Nvidia GT300 core: Speculation

Status
Not open for further replies.
Between Beta 2 CPP (a fairly reasonable "almost-ready" version) and Vista RTM a full year elapsed (with a few post-B2 and RC builds added to the mix).
Frankly, i don't expect Windows 7 to be much different, except that perhaps Beta 2 will be more like a Release Candidate was for Vista (if there's any Beta 2, that is).

In that timetable, both GT300 and DX11 could conceivably be released without waiting of Win7 RTM, but i doubt Microsoft would deny itself the chance to pimp DX11 as part of Win7's new features in mainstream media, while quietly releasing it for Vista -as scheduled- a bit later.
 
I doubt it will be delayed since the planned launch window is (at least haven't seen anyone saying otherwise) still Q4/09-Q1/10
I still think it will hit Q3, early Q4 (october) latest. They have announced that the win 7 upgrade program will start july, 1st (meaning when you buy a computer with vista on or after that date, you receive a free win7 update). The Vista upgrade programm started october 2006. Go figure.

Edit:
Between Beta 2 CPP (a fairly reasonable "almost-ready" version) and Vista RTM a full year elapsed (with a few post-B2 and RC builds added to the mix).
Frankly, i don't expect Windows 7 to be much different, except that perhaps Beta 2 will be more like a Release Candidate was for Vista (if there's any Beta 2, that is).
Rumor has it that there will be no Beta2 and just a single RC.
 
If there is no GT212 in NVIDIA plans why there is "GT212 info" in the newest 185.20 drivers?

Moreover NVIDIA can`t delay GT212 to Q3 because there will be GT300 around the corner (Q4) and then there will be absolutely no reason to release Highend GT2xx based chip.

PS. KonKort said a few days ago that he has got confirmed info from someone in NVIDIA about specs and possible release date (5-6 months).
 
Charlies an idiot. I don't think there is anyone left that believes a word he says these days.

Still, thanks for the confirmation :smile:
 
With that rumored sp count GT212 will be a huge chip even at 40nm, which incidentally is a new process. Suppose that's true, it leaves little room for GT300 to improve unless one is to expect a RV770 type miracle in less than two quarters. This also departs form the very sensible tick-tock(G80-G92) style Nvidia has employed.

According to floating rumours GT212 should increase the sp count, by a relatively small degree the TMU count and dicrease the ROP count by half; considering that if memory serves well ALUs capture roughly 1/4th of a GT2x0, you may play around with some speculative math and make a rough estimate how "huge" such a chip would really be under 40nm.

As for GT300 I'd be very surprised if it being a D3D11 architecture it would end up with that much in common with GT2x0/G8x0 whatever. After all I don't think they skipped 10.1 so far purely out of strategical reasons.

I concur with the post above that G2xx family is generally late by six month. GT200 was supposed to be out at early 2008 and it makes the most sense that a shrink GT212 or whatever it's called should have been out about now.

AMD seems to introduce its first mainstream chip under 40nm and then will move in due time to a larger performance chip. If NV is planning first a performance 40nm chip and later on GT212 I don't see why the latter should be out by now.

As for Charlie's stuff that it has been canned, Fudzilla also reported shortly before the 295 announcement that AIBs claimed that the dual-GPU config has been canned. That shouldn't mean that they might or might not can the 212, but we'll probably know what is going on when the time for the 212 projected release comes and not now. AFAIK it was projected for Q2 09' and it'll still take 7 months until we reach the end of Q2.
 
Charlie is not saying 212 has been canned, just that it probably won't be out before Q3 (which fits konkorts 5-6 months btw), and that it maybe wouldn't make that much sense to release it that late.
However, if it's a < 300mm2 chip, I don't see why it wouldn't make sense along the gt300, if the gt300 is following tradition and is 400+
 
Last edited by a moderator:
That their 3 lead customers are AMD, NVIDIA and Altera. And I've seen musings Altera will be the last, and they've already got samples... Of course, I could be reading too much into that or it might simply be wrong.
Hmm, interesting, I got the distinct impression (admittedly from Altera's technical/marketing bumf) that Altera would be first to market and is in a much stronger partnership with TSMC than either AMD or NVidia.

Jawed
 
I thought Altera was to be the one first out as well.
I also have the impression that an FPGA would be easier to tape out and verify than a GPU.
 
Now that I think about it, I think those musings were about *mass production* - so indeed it'd be very plausible that Altera taped-out first and is sampling their customers before AMD/NVIDIA. Which brings us right back to square one, heh.
 
Is it conceivable that their first 40nm product could actually be one of the "low-power chips" they talk about here ?
Or are we talking Tegra instead ?
 
If 7 is delayed, would that mean delaying DX11 for Vista as well?

Wasn't MS planning to release DX11 "before" Windows 7? I'm pretty sure I've read in various places that DX11 was targeted for a summer/fall rollout while Windows 7 is currently on tract for a Winter/Spring rollout.

And even if that were not the case. MS has already stated that DX11 will be available on Vista so there would be no reason to delay it. Unlike the case with Windows XP and DX10 since DX10 required the (then) newly revamped rendering layer that debuted with Vista. Since Vista and Windows 7 share the same graphics under pinning there is no need delay it's release to coincide with the OS.

Regards,
SB
 
However, if it's a < 300mm2 chip, I don't see why it wouldn't make sense along the gt300, if the gt300 is following tradition and is 400+
Theoretically it makes sense to release such a chip as a high end single chip GPU if it won't face any delays and let it slip with the release of the GT3x0 into the performance segment. If there should be a delay after all it'll probably end up being a performance GPU since its release. In such a case the only other sensible gap filler for mid 09' sounds to be a 295 follow up with higher frequencies. "Nothing" doesn't seem likely for both IHVs for this year's summer.
 
Wasn't MS planning to release DX11 "before" Windows 7? I'm pretty sure I've read in various places that DX11 was targeted for a summer/fall rollout while Windows 7 is currently on tract for a Winter/Spring rollout.
Just rumors. Microsoft even gave some hints that DX11 would come after Win7. Later is was clarified that Win7 would ship with DX11.
And even if that were not the case. MS has already stated that DX11 will be available on Vista so there would be no reason to delay it. Unlike the case with Windows XP and DX10 since DX10 required the (then) newly revamped rendering layer that debuted with Vista. Since Vista and Windows 7 share the same graphics under pinning there is no need delay it's release to coincide with the OS.
There may be no technical reasons to delay DX11, but from a marketing point of view it would be incredibly stupid not to tie the release of DX11 to the release of Win7. Actually I think that MS may even delay DX11 for Vista a bit (after Win7's release) in order to give Win7 a headstart.
 
Shoot me for the OT, but considering what I've read so far about win7 beta I don't think Microsoft needs anything else but to reduce the projected prices for it against Vista.
 
Shoot me for the OT, but considering what I've read so far about win7 beta I don't think Microsoft needs anything else but to reduce the projected prices for it against Vista.
I think Microsoft would really like to keep it's current price level and still sell as many Win7 copies as possible. I heard they really like their margins. ;)
 
Status
Not open for further replies.
Back
Top