HYPOTHETICAL: R420 two months earlier than nV40?

[digitalwanderer]

Ok, assuming that Fuad's latest story in the Inquirerererer is right and nVidia canned the nV40 and pushed up the nV45 to replace it to compete with the R420, (which I'm actually believing for the moment since it makes so much sense), and that there currently isn't any working silicon and we're looking at an April paper launch with a target around May sometime for the actual card and who knows when for availability.

I know, I know...it's a lot to assume. That's why I put the big "HYPOTHETICAL" in the thread title, and it's just the hypothetical situation I want to speculate about a bit.

How much of an advantage for ATi/disadvantage for nVidia do you think a 2 months give? Normally I'd think that nVidia would be hot selling the sheet out of their paper launch and trying to FUD people to delay, but even if they're telling the truth no one is going to wait for another "soon coming wondercard" from nVidia after the nV30 fiasco.

I think the hypothetical delay could really hurt 'em, anyone else got an opinion?

 

[Florin]

I think a lot of people won't buy either until they've seen benchmarks between the both of em.

 

[MrGaribaldi]

How much of an advantage for ATi/disadvantage for nVidia do you think a 2 months give? Normally I'd think that nVidia would be hot selling the sheet out of their paper launch and trying to FUD people to delay, but even if they're telling the truth no one is going to wait for another "soon coming wondercard" from nVidia after the nV30 fiasco.

I think the hypothetical delay could really hurt 'em, anyone else got an opinion?

Such a delay would potentially hurt them, but it all depends on the lengths nvidia are wiling to go to stop ATI from gaining a bigger marketshare.

They would loose the high-end and the sales associated with having the top-performer, but they could limit this damage, by cutting the prices on their current line-up. If the reduction is big enough, I think many people might be swayed to buy nvidia instead, but this might require nvidia to loose money with each card sold.

They could possible try to get developers to go out and say buy NV40, as it'll perform great in our game, but not certain how much impact that would have nor how much it would cost. iD have gone out and said the NV3x will have good performance in D3, but by doing that now, I'm afraid the same tactic won't work later (with other game devs).


What I'm basicly saying is that if this comes to pass, it'll hurt nvidia, but how they choose to take the loss (money or marketshare) is up to them.

 

[John Reynolds]

This assumes ATI won't sit back and continue selling their current parts for another month or two. Why go through the costs of transitioning your current line prematurely (which isn't entirely vertical in that it only addresses the mid and upper ends of the market) when it's currently outselling the competition by good margins. The enthusiast mind thinks, "yeah, launch the next generation and pull even further ahead, albeit even if for a short time", but that's not necessarily good business. IMO

 

[Johnny Rotten]

Without Doom3 and/or Half-Life 2 I dont think a relatively small 2 month delay will really hurt them much at all (I myself for example wont be considering ANY new card until one of those two titles are out). However if there is a bigname, big publicity, triple A title out there to for a single card to monopolize....

 

[digitalwanderer]

This assumes ATI won't sit back and continue selling their current parts for another month or two. Why go through the costs of transitioning your current line prematurely (which isn't entirely vertical in that it only addresses the mid and upper ends of the market) when it's currently outselling the competition by good margins. The enthusiast mind thinks, "yeah, launch the next generation and pull even further ahead, albeit even if for a short time", but that's not necessarily good business. IMO

Ah, but there in lies the rub! IF the R420 is close or faster than the nV40 than it would be a great strategy, but what if the nV40 is about 20-30% faster than the R420? Wouldn't launching early and taking some of the marketshare/mindshare be a good thing?

Rather than be almost as fast as the competition, they'd be the hottest thing around and totally un-equaled for a couple of months...it's GOTTA be tempting for 'em.

 

[THe_KELRaTH]

There are ppl that will buy the ATI product whether or not it's faster than the NV40, for this group it makes little odds if the R420 is out before or after the NV40.
Then there's the group that will wait till both products are out though if one product is significantly later (over 2+ months) I would think this would change - especially for the ppl who waited for the NV30 (Deshavue).

Therefore I see no reason why ATI wouldn't just stick to their planned timetable.

