Impact of a "Paper" NV30 launch?

[Joe Cool]

Hopefully Nvidia will be able to release their NV30 product in volume when they do so, that would be a good thing for compeititon's sake no matter whose "side" you're on.

But let's say it comes out and is only available in limited quantities for a substantial time. That is, whereas reviewers and a few lucky people at conventions can get it, the vast majority cannot - because it won't be available on most store shelves. Making it perhaps even a "paper launch."

If you're a big Nvidia fan and have been waiting for the NV30 (for whatever reason), would this impact on your attitude towards them? Do you think the industry's attitude towards them would change? And deserve to change?

The reason I bring this up is that I've noticed a bit of a cooling amongst certain hardware sites towards AMD lately for their own paper launches of their top end CPU's. This even though the currently available AMD products are still often very competitive, at least on a price/performance ratio. This sometimes seems to be forgotten amongst the grumbles.

You see, Nvidia, ironically, might be in that boat too if the NV30 is a paper launch. There's nobody that can compete with them on a price/performance ration when it comes to their GF4 Ti4200 cards, and that's not likely to change either on Monday, however the NV30 launch is carried off.

(The Radeon 9500 is an unreleased and unproven product itself, so I'm discounting it)

But back to my main point - especially if if you're a fan of Nvidia - would they be worthy of criticism if their NV30 is a paper one? Or now that companies like AMD are doing it more and more - is this just something to be expected?

[Nutty]

I think these paper launches and delays are just a by-product of pushing the envelope. The faster you try and make the hardware the more mistakes happens, problems arise, low yields, expenses etc...

My opinion of nVidia hasn't changed much cos of the delays of nv30.

[legion88]

Originally posted by Joe Cool
Hopefully Nvidia will be able to release their NV30 product in volume when they do so, that would be a good thing for compeititon's sake no matter whose "side" you're on.

But let's say it comes out and is only available in limited quantities for a substantial time. That is, whereas reviewers and a few lucky people at conventions can get it, the vast majority cannot - because it won't be available on most store shelves. Making it perhaps even a "paper launch."

If you're a big Nvidia fan and have been waiting for the NV30 (for whatever reason), would this impact on your attitude towards them? Do you think the industry's attitude towards them would change? And deserve to change?

The reason I bring this up is that I've noticed a bit of a cooling amongst certain hardware sites towards AMD lately for their own paper launches of their top end CPU's. This even though the currently available AMD products are still often very competitive, at least on a price/performance ratio. This sometimes seems to be forgotten amongst the grumbles.

You see, Nvidia, ironically, might be in that boat too if the NV30 is a paper launch. There's nobody that can compete with them on a price/performance ration when it comes to their GF4 Ti4200 cards, and that's not likely to change either on Monday, however the NV30 launch is carried off.

(The Radeon 9500 is an unreleased and unproven product itself, so I'm discounting it)

But back to my main point - especially if if you're a fan of Nvidia - would they be worthy of criticism if their NV30 is a paper one? Or now that companies like AMD are doing it more and more - is this just something to be expected?


A "paper launch" is the start of a large advertising campaign. Why would anyone whine and moan about an advertising campaign? It is like bitching and moaning to 3dfx for telling us what the Voodoo5's can do for nearly a year before the product actually hit the stores.

The only criticism NVIDIA should get (especially from investors) is if they are unable to supply the public a worthy product for the Christmas season after making such a promise. And would that criticism mean anything? Nope.

The product, when it finally hit the stores, will be judged on how well it stacks up against the competition, not how late it was, just like the V5 before it.

[Fotis]

Missing a product cycle isn't serious?:rolleyes: :rolleyes:

[Wenzula]

Sorry if my topic does not belong to the thread topic, but I have a question and did not want to start a extra thread for it. :)

Can somebody please tell me where and when I can watch the launch event of the GeForce FX aka NV30 live via live-stream in the internet???

THX for your help.


Greetz

Wenzula

P.S. Sorry for my bad english but it is not my native language.

[legion88]

Originally posted by Fotis
Missing a product cycle isn't serious?:rolleyes: :rolleyes:

The issue is whether or not people's attitudes towards NVIDIA would change because it was late.

