No low-k for NV in the near future...

Huang:As for products, Nvidia’s GeForce FX is no doubt the performance leader on the market now.[

What he said few months ago bout ATI's CEO? Taking hallucinogens? :D
 
Not using copper currently? Strange. Well, whatever makes the NV35 as fast/faster than the FX Ultra and cooler, I'll support it.

Side note: the #1 thing I want to see from nVidia next is FSAA on par with ATI's (in algorithm, not necessarily number of samples).
 
Chalnoth said:
Not using copper currently? Strange. Well, whatever makes the NV35 as fast/faster than the FX Ultra and cooler, I'll support it.

Yeah - that's crucial. It'd be nice to get confirmation that current NV30 dice use Al interconnects. I had presumed low-k (+ die tweaks) to get the NV30>>>NV35 clock bump, but perhaps they are changing interconnect tech.

MuFu.
 
nVidia *insisted* several times that the NV30 was using Copper during the live product launch, so it would really be a last minute change if they were using Aluminium. And anyway, they aren't. This quote was just not interpreted correctly: Jen Hsun Huang wasn't very clear.

This quote is incomplete, and does not reveal the true meaning of the sentence. Here's a more complete quote ( bold is my addition ) :
Q: How has the market responded to Nvidia’s first 0.13-micron process graphics chip, the GeForce FX (NV30)? Could you tell us more about Nvidia’s production plans in 2003?

A: At the initial stage, Nvidia will ship 1.5 million NV30 chips and will also make cards based on the top-end GeForce FX 5800 Ultra (NV30 core) itself. The lower-end GeForce FX 5800 chips, however, will be released to downstream card manufacturers for production.

Currently, Nvidia’s wafer orders at Taiwan Semiconductor Manufacturing Company (TSMC) are mostly intended for the new NV30-family chips. With launches of other NV30-family products in the near future, we expect that NV30-based products will account for more than 50% of Nvidia’s overall shipments by year-end.

So far, most of Nvidia’s graphics chips have been manufactured on the 0.15-micron process at TSMC’s 8-inch fabs, but we expect the ratio of 0.13-micron production will grow greatly this year. The timing of the transfer of production to the 12-inch fabs depends on TSMC’s schedule. When we move production to the 12-inch fabs, we will also start adopting the copper process. For the moment, we are not considering using low-k technology due to its immaturity at the upstream stage.

In the part quoted by McElvis, Jen Hsun Huang is talking about *most of nVidia's graphics chips* , and commenting about nVidia's production plans for 2003.
Thus, when he's saying "we will also start adopting the copper process", he means nVidia will adopt the copper proccess for *mainstream* parts at the same time as 12-inc fabs.

Thus, what nVidia's CEO is really trying to say here, is that the NV31 and NV34 both don't use copper but aluminium, that mainstream won't use Low K for a *long* while and that the NV36 will likely use Copper.
I don't think he really meant to talk about high-end products there, so Low K is still possible for the NV35 ( although recent comments about the immaturity of the Low K proccess makes it less and less likely )


Uttar
 
You could take what he says both ways, so yes, he was not being very clear.

And, as in my original post, I always thought that the NV30 was on copper. I never new that they insisted it was using copper at the launch, but that's what I heard around here.

Sorry for any confusion caused :oops:
 
Chalnoth said:
Side note: the #1 thing I want to see from nVidia next is FSAA on par with ATI's (in algorithm, not necessarily number of samples).

Is there any confirmation by Nv that the NV35 will atleast do this?? Do can we assume IQ won't be matched until NV40.
 
Just seen a post from Hellbinder @ nvnews concerning the R400.. not being available in for the end of the year, as the low-k process not being available..

Anyone in the know.. also want to confirm this statement..

At the moment.. Im :cry: :cry: :cry:

/me kicks a stone, as he trudges out of the thread. :(
 
demonic said:
Just seen a post from Hellbinder @ nvnews concerning the R400.. not being available in for the end of the year, as the low-k process not being available..

Anyone in the know.. also want to confirm this statement..

At the moment.. Im :cry: :cry: :cry:

/me kicks a stone, as he trudges out of the thread. :(

I read a quote the other day from ATI saying that the R350 would be the performance leader till at least the end of the year.

Its on this board somewhere. Straight from the horses mouth
 
Yeah...we're likely not to get anything other than the more traditional "incremental" upgrades for the next 12 months.

On the other hand, Hellbinder indicated that ATI is in fact, now planning on a 0.13 R350 (9900) to bring to market inbetween now and the R400 release.

Assuming this is true, that is not something that would be undertaken lightly, and does indicate to me, that the "Best" case scenario for R400 is for the spring '04 cycle. The good news is, that ATI is not willing to let the 0.15u R350 ride it out for a full year....
 
He didn't say the R350 would maintain the performance lead throughout the year, he said it ensured that ATI would maintain the performance lead throughout atleast the calender year (i.e. beat the GFFX, presumably beyond question).



I especially liked the part about how they had 0 problems going to .13 ;)
 
Back
Top