Go Back   nV News Forums > Hardware Forums > Rumor Mill

Newegg Daily Deals

Reply
 
Thread Tools
Old 08-17-02, 02:47 PM   #25
Fotis
Radeforce GTX7970
 
Fotis's Avatar
 
Join Date: Jul 2002
Location: Greece
Posts: 1,346
Default

Quote:
Originally posted by SavagePaladin
The memory speed at the very least is known to be higher. Thats one rumor I don't doubt.

Anyway, carry on while I go back to sleep.
Did you drink your milk first?
I know the memory will be higher I was talking about core clock.
Everything I right is based on facts and past performance by nvidia.

Last edited by Fotis; 08-17-02 at 03:06 PM.
Fotis is offline   Reply With Quote
Old 08-17-02, 03:18 PM   #26
Matthyahuw
Registered User
 
Matthyahuw's Avatar
 
Join Date: Jul 2002
Location: AZ
Posts: 919
Send a message via ICQ to Matthyahuw Send a message via AIM to Matthyahuw Send a message via Yahoo to Matthyahuw
Default

of course the clock will be higher...that's one of the advantages of moving to a smaller die, that and less heat...
__________________
Shalom!
Matthyahuw is offline   Reply With Quote
Old 08-17-02, 03:25 PM   #27
Fotis
Radeforce GTX7970
 
Fotis's Avatar
 
Join Date: Jul 2002
Location: Greece
Posts: 1,346
Default

I know all that and I hope it goes up to 400mhz(Dangerously closing my cpu speed )
I'm just posting what has happened with other nvidia products gf2 on 0.18 and got 250 and gf3 on 0.15 got 200 if I'm right.
Moreover nv30 has almost double the transistors of a gf4 don't think that nv30 will be a fridge!!

Last edited by Fotis; 08-17-02 at 03:29 PM.
Fotis is offline   Reply With Quote
Old 08-17-02, 05:04 PM   #28
SavagePaladin
info*****
 
Join Date: Jul 2002
Posts: 1,395
Default

Quote:
Originally posted by Fotis

Did you drink your milk first?
I know the memory will be higher I was talking about core clock.
Everything I right is based on facts and past performance by nvidia.
Chocolate milk, sure.
o.O
SavagePaladin is offline   Reply With Quote
Old 08-17-02, 06:35 PM   #29
Bigus Dickus
GF7 FX Ti 12800 SE Ultra
 
Join Date: Jul 2002
Posts: 651
Default

Quote:
Originally posted by StealthHawk
ArtX engineers designed R300 entirely on their own without any input whatsoever from other engineers working at ATI? unless that above statement is true, the parallel you are drawing is not accurate.
And one single group of engineers at NVIDIA designed both the NV2A and NV30?

I'd guess there was at least some input from other engineers, perhaps even different design teams worked on the two chips.

Again, the bottom line is that the ArtX group designed a console and PC chip, both came out fine, and on time. The XBOX is not an "excuse" for NVIDIA's delays; it's simply one cause of the delays... one which doesn't affect everyone.
Bigus Dickus is offline   Reply With Quote
Old 08-17-02, 06:39 PM   #30
Bigus Dickus
GF7 FX Ti 12800 SE Ultra
 
Join Date: Jul 2002
Posts: 651
Default

Quote:
Originally posted by SurfMonkey
I guess the distinction that people are making is that ATi bought ready made technology and shoe horned it into a PC chip. nVidia, on the other hand, bought the rights to a whole load of theoretical technologies and had to do their design from the ground up.
And that's a false distinction.

ATI didn't "buy the technology and shoe horn it into a PC chip;" they bought the group of engineers that designed it, and they then designed the R300. No doubt they used knowledge gained from the Flipper Chip in the design of the R300, but isn't that just good engineering practice? Are you saying that since NVIDIA designed the NV30 "from the ground up" that they didn't use any knowledge gained from the NV2A and NV20?

I don't understand how people can see the two situations as different.

Group A engineers designed a console chip, then designed a PC chip, and both were on time.

Group B engineers designed a console chip, then designed a PC chip, and the latter is significantly late. (Assuming group B designed both... we know group A did).

