HYPOTHETICAL: R420 two months earlier than nV40?

[Fodder]

Hum, i have not really the technical knowledge but from my readings here since late 2002 and the whole nv 30-nv 35 story, you just can't decide to skip a step and keep same time frame for release.

Remember, the original timeframe for NV40 was Q4 '03, when NV38 came out, or so the rumourmill spake.

 

[DoS]

Hum, i have not really the technical knowledge but from my readings here since late 2002 and the whole nv 30-nv 35 story, you just can't decide to skip a step and keep same time frame for release.

Remember, the original timeframe for NV40 was Q4 '03, when NV38 came out, or so the rumourmill spake.

Indeed, however there were some delays (i don't know why) and by the end of 2003 all the readings were saying that NV40 had taped sometime early December. What have they been up to the past 3 months ? I insist, a late May launch sounds just right.

 

[Doomtrooper]

Then ATI's in trouble, aren't they?

Why would they be in trouble ??

 

[Hanners]

Then ATI's in trouble, aren't they?

Why would they be in trouble ??

Assuming the rumours are true, ATi will be pretty much equal with nVidia on AA quality, and behind on the level of shaders implemented. Therefore, going by your earlier prediction, ATi would be in trouble.

 

[MrGaribaldi]

Going the "i ll believe it when i see it" route is fair enough. However, IMHO NVIDIA took the decision to can the old NV40 and proceed with NV45 earlier than most ppl in this forum think. The reason i am making this assumption is that we heard some time ago (by multiple sources) that the original NV40 had tapped-out by late 2003...it's almost 4 motnhs since then and as someone else pointed out in this thread the CEBIT launch was never the plan as of January. What i am saying is that all indications are that NVIDIA took the decision to proceed with NV45 earlier than most here think, and if that was late December last year then a May launch makes sense. I strongly expect a May launch for both products

my 2 eurocents

Ailuros said here that the NV40 & NV45 taped out at the same time, so wouldn't that mean that they were both "ready" in dec 2003?

But I guess the question is what is really the difference between the NV40 & NV45? Is the NV45 the refresh part to the NV40, or just NV40 with PEG?

He actually performs these tasks? Gosh. I woulda thought there's a solid DevRel dept at NVIDIA...

But wouldn't sending Kirk out lend more credibility to what they're saying, since I would guess most Dev's are taking everything Nvidia says about their nextgen card with a pinch of salt.
Ie. A dev thinking "This DevRel guy says the NV40 will rock and blow ATI out of the water. But that's what he said about the NV30 in 2002, so why should I believe him?" compared to "Man, Kirk says the NV40 will rock and blow ATI out of the water. Had it been a DevRel guy I might not have believed it, but Kirk should know..."

 

[Joe DeFuria]

Ailuros said here that the NV40 & NV45 taped out at the same time, so wouldn't that mean that they were both "ready" in dec 2003?

No, tape-out doesn't mean "ready." There is a significant length of time between when a chip tapes out, and when it's "ready" for production (and then there are issues with production itself that can lead to significant time frame for "ready to ship in quantity.)

Both could tape-out at the same time, but both could be "commercially ready" at very different times. We have no real way of knowing either way.

But wouldn't sending Kirk out lend more credibility to what they're saying..

Actually, I'd think it would be less. (The higher up you go, the less credibility, IMO. )

 

[digitalwanderer]

But wouldn't sending Kirk out lend more credibility to what they're saying..

Actually, I'd think it would be less. (The higher up you go, the less credibility, IMO. )

LOL! Thanks Joe for being the first to say that, it probably would have sounded much more fanboyish coming from me.

 

[oddfellow]

But wouldn't sending Kirk out lend more credibility to what they're saying..

Actually, I'd think it would be less. (The higher up you go, the less credibility, IMO. )

Exactly! My thinking would be; "Wow, nVidia must be desperate if they're sending a big gun like Kirk out to tell us that it rocks"

 

[MrGaribaldi]

Ailuros said here that the NV40 & NV45 taped out at the same time, so wouldn't that mean that they were both "ready" in dec 2003?

No, tape-out doesn't mean "ready." There is a significant length of time between when a chip tapes out, and when it's "ready" for production (and then there are issues with production itself that can lead to significant time frame for "ready to ship in quantity.)

Both could tape-out at the same time, but both could be "commercially ready" at very different times. We have no real way of knowing either way.

Ah, thank you for clearing that up. I thought that when a tape-out was complete, it would take the same time for both to be "commercially ready" at the same time (allthough depending on yields for both cores).

But wouldn't sending Kirk out lend more credibility to what they're saying..

Actually, I'd think it would be less. (The higher up you go, the less credibility, IMO. )

Hmm... Didn't look at it like that...

I thought that the higher up the person was, the less likely the person would be to pull a fast one, since then that person would loose all credibility. (which would mean the usefullnes of that person in the future would decrease)

 

[MrGaribaldi]

Actually, I'd think it would be less. (The higher up you go, the less credibility, IMO. )

Exactly! My thinking would be; "Wow, nVidia must be desperate if they're sending a big gun like Kirk out to tell us that it rocks"

But wouldn't they think nvidia was desperate if they couldn't get anything to market before ATI, no matter who they send out?