 

[thatdude90210]

Ok, assuming that Fuad's latest story in the Inquirerererer is right and nVidia canned the nV40 and pushed up the nV45 to replace it to compete with the R420, (which I'm actually believing for the moment since it makes so much sense),

Didn't we go through this exact same theory/story for the NV30. Before it's release, there were speculation that original Nv30 was canned, that the part being released as "Nv30," aka Dustbuster, would be the Nv35 being pushed up to meet the R300. Are we sure it isn't just the marketing team at work here, trying to make people feel like they are getting more than they are getting?

 

[digitalwanderer]

Ok, assuming that Fuad's latest story in the Inquirerererer is right and nVidia canned the nV40 and pushed up the nV45 to replace it to compete with the R420, (which I'm actually believing for the moment since it makes so much sense),

Didn't we go through this exact same theory/story for the NV30. Before it's release, there were speculation that original Nv30 was canned, that the part being released as "Nv30," aka Dustbuster, would be the Nv35 being pushed up to meet the R300. Are we sure it isn't just the marketing team at work here, trying to make people feel like they are getting more than they are getting?

If it's delayed like we're hearing, then it better NOT be just marketing speak now 'cause they'll be needing a killer card if it comes out late!

 

[martrox]

If it's only 2 months I don't think it's a big deal.....IF the NV40 is a goodly deal faster/better. However, if it's 6 months - and if there's no working silicon ATM, who knows - and/or the NV40's slower/lesser IQ then all bets are off.....

I don't think nVidia can even remotely stand another debacle like the last 2 years....... and they will be under a microscope!

 

[digitalwanderer]

If it's only 2 months I don't think it's a big deal.....IF the NV40 is a goodly deal faster/better. However, if it's 6 months - and if there's no working silicon ATM, who knows - and/or the NV40's slower/lesser IQ then all bets are off.....

It's the "there's no working silicon ATM" bit that scares me, what if two months late is nVidia's best case scenerio?

 

[Arun]

Dig, A0/A1/A2 are fully working and operational.
The 210M part wouldn't seem to be working anywhere yet, but since it's fundamentally based on the old architecture, I doubt it'll be much of a problem.

The problem here, I assume, is mass producing the part.
This means NVIDIA would have no difficulty to synchronize a "reviews at launch" movement synchronized, or rather a few days before, ATI's retail avaibility. So, if the NV40 rocks, they'd obviously have some, errr, difficulties selling them if they barley have any samples, but they'd have no problem ruining ATI's sales and generating tons of hype, which is what matters for them at this point, I assume.


Uttar

 

[BRiT]

The problem here, I assume, is mass producing the part.
This means NVIDIA would have no difficulty to synchronize a "reviews at launch" movement synchronized, or rather a few days before, ATI's retail avaibility. So, if the NV40 rocks, they'd obviously have some, errr, difficulties selling them if they barley have any samples, but they'd have no problem ruining ATI's sales and generating tons of hype, which is what matters for them at this point, I assume.

If Nvidia's supply of their high-end NV40 is as scarce as their NV30 was, I'd like for nothing better than for ATI to do the same. ATI should produce and demonstrate a working 2-chip R42x board. I'm sure that would claim the performance crown. 8)

 

[John Reynolds]

So, if the NV40 rocks, they'd obviously have some, errr, difficulties selling them if they barley have any samples, but they'd have no problem ruining ATI's sales and generating tons of hype, which is what matters for them at this point, I assume.

Uttar

I don't think nVidia gets the free pass onto the hype highway like they used to, not after last year. But, yes, they would certainly strive to ruin a real launch by a competitor with a paper (spec) one of their own. As would ATI, if the roles become reversed.

 

[Bouncing Zabaglione Bros.]

This assumes ATI won't sit back and continue selling their current parts for another month or two. Why go through the costs of transitioning your current line prematurely (which isn't entirely vertical in that it only addresses the mid and upper ends of the market) when it's currently outselling the competition by good margins. The enthusiast mind thinks, "yeah, launch the next generation and pull even further ahead, albeit even if for a short time", but that's not necessarily good business. IMO

Ah, but there in lies the rub! IF the R420 is close or faster than the nV40 than it would be a great strategy, but what if the nV40 is about 20-30% faster than the R420? Wouldn't launching early and taking some of the marketshare/mindshare be a good thing?

Rather than be almost as fast as the competition, they'd be the hottest thing around and totally un-equaled for a couple of months...it's GOTTA be tempting for 'em.

That's what I think. If NV40 is months away, then ATI can get a big jump on Nvidia. The longer ATI delay, the more time Nvidia has to make a better part. The quicker ATI launch, the more time they have to dominate the market and consolidate their grip on the high/mid end.

If NV40 arrives and is better, then ATI have a (maybe) six month lead in the market, and can look to launching the refresh product or price cuts to fight back. If NV40 arrives and is worse, then ATI have already got a grip on this generation's market and can just put even more pressure on Nvidia.

 

[digitalwanderer]

Dig, A0/A1/A2 are fully working and operational.
The 210M part wouldn't seem to be working anywhere yet, but since it's fundamentally based on the old architecture, I doubt it'll be much of a problem.

You're sure on that? :| (And the "AO/A1/A2" is the first three runs of the nV40, right? Or is it the 'refreshed nV45'? Or should I go outside until the bleach fumes in here die down a bit as I'm all just muddled up again? )

The problem here, I assume, is mass producing the part.
This means NVIDIA would have no difficulty to synchronize a "reviews at launch" movement synchronized, or rather a few days before, ATI's retail avaibility. So, if the NV40 rocks, they'd obviously have some, errr, difficulties selling them if they barley have any samples, but they'd have no problem ruining ATI's sales and generating tons of hype, which is what matters for them at this point, I assume.

I don't think a few reviewer samples out will generate quite the hype/impact they expect after the last fiasco, I really don't.

 

[John Reynolds]

That's what I think. If NV40 is months away, then ATI can get a big jump on Nvidia. The longer ATI delay, the more time Nvidia has to make a better part.

Sure, because ASIC layouts are simple designs easily modified at the last minute to respond to competition-based surprises.

Guys, don't make me use the eye rolling emoticon.

 

[Joe DeFuria]

Dig, A0/A1/A2 are fully working and operational.
The 210M part wouldn't seem to be working anywhere yet, but since it's fundamentally based on the old architecture, I doubt it'll be much of a problem.

Can you explain what you mean here?

Are you talking about the A0/A1/A2 "original, but now canned" NV40 was working and operational, but the "NV45" (210m transistor replacement) is not yet up and running?

If this is all true (and this is what you're saying), then nVidia is further behind than I thought.

 

[digitalwanderer]

That's what I think. If NV40 is months away, then ATI can get a big jump on Nvidia. The longer ATI delay, the more time Nvidia has to make a better part.

Sure, because ASIC layouts are simple designs easily modified at the last minute to respond to competition-based surprises.

Guys, don't make me use the eye rolling emoticon.

Fair enough, how about the longer ATi needlessly delays ("needlessly" is assuming the card is ready to go) the less late the nV40 will look?

Just from a psychological perspective, the quicker the R420 comes out the later the nV40 comes out....and with nVidia's last release delay I'm sure that could easily work in ATi's favor.

 

[Bouncing Zabaglione Bros.]

That's what I think. If NV40 is months away, then ATI can get a big jump on Nvidia. The longer ATI delay, the more time Nvidia has to make a better part.

Sure, because ASIC layouts are simple designs easily modified at the last minute to respond to competition-based surprises.

Guys, don't make me use the eye rolling emoticon.

What if NV40 is better? ATI would get advantage from launching *now*, not waiting until a better competitor arrives. NV40 is months away, and that extra development time may mean Nvidia can get a better product out than if they had to launch tomorrow.

Of course I'm not suggesting that Nvidia will redesign their chip because of ATI (though they may have scrapped one or two already), but the longer ATI leave it, the more chance that Nvidia can get NV40/45 working optimally, or decide to clock at a higher, rather than lower speed. Nvidia may be able to take advantage of process refinements or otherwise use the extra development time to improve the product.

My point is that NV40/45 is still in development and can benefit from improvements. R420 is finished and is effectively standing still - it isn't going to get any better sitting around in warehouses waiting for NV40 to launch.