We were never talking about any problems caused by reduced revenues due to missing a product cycle (i.e. 3dfx's financial problems). 3dfx missed their product cycle but there were still fans by the boatload talking about how great the V5 was.

[Uttar]

They'll be worthy of criticism if a competitor's product is released less than 3 months after the NV30 is faster.

If the NV30 is in stores before the R350 *and* faster than the R350, then it's certainly not a good idea to criticize nVidia.

However, if that isn't the case... Bad nVidia! Bad! :)
I most certainly wouldn't switch sides for so little, however. The NV30 is superior in features no matter what, so I still find it to be an interesting product.


Uttar

[Dr_Colossus]

Dead rumor mill.

[LORD-eX-Bu]

once 3dfx missed the cycle they were never the same. The card that they finally put out wasn't good and they lost alot of customers. Including me.;) :D

[thcdru2k]

let's hope that doesn't happen to nvidia :D

[StealthHawk]

Originally posted by Joe Cool
The reason I bring this up is that I've noticed a bit of a cooling amongst certain hardware sites towards AMD lately for their own paper launches of their top end CPU's. This even though the currently available AMD products are still often very competitive, at least on a price/performance ratio. This sometimes seems to be forgotten amongst the grumbles.

the reason being that AMD paper launched new high end CPUs then paper launched more high end CPUs a month later or so. at the time of the new launches, the previous CPUs were still not available in any conceivable volume.

which is utterly ridiculous.

[suburbanguy]

Erm... I think that if ATI launches a R350 it will match NV30 in features and cream it in performance (256-bit bus don't forget)
provided ATI uses GDDR-3 which is ment to be significantly faster than DDR-II that Nvidia is using.

ATI is rumored to launch R350 by the spring....

[StealthHawk]

Originally posted by suburbanguy
Erm... I think that if ATI launches a R350 it will match NV30 in features and cream it in performance (256-bit bus don't forget)
provided ATI uses GDDR-3 which is ment to be significantly faster than DDR-II that Nvidia is using.

ATI is rumored to launch R350 by the spring....

the GDDR3 being tested runs at 1GHz last we heard. that isn't "significantly faster" than the DDRII on NV30. hell, that probably won't be any faster.

[DrAkO]

It's not correct to compare NV30 to R9700 because we have a lot of month between each card is available...

And if R350 or R400 are available a couple of month after the NV30, it will be false to compare the next ATI card VS NV30...

To compare 2 cards, they will be available at the same time...and there's no situation at this moment...

[suburbanguy]

Nv30/GeForceFx is sounding less and less impressive by the day IMHO. I'll hold off until I have a rough idea of when NV35 is coming out and what performance/features it is likely to have.
I am not getting a NV30/GeForceFX based video card since it's only going to have a 128-bit bus, (and now seems likely to have less triangle performance than R300) It's likely though, that NV35/GeforceFX2 will move to a 256-bit bus.

My speculation is, it is also probably very likely at this point, in terms of performance, that NV35 will be to the NV30 what GF2GTS was to the original GeForce.

To be fair, I can say the same about R300 and the rumored R350.
R300 has 256-bit bus already BUT it only has 1 TMU per pipeline.
R350 should bring that to 2 TMUs per pipe thus 16 TMUs in total, equaling NV30. (just speculation here)

If R350 and NV35 are coming out sometime in 2003, I will get one of those. The extra performance can only help in games like DOOM3. It's usually a wise decision to buy the REFRESH of a new GPU architecture, rather than buying into the first iteration.

[thcdru2k]

??? double post ???

[tamattack]

Originally posted by StealthHawk
the GDDR3 being tested runs at 1GHz last we heard. that isn't "significantly faster" than the DDRII on NV30. hell, that probably won't be any faster.

Not that I have an opinion on GDDR3 one way or the other, but if the rumours are true, then nv30 is 128 bit.

1 GHz GDDR3 on 256 bit = 2 x (1 GHz DDR2 on 128 bit) physical bandwidth.

[Bigus Dickus]

Paper launches are not good. This has always been the case, be it ATi, Intel, AMD, Via, or nVidia. Try and sugar-coat it any way you wish, but it is still an indication that the company knows their product is late.