And of course, I'm not suggesting that Kirk should be sendt around to every Dev out there, but to those judged as having a major influence on the rest of the market.

If he were sendt to small garage dev's to talk about the NV40, then it would be another matter entirely.

 

[digitalwanderer]

Hmm... Didn't look at it like that...

I thought that the higher up the person was, the less likely the person would be to pull a fast one, since then that person would loose all credibility. (which would mean the usefullnes of that person in the future would decrease)

*COUGH-COUGH* Derek Perez & Brian Burke *COUGH-COUGH*

 

[MrGaribaldi]

*COUGH-COUGH* Derek Perez & Brian Burke *COUGH-COUGH*

LoL, it would seem my head still hadn't managed to get that through my thick skull.

But hopefully this time it's been

into my head.

I'll just go sit in the corner for a while, ok?

 

[digitalwanderer]

*COUGH-COUGH* Derek Perez & Brian Burke *COUGH-COUGH*

LoL, it would seem my head still hadn't managed to get that through my thick skull.

But hopefully this time it's been

into my head.

I'll just go sit in the corner for a while, ok?

Just remember that golden rule, "shit rolls downhill"...

 

[Joe DeFuria]

LoL, it would seem my head still hadn't managed to get that through my thick skull.

Seriously, the higher up you are, the less "in the trenches you are" so to speak, and typically the less grounded in "practical reality" you are. Higher-ups also tend to fall victim to their reports telling them "what they want to hear", vs. "what they need to hear" .

Higher-up doesn't automatically mean "more likely to spew B.S.", but it does more often mean a more "birds eye view" of things, which usually doesn't tell the whole story, which means there is more room for B.S.

 

[StealthHawk]

Then ATI's in trouble, aren't they?

Why would they be in trouble ??

Assuming the rumours are true, ATi will be pretty much equal with nVidia on AA quality, and behind on the level of shaders implemented. Therefore, going by your earlier prediction, ATi would be in trouble.

Exactly. And NVIDIA already has a better implementation of AF than ATI, IQ-wise. If the rumors are true and NV40 has "ATI-like" FSAA, and it also supports PS3.0/VS3.0 while R420 doesn't...going by Doomtrooper's metric of "features selling cards" it would look like a clear win for NV40. That is, assuming that NV40 has what it supposedly has, and that R420 doesn't have what it supposedly doesn't have.

Personally, I'm not so sure that DT's criteria is correct. For example, R3xx was so great not because it supported PS2.0, and did it fast, but because it was so much faster with FSAA/AF than R200 or NV25. Not only that, but the FSAA quality was better. Two birds, one rock. I think as games use more shaders, they will be more GPU limited than CPU limited, so if NV40 and R420 are a lot faster in that department as is claimed, we should see some substantial performance gains in that department with the onset of the next generation cards.

 

[Doomtrooper]

I see based off forum speak and rumors the R420 is not up to speed to a NV40, I also see we are barking up the same tree that was spread here by the usual suspects how the Nv30 was going to destroy the R300.

Surprises are in for alot of people that believe these 'rumors'.

 

[ben6]

Hum, i have not really the technical knowledge but from my readings here since late 2002 and the whole nv 30-nv 35 story, you just can't decide to skip a step and keep same time frame for release.

Remember, the original timeframe for NV40 was Q4 '03, when NV38 came out, or so the rumourmill spake.

Indeed, however there were some delays (i don't know why) and by the end of 2003 all the readings were saying that NV40 had taped sometime early December. What have they been up to the past 3 months ? I insist, a late May launch sounds just right.

I would suggest late May to be incorrect. IF they were to launch in May it'd be around E3 time like they did 5900 U, which is May 11-13.

 

[John Reynolds]

I see based off forum speak and rumors the R420 is not up to speed to a NV40, I also see we are barking up the same tree that was spread here by the usual suspects how the Nv30 was going to destroy the R300.

Surprises are in for alot of people that believe these 'rumors'.

If ATI still command a significant PS 2.0 performance lead with this next gen., they need a killer app to show a real world benefit of that lead. IMO. Otherwise, bulletable features supported could win the day via good PR/marketing.

 

[Tahir2]

I believe the next 'true' NVIDIA card (whatever it has been designated internally) is late. It looks like NVIDIA may be significantly behind ATI in release schedule for the second time (R300 vs NV30) for similar reasons. I also believe the next ATI card (whatever it has been designated internally) is on time for a May release.

If ATI is indeed ready we can expect a paper launch very soon.

NVIDIA, apart from the PCI-Express noises it has been making, is expected to respond with something that will beat the new ATI card in everything from Quake3 through to Doom3/HL2 or not respond at all, and I think it is pretty safe to assume that NVIDIA has opted for the latter.

Edit: I meant opted for releasing an ATI card killer (speed and featurewise) late rather than a card that is barely faster with hacks and degraded IQ.

P.S. I believe NVIDIA still has not gotten their AA implementation as good as ATI's.

 

[Bouncing Zabaglione Bros.]

I believe the next 'true' NVIDIA card (whatever it has been designated internally) is late. It looks like NVIDIA may be significantly behind ATI in release schedule for the second time (R300 vs NV30) for similar reasons.

You're implying design/fabrication issues? If so, who knows what kind of problems or compromises the end product will have?