That each group probably had input from other engineers and use previously gained knowledge from past designs is a given, and completely irrelevant (besides, it's the same for both sides, so even if it wasn't irrelevant... it still would be ).
Bigus Dickus is offline   Reply With Quote
Old 08-17-02, 06:52 PM   #31
Bigus Dickus
GF7 FX Ti 12800 SE Ultra
 
Join Date: Jul 2002
Posts: 651
Default

Quote:
Originally posted by Matthyahuw
of course the clock will be higher...that's one of the advantages of moving to a smaller die, that and less heat...
I think it's more likely that the NV30 will be clocked higher (core) than the R300, but I don't think it's a certainty (I'd guesstimate something like 70% probability it's 325MHz or greater).

In fact, I think the 9700Pro's clock speed is probably at or slightly above the original target for the NV30.

Most people thought the R300 would never see speeds above 250MHz, and since it's a more complex chip than Matrox's Parhelia (which is stuck at 220Mhz), many thought it would be in the low 200's.

With that kind of expected clock speed, it was "a given" that the NV30 would be significantly faster. However, now that ATI has pulled a rabbit out of their ass with 325Mhz, it's not "a given" that NVIDIA will be able to do the same, and just arbitrarily raise the core speed to ~400Mhz.

In fact, part of the delays might simply be hand tweaking of the core design for improved frequencies... something ATI did to achieve 325Mhz, and something NVIDIA probably wasn't planning on having to do.

ATI is drawing over AGP spec, with a hand tweaked design, on a very mature and well refined process. Whether a better but immature .13u process will be a benefit or not remains questionable (I think it will be a benefit... but perhaps not initially). Has NVIDIA hand tweaked the NV30 in the same manner? Are they willing to go beyond AGP specs? (If the NV30 was purely digital logic, it should be able to run >400MHz and still be below the 40W maximum, but with RAMDACS and other components, that is still uncertain.)

My feeling is that NVIDIA will do whatever it takes... power connector, more time, whatever... to clock the NV30 above 325MHz, even if it only by 25Mhz or so. I think they have to, because that will be the NV30's most significant advantage.
Bigus Dickus is offline   Reply With Quote
Old 08-17-02, 11:43 PM   #32
jbirney
Registered User
 
jbirney's Avatar
 
Join Date: Jul 2002
Posts: 1,430
Default

Quote:
What I meant was the Gigapixel and 3dfx technologies were all theoretical, nothing had ever been, successfully, built with them. nVidia had to do all the design work and implementation work from scratch. ATI bought a ready made design and re-engineered to become a PC gfx chip.
Gigapixel had working silicon when 3dfx bought them. 3dfx has first pass ramage parts back from the fab and had just started working on them when they were sold. Working parts even if they are still alpha/beta is not my def of theoretical.


I also dont think its a give that the NV30 is gonna to be that much faster than the R9700. The part can only be as fast as its slowest data path. ATI hand tweaked theirs. BD brought up lots of good points. Remember this chip has 120+ million transistors so its not the same thing as a straight die shrink from a GF4. Also the GF3 was clocked slower than the Ultra....
jbirney is offline   Reply With Quote

Old 08-18-02, 09:22 AM   #33
legion88
WhatIfSports.com Junkie
 
Join Date: Jul 2002
Posts: 135
Default

Quote:
Originally posted by jbirney


Gigapixel had working silicon when 3dfx bought them. 3dfx has first pass ramage parts back from the fab and had just started working on them when they were sold. Working parts even if they are still alpha/beta is not my def of theoretical.


I also dont think its a give that the NV30 is gonna to be that much faster than the R9700. The part can only be as fast as its slowest data path. ATI hand tweaked theirs. BD brought up lots of good points. Remember this chip has 120+ million transistors so its not the same thing as a straight die shrink from a GF4. Also the GF3 was clocked slower than the Ultra....
Gigapixel had more than just "working silicon". They were actually showing off their GP1s and GP2s to the media. Few websites had articles on them and they were even commenting on the FSAA quality.

The same can not be said with Rampage.
legion88 is offline   Reply With Quote
Old 08-18-02, 02:09 PM   #34
jbirney
Registered User
 
jbirney's Avatar
 
Join Date: Jul 2002
Posts: 1,430
Default

Thanks for the info Rosco. I was not sure how far Gigipixels products were thats why I took a safe stance and just said working silicon
jbirney is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


Similar Threads
Thread Thread Starter Forum Replies Last Post
December 2012 OpenGL status News Archived News Items 0 12-30-12 10:20 AM
NV30 not taped out yet? Megatron Rumor Mill 35 08-01-02 04:19 PM

All times are GMT -5. The time now is 11:50 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.