bit-tech.net

High-end Fermi will ship this year

High-end Fermi will ship this year

A senior official at Nvidia has said that they're confident GeForce products based on Fermi will ship this year.

A senior official inside Nvidia has said that they're "confident" that the high-end GF100 Fermi-based graphics cards will be released this year.

Nvidia's top executives have tried to remain tight-lipped about its release plans for its next-generation GeForce products, refusing to even talk about the graphics capabilities of the new Fermi chip in any great detail.

The execs have continually stated that the first cards will launch "on Tuesday", but weren't even ready to commit to a Q4 2009 or Q1 2010 launch officially. This is because the GPU is still in the bring up phase, which is where the engineering team checks for bugs and begins the process of productisation, which involves working out potential clock speeds, writing and optimising drivers and more.

It's good to hear some positive news about Fermi after the visual stunts on stage during the opening keynote.

There's no word on pricing or specific availability dates, but we'll keep digging for more. I've got quite a lot of information to decipher and sift through after many hours of probing Nvidia execs for more intricate Fermi architectural details. We'll be publishing a more detailed look at the new architecture (complete with some educated speculation on the graphics architecture based on the hints I've had this week) on bit-tech very soon.

Discuss in the forums.

14 Comments

Discuss in the forums Reply
Goty 3rd October 2009, 02:39 Quote
I would be extremely surprised if NVIDIA is actually able to get cards to consumers before 2010.
Kyocera 3rd October 2009, 06:56 Quote
I hope you'll look into the PC usage of such GPU/CPU hybrid since it's without sense to buy a hybrid CPU/GPU solely for rendering frame.
Such can be done cheaper, more efficiently and simpler with ATI approach.
Are we to expect Fermi based motherboards,without x86 processors?
p3n 3rd October 2009, 09:38 Quote
Quote:
Originally Posted by Goty
I would be extremely surprised if NVIDIA is actually able to get cards to consumers before 2010.

They aren't going to sell that many at £800 each...
Paradigm Shifter 3rd October 2009, 13:22 Quote
There might be workstation-class cards available, as they're ridiculously expensive so actual volume sold is low, which would allow nVidia to say, "Look, cards are in the market" but consumer-grade cards before next year? I'm doubtful.

Not to say I won't be happy if they do get them out, but I smell paper launch with possibly a few cards going to review sites, then no real availability for end-users until late January/February next year.
AshT 3rd October 2009, 17:31 Quote
This and Tims previous writeup of Nvidia shenanigans make me think Nvidia are desperate for that one-up on the new ATI cards. It happens everytime one of them bring their cards to market, but this time I think Nvidia are maybe a little too desperate and getting it wrong?
gavomatic57 3rd October 2009, 18:10 Quote
One thing that can be said for Nvidia is that they are timely - I'd imagine consumer Fermi parts will arrive when there is a need for them. Redesigning G80/G92 to support DX10.1 would have been a huge waste of time and money when you consider the number of titles that came out to make use of it. DX11 is coming but the titles that will use it releasing in the next 6 months you can count on a pigeon's foot. Nvidia will have learned from DX10 and the G80 that releasing parts before there are titles to test them on is foolhardy. I'd rather they took their time and got it right.
chumbucket843 3rd October 2009, 23:08 Quote
it will be here this year. in very limited quantities. almost like the gtx 295 launch.
HourBeforeDawn 4th October 2009, 00:14 Quote
all I know is if they didnt work on lowering the power requirements like for example what AMD did with the ATI 5870 then no thanks. Im tired of inefficient power hungry heaters ~_~
Tim S 5th October 2009, 04:23 Quote
Quote:
Originally Posted by HourBeforeDawn
all I know is if they did work on lowering the power requirements like for example what AMD did with the ATI 5870 then no thanks. Im tired of inefficient power hungry heaters ~_~

The 285's idle board power is the same as the 5870's, so they're pretty efficient at idle (i.e. most of the time). See here: http://www.bit-tech.net/hardware/graphics/2009/09/30/ati-radeon-hd-5870-architecture-analysis/17
HourBeforeDawn 5th October 2009, 04:26 Quote
Quote:
Originally Posted by Tim S
The 285's idle board power is the same as the 5870's, so they're pretty efficient at idle (i.e. most of the time). See here: http://www.bit-tech.net/hardware/graphics/2009/09/30/ati-radeon-hd-5870-architecture-analysis/17

who cares about idle, its not to hard to make something use less power when not under load, its load that Im more concerned about which causes the need for higher wattage power supplies, thats what really counts on my books, dont get me wrong idle is important too but its the overall size of the psu needed at load that Im more concerned about.

Which bye the way the 285 has a horrible load compared to the 5870
Tim S 5th October 2009, 06:11 Quote
Did you read the text underneath the graph?
HourBeforeDawn 5th October 2009, 06:16 Quote
Quote:
Originally Posted by Tim S
Did you read the text underneath the graph?

ur point? Im sure under furmark it would push a card past its designed operating spec/normal load conditions, regardless even still it draws less then the 285, so what did the 285 draw at that particular scene in Vantage?
Elton 14th October 2009, 05:19 Quote
If it does, I'll eat my shoes.

By the way, more often than not the card is idle, you don't spend every living moment gaming you know.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums