid nitrogen...

These people could easily turn around and sell their 9700 if the FX is that much better.

Reminds me so much of the original Athlon and P3 wars, the Athlon outperforming the P3 by a VERY good margin, cheaper yet hardly anyone bought them...the old saying of 'if you build it they will come' even if the product is superior should be changed to 'if we hype it, they will wait till we get it right'

 

[T2k]

I still believe that GFFX cannot beat R9700Pro in any terms w/ 4x AA + 8x AF - means, I don't see the points why will anybody prefer that card for approx. 2x more $$...

 

[Sabastian]

I still believe that GFFX cannot beat R9700Pro in any terms w/ 4x AA + 8x AF - means, I don't see the points why will anybody prefer that card for approx. 2x more $$...

I guess that is my point. We don't yet know why because there aren’t any reviews. Yet the actual card launch was nearly 1 1/2 months ago. As for the Radeon 9700 being a better card we still don't know and have to pass the time more to find out. Nvidia would be hyping it up right now if they could. Supposedly the card is going to be for sale on Jan 22 for $700. But there is nothing not even a half assed preview. We ought to be seeing something err at least I thought we would have by now.