Nvidia's mainstream Fermi parts arrive in late Q1

Nvidia's mainstream Fermi arrives in late Q1

http://img98.imageshack.us/img98/1849/geforce.jpg

Launching in March if all goes well

Quote:

---

Nvidia's mainstream market version of its GT300 / Fermi chip is scheduled for a Q1 2010 launch. This is not particularly good news for the company as ATI is already shipping its mainstream card based on the Juniper chip at pricing around €130 / $169 for the more expensive parts. Of course, we are talking about the HD 5770 and HD 5750 that both started shipping yesterday.

In perspective, Fermi will be released as a performance chip card, a high-end single chip card, and a dual-GPU card that might launch a few weeks later.

When it comes to entry level and mainstream segments, we will most likely have to wait until around March, if not even later. Despite the huge flexibility of Nvidia's Fermi GF100 architecture, it still takes time to derive slower performing chips from the original design.

We have a feeling that ATI might capture a much bigger piece of the DirectX 11 market than anyone expected due to being first to the market and shipping a month before its competition. Both of these factors are very beneficial for raking in good sales numbers. While Nvidia’s mainstream cards might end up faster, they will unfortunately come a month behind the competition.

---

Source

God we need a separate area for all this stuff

http://www.xtremesystems.org/forums/...50#post4062750

Whats this "a month" chat? If March is true, ATI have 6!!! MONTHS of unrivalled DX11 sales.

ATI... if somehow you read this.... DO NOT DARE **** away this lead. Most companies would kill for the advantage over a rival you have right now


EDIT: Its mainstream Fermi. Not Fermi of any kind..... so not a 6 month lead at all. I'll leave this post up so people can point and laugh at me if they want

He was talking about mainstream - Cedar and Redwood. They're Q1 I think.

Nope, Juniper is also mainstream. Midrange actually.

159 and 129$ are mainstream prices.

Cedar and Redwood are low end.

So ati does have 6 months lead in the DX11 mainstream market

Is anyone really surprised?

Quote:

---

Despite the huge flexibility of Nvidia's Fermi GF100 architecture, it still takes time to derive slower performing chips from the original design.

---

that statement makes no sense and is contradicting with itself...
if gt300 sports HUGE flexibility... then why does it take so long to cut it down?
then what would you call atis rv8xx? it sports MONSTROUS flexibility? cause they could launch fully fledged and cut down parts within weeks from each other... not a quarter "BEST CASE" scenario...

im really curious why it takes so long...
i cant believe that its an engineering thing, it looks more like bad planning and/or execution...

Quote:

---

Originally Posted by saaya

that statement makes no sense and is contradicting with itself...
if gt300 sports HUGE flexibility... then why does it take so long to cut it down?
then what would you call atis rv8xx? it sports MONSTROUS flexibility? cause they could launch fully fledged and cut down parts within weeks from each other... not a quarter "BEST CASE" scenario...

im really curious why it takes so long...
i cant believe that its an engineering thing, it looks more like bad planning and/or execution...

---

That's not the only reason. Another reason is they want more people to buy the more expensive high end part. GT200 is very flexible as well. Look at the GT216 (used in GT220) with its ~100mm^2 die size and DX10.1 support.

anyone else notice he said we should see a fermi x2 within a few weeks of fermi's release?

Quote:

---

Fermi will be released as a performance chip card, a high-end single chip card, and a dual-GPU card that might launch a few weeks later

---

100% not possible.

Quote:

---

Originally Posted by Manicdan

anyone else notice he said we should see a fermi x2 within a few weeks of fermi's release?

100% not possible.

---

It is still way too late because Fermi itself is late.

Quote:

---

Originally Posted by saaya

that statement makes no sense and is contradicting with itself...
if gt300 sports HUGE flexibility... then why does it take so long to cut it down?

---

yeah, really... What's the definition of "huge flexibility" anyway, if it takes too long to flex it, how come it's hugely flexible??

Quote:

---

Originally Posted by saaya

im really curious why it takes so long...
i cant believe that its an engineering thing, it looks more like bad planning and/or execution...

---

It is an engineering/design thing...
I wish we knew more info about how exactly they design the lower end chips. I have a few theories but nothing more than that.