[StealthHawk]

Originally posted by tamattack
Not that I have an opinion on GDDR3 one way or the other, but if the rumours are true, then nv30 is 128 bit.

1 GHz GDDR3 on 256 bit = 2 x (1 GHz DDR2 on 128 bit) physical bandwidth.

and NV30 can't have a 256bit bus with 1+GHz RAM?

and i still say it's premature to be judging NV30 without having seen anything about it yet. we don't know how well it makes use of its bandwidth. that will be the deciding factor.

*if* the EBW numbers are real, and i doubt they are, everyone who was complaining about the 128bit bus will be eating their words.

[tamattack]

Originally posted by StealthHawk
and NV30 can't have a 256bit bus with 1+GHz RAM?

I never said that it can't. I merely said that the current rumours have it pegged at a 128 bit bus. And if so, then your earlier statement of:

Originally posted by StealthHawk
the GDDR3 being tested runs at 1GHz last we heard. that isn't "significantly faster" than the DDRII on NV30. hell, that probably won't be any faster.

is incorrect (again, if the 128 but bus rumour is true).

Originally posted by StealthHawk
and i still say it's premature to be judging NV30 without having seen anything about it yet. we don't know how well it makes use of its bandwidth. that will be the deciding factor.

I don't believe that I made any judgements in my earlier post. (Unless, you weren't talking specifically to me, in which case, never mind.)

Originally posted by StealthHawk
*if* the EBW numbers are real, and i doubt they are, everyone who was complaining about the 128bit bus will be eating their words.

True enough, although the cynic in me (partially because I work in marketing) tends to view these types of theoretical numbers (ie: EBW) as marketing sleight-of-hand designed to fool consumers who automatically think that a bigger number = better. (That also implies that the actual number is smaller than the competition).

That is, it's nothing more than a 'professional' way to have a penis waving match. ;)

[StealthHawk]

Originally posted by tamattack
I never said that it can't. I merely said that the current rumours have it pegged at a 128 bit bus. And if so, then your earlier statement of:

is incorrect (again, if the 128 but bus rumour is true).

i wasn't disputing your statements but this oneATI uses GDDR-3 which is ment to be significantly faster than DDR-II that Nvidia is using.

nowhere does it mention the total bandwidth that R350(or a R300) would have. it says that the memory itself is supposed to be "significantly faster." of course why NV35 or whatever can't simply use faster DDRII or even the same GDDR3 is beyond me.

I don't believe that I made any judgements in my earlier post. (Unless, you weren't talking specifically to me, in which case, never mind.)

nope, that chiding wasn't for you. it was for all the naysayers and their "16GB of raw bandwidth automatically = suckage."

[tazdevl]

Originally posted by Uttar
They'll be worthy of criticism if a competitor's product is released less than 3 months after the NV30 is faster.

If the NV30 is in stores before the R350 *and* faster than the R350, then it's certainly not a good idea to criticize nVidia.

However, if that isn't the case... Bad nVidia! Bad! :)
I most certainly wouldn't switch sides for so little, however. The NV30 is superior in features no matter what, so I still find it to be an interesting product.


Uttar

You sure the NV30 features will be superior to the R350? :rolleyes:

Stealth there was a link on the front page from a David Kirk interview a few days ago where he stated that a 256 bit bus isn't economically justifiable. That point is pretty much dead. It's 128 as confirmed by every preview.

[StealthHawk]

Originally posted by tazdevl
Stealth there was a link on the front page from a David Kirk interview a few days ago where he stated that a 256 bit bus isn't economically justifiable. That point is pretty much dead. It's 128 as confirmed by every preview.

i never said it would have a 256bit bus. i did say that before today, it was an unknown speculative assumption to say it 100% absolutely couldn't have one.

and i haven't "defended" a 256bit bus since the launch today. i'm not sure what exactly your point is.

obviously if David Kirk says that NV30 has a 128bit bus they have to make up the bandwidth somehow with special optimizations and techniques. it just doesn't make sense to clock a card so high and also require an external power unit unless the card can put the speed to good use(ala, we probably won't see another repeat of the GF2)