Quote:

---

Originally Posted by Manicdan

anyone else notice he said we should see a fermi x2 within a few weeks of fermi's release?



100% not possible.

---

How is it not possible to take two performance chips and slap them both on a single PCB? We've already seen it done three times (GX2s, GTX 295s).

Quote:

---

Originally Posted by AuDioFreaK39

How is it not possible to take two performance chips and slap them both on a single PCB? We've already seen it done three times (GX2s, GTX 295s).

---

PCI-E power limitation, for instance.

Quote:

---

Originally Posted by Piotrsama

PCI-E power limitation, for instance.

---

thats what i was thinking. im expecting this to use 400W unless massively underclocked

Quote:

---

Originally Posted by AuDioFreaK39

How is it not possible to take two performance chips and slap them both on a single PCB? We've already seen it done three times (GX2s, GTX 295s).

---

weren't the gx2's dual pcb, not single?

Quote:

---

Originally Posted by Flanman

weren't the gx2's dual pcb, not single?

---

The latest dual chip card is single-PCB, but that's not the point.

Quote:

---

Originally Posted by AuDioFreaK39

The latest dual chip card is single-PCB, but that's not the point.

---

i realize that much but you claimed it has already been done three times, using the gx2's as example. was just pointing it out.

what was the most power hungry single card that was turned into a duel? and should the g300 be higher or lower in tdp than that card?

Well now the question becomes: will this delay 5870X2? Not that there's that much going on in the gaming market by Q1's end.

Quote:

---

Originally Posted by 003

That's not the only reason. Another reason is they want more people to buy the more expensive high end part. GT200 is very flexible as well. Look at the GT216 (used in GT220) with its ~100mm^2 die size and DX10.1 support.

---

yeah, very flexible... it only took them 14 months to come up with GT216 after GT200 was launched... thats amazingly flexible! :D

seriously, bad example man... :D

Quote:

---

Originally Posted by AuDioFreaK39

The latest dual chip card is single-PCB, but that's not the point.

---

The point if anything is all of Nvidia's dual gpu cards came out much later than the launch of the initial architecture ( eg G80 then G92 - 9800X2 then GT200 - GT200b to GTX295 ) If they truly wanted to offer a dual gpu card early on they *could* but using the current process it would be crippled and running at highly reduced clocks which is less than optimal ( plus that approach leaves no room to counter the competition if they decided to do a refresh themselves ) I fully expect Nvidia to market the living crap out of CUDA like never before given they won't be able to claim the performance crown until a dual gpu card is feasible. If you can't beat the competition, just convince everyone that your product is better... for reasons they likely don't need to care about XD

I am expecting a potential November paper launch with late December availability *best* case ( more likely January though )

Still, they took way longer to scale down GT200, so if this is true, that is good news for them. G92 has lasted them the longest of any architecture ever but they can't beat that dead horse any longer.

Quote:

---

Originally Posted by RagzaroK

Well now the question becomes: will this delay 5870X2? Not that there's that much going on in the gaming market by Q1's end.

---

Highly doubt it. It would be a smart move to take advantage of the lead ATi has at the moment. Why save your enthusiast chip until your competitor releases theirs? Use the time to attract impatient enthusiasts from the green side to potentially increase the amount of loyal customers.

Atleast that's how I see it.

Quote:

---

Originally Posted by saaya

yeah, very flexible... it only took them 14 months to come up with GT216 after GT200 was launched... thats amazingly flexible! :D

seriously, bad example man... :D

---

Uhh no not really. It's a great example. Nvidia really hasn't had a need to shrink down the GT200 because of G92, but when the OEMs wanted DX10.1 support to have on paper, it was easier to rework GT200 to support it.

Quote:

---

Originally Posted by 003

Uhh no not really. It's a great example. Nvidia really hasn't had a need to shrink down the GT200 because of G92, but when the OEMs wanted DX10.1 support to have on paper, it was easier to rework GT200 to support it.

---

right, but oems wanted that last christmas or maybe in early 2009, and its not like they only told them then, they probably told nvidia in spring 2008 that they want 10.1 in winter 2008 or spring 2009... but it took nvidia until NOW to finally get it done. so again, how is that flexible?

how can you say a product that took more than a year from concept to retail is a good example of how flexible the architecture its based on is? 2 or less quarters is already not exactly flexible... 1 quarter or less is kinda flexible... 2 months or less is VERY flexible...

or what, are you claiming nvidia only started working on gt2xx 6 months ago? :rolleyes: sure...

you cant really compare gt200 and gt2xx with rv870 and rv840 though, cause rv840 is a cut down rv870, gt2xx is a reworked gt200... they added a new imc, dx10.1 and probably tweaked something else... thats obviously going to take a lot more work and time to do... the problem is that nvidia hasnt done any cut down chips in ages, they always did reworks... they always changed the architecture and didnt just chop some parts of, which takes a lot more time... a chopped down gt200 might be possible within 2 months like ati did it, yes, gt200 MIGHT very well be very flexible... but how could we know? even nvidia cant know cause they never even tried it afaik... so how can you say gt200 is flexible if they never created a cut down part of it and brought it to market quickly... they did a very slow rework... drawing a conclusion from that to how fast they are at cutting down a chip by chopping some bits off is impossible... but if anything youd think itll take them a while as their rework wasnt exactly a fast one...

this might be the main issue they are facing actually, as that dragged their entire roadmap along and slowed everything down...
instead of cutting a chip down and THEN reworking/refreshing it, they always did both at the same time... they basically tried to do a tick and tock at the same time... and thats really stupid... both intel and amd learned a while ago that you really dont want tick and tock at the same time...

Quote:

---

Originally Posted by saaya

right, but oems wanted that last christmas or maybe in early 2009, and its not like they only told them then, they probably told nvidia in spring 2008 that they want 10.1 in winter 2008 or spring 2009... but it took nvidia until NOW to finally get it done. so again, how is that flexible?

how can you say a product that took more than a year from concept to retail is a good example of how flexible the architecture its based on is? 2 or less quarters is already not exactly flexible... 1 quarter or less is kinda flexible... 2 months or less is VERY flexible...

or what, are you claiming nvidia only started working on gt2xx 6 months ago? :rolleyes: sure...

you cant really compare gt200 and gt2xx with rv870 and rv840 though, cause rv840 is a cut down rv870, gt2xx is a reworked gt200... they added a new imc, dx10.1 and probably tweaked something else... thats obviously going to take a lot more work and time to do... the problem is that nvidia hasnt done any cut down chips in ages, they always did reworks... they always changed the architecture and didnt just chop some parts of, which takes a lot more time... a chopped down gt200 might be possible within 2 months like ati did it, yes, gt200 MIGHT very well be very flexible... but how could we know? even nvidia cant know cause they never even tried it afaik... so how can you say gt200 is flexible if they never created a cut down part of it and brought it to market quickly... they did a very slow rework... drawing a conclusion from that to how fast they are at cutting down a chip by chopping some bits off is impossible... but if anything youd think itll take them a while as their rework wasnt exactly a fast one...

this might be the main issue they are facing actually, as that dragged their entire roadmap along and slowed everything down...
instead of cutting a chip down and THEN reworking/refreshing it, they always did both at the same time... they basically tried to do a tick and tock at the same time... and thats really stupid... both intel and amd learned a while ago that you really dont want tick and tock at the same time...

---

The main point here is that they spent a lot of time with figuring out how to scale down the GT200 and have created Fermi in such a manor that it will be very easy to scale down, and Nvidia has confirmed this in either some press report or an interview, I forgot where.

But point being is they have learned much from GT200 and applied what was learned to Fermi.

Quote:

---

Originally Posted by 003

But point being is they have learned much from GT200

---

Yes.

Quote:

---

Originally Posted by 003

and applied what was learned to Fermi.

---

No.

I think fermi is too soon to apply 'what you learnt'... it was already either designed as scalable, or it wasnt. It'll be at least GT400, possibly 500, before the lesson learnt shows up in the architecture.

People seem to consider dev times so selectively around here.

Quote:

---

Originally Posted by 003

The main point here is that they spent a lot of time with figuring out how to scale down the GT200 and have created Fermi in such a manor that it will be very easy to scale down, and Nvidia has confirmed this in either some press report or an interview, I forgot where.

---

We'll know how easy Fermi will actually scale down by how long it will take Nvidia to have value parts based on Fermi.

And I declare that Fermi should be now known as Hermi, until it's actually available.