The GT300/Fermi Thread - Part 2!

Quote:

---

Originally Posted by Jamesrt2004

if that happens then its VERY bad for nvidia as it will be nearly a whole gen behind :/ ATI could potentially have 28nm cards while fermi's still on 40nm

and it doesnt help that TSMC are thinking about changing to 28nm now aswell as im sure nvidia would have sorted their plans for a die shrink to be at 32nm.

we shall see.. nvidia will be fine.. but it still isnt good to be so late :/



(though tbh 6 series wont come out till Q1 2011 (OUR 2011))

---

Anand stated this in his recent article on Evergreen:
"The Northern Islands GPUs, due out later this year, were surely designed before anyone knew how RV870 would play out. Much less whether or not Fermi/GF100 would be this late."

Maybe Northern Islands have been pushed back since. But I get the impression from Anand's article that ATI is really schedule-oriented these days (at the sacrifice of possible features).

Cypress was released in Sept 2009. If Northern Islands get out on, say, Nov 2010, that would be a 14 month release cycle. Not exactly a stretch.

Quote:

---

Originally Posted by ***Deimos***

http://vr-zone.com/forums/561413/nvidia-quot-fermi-quot-gpus-delayed-yet-again.html

Sounds a lot like a few thousand A3 GPU boards will be shipped to consumers as is (March),
while they re-spin again in hopes of fixing whatever is causing all the trouble.

Its just rumour, but Charlie may be right about "hope TSMC fabs perfect huge chip vs respin adding redudancy like double via to make chip even bigger".
Very tough situation indeed. But unlike G200 which went with conservative 65nm, there is no "55nm" equivalent half-node shrink.. 40nm is already that!

One way or another nVidia needs to fix/work-around 40nm issues till end of '10.

---

Hahaha, so I was correct then! Any bets on the price of these cards in may? :p:
This is going to be ASUS MARS 2.0! :D

Quote:

---

Originally Posted by ***Deimos***

SSAO is Screen Space Ambient Occlusion. It is implemented entirely on GPU as pixel shader for games like Crysis. Its an approximation to Global illumination. It allows the ambient light sources and colored surfaces to contribute to the lighting of an object (NOT specular reflections) that may otherwise be in shadow. ie when lights are on in closed washroom, everything outside washroom door is a little less dark because of light reflecting off the walls.

nVidia has a driver option for this if I recall.

---

Yep. Nice option, actually! :up: I used it all the time with my 8800.

lets just look at those official nvidia replies from last week to charlies article about fermi being delayed to Q2 again :D

Quote:

---

Originally Posted by mapel110

Reply from nvidia.
http://twitter.com/RS_Borsti
Oh Charlie... that just another hilarious post

http://twitter.com/Igor_Stanek
Oh Charlie... that just another hilarious post me: I think with this post Charlie totally destroyed his credibility :)

---

sounds like nvidia PR doesnt have a clue whats actually going on :D
so much for credibility huh igor? :lol:

Quote:

---

Originally Posted by saaya

lets just look at those official nvidia replies from last week to charlies article about fermi being delayed to Q2 again :D


sounds like nvidia PR doesnt have a clue whats actually going on :D
so much for credibility huh igor? :lol:

---

its actually rediculous.. it actually is really starting to pi$$ me off... ok I understand to a degree that nvidia have to keep hopes up on the card... but just say.. "we ARE having problems.. we are fixing them as soon as we can if you wait it'll be WELL worth it, but like we said you have to wait and see why..;) "

now if they said that your like LAME its delayed etc.. etc.. but then intreaged that why they are saying the wait will be worth it and to me would atleast make me wait to see what they are hiding...

all this crap with even their PR flaming someone.. whos actually saying mostly the truth (like i say normally charlie is right to a certain extent) .. is just plain wrong.. stop saying the card will be out the next month.. and say we dont know but what we do know... is that you will LOVE it.

seriously... I should have the marketing job... lol

Quote:

---

Originally Posted by Jamesrt2004

its actually rediculous.. it actually is really starting to pi$$ me off... ok I understand to a degree that nvidia have to keep hopes up on the card... but just say.. "we ARE having problems.. we are fixing them as soon as we can if you wait it'll be WELL worth it, but like we said you have to wait and see why..;) "

---

They can't, Charlie will write like he's on fire once they confirm he's correct! Besides, his info HAS to stay semi-accurate! :D
Besides, they make people hold off their ATI purchases that way, not everyone is following the rumours and release dates so closely.

might as well load up Fermi Thread Part 3

All I know is I am giving them a little longer and if there cards are not out I am going to get an ATI.

Quote:

---

Originally Posted by HuffPCair

All I know is I am giving them a little longer and if there cards are not out I am going to get an ATI.

---

+1

But fermi is suppose to be released "paper" around first week of March and stock may come by around mid march :shrug:

I am waiting for fermi because they may drive the prices down quite a bit and i may be able to afford a 5970 with a custom water block :up:

Otherwise i'll get either a 5870 with 2GB ram "Either a eyefinity edition or whatever" or if its good enough a GTX 470!!

Quote:

---

Originally Posted by zalbard

They can't, Charlie will write like he's on fire once they confirm he's correct! Besides, his info HAS to stay semi-accurate! :D
Besides, they make people hold off their ATI purchases that way, not everyone is following the rumours and release dates so closely.

---

I guess just anoyed :( lol :p:

All i know is, IF the report is true, that Fermi wide availability would finally come in Q2 '10, any XSys forum proponents/supporters of waiting game should apologize or make a new suggestion/recommendation of waiting for ATi's next gen GPU family. You know who YOU are. :rolleyes:

Quote:

---

Originally Posted by spursindonesia

All i know is, IF the report is true, that Fermi wide availability would finally come in Q2 '10, any XSys forum proponents/supporters of waiting game should apologize or make a new suggestion/recommendation of waiting for ATi's next gen GPU family. You know who YOU are. :rolleyes:

---

why? it's still wise to wait.. its safe to assume when fermi hits ati will drop prices.. even if you have to wait even 4/5 more months then the price drop will warrant the wait in my and alot of people eyes.. i got burned last time when i brought a 4870 for it to drop like £80 a few weeks later.

Quote:

---

Originally Posted by generics_user

nvidia is going to get owned by ati in the mainstream market as they have no competitors at all in the 100-200$ range even after fermi is released; what makes things worse is that they seem to prepare high-end fermis only which may open the gap in their lineup from 100-200$ to 100-300$ :eek:

i thought that fermir was late but things seem to be much worse than expected; this thing has been delayed from windows 7 launch timeframe to may = 8 months....

---

It seems Nvidia has learned from their last mid-range mistake; the lack of a midrange GT200 series cards. As of the last Fermi info brief, it seems Fermi is built to be much easier to cut down then its previous generation GPU's.

Time to twiddle our thumbs, save up our money, and wait.

Quote:

---

Originally Posted by ajaidev

+1
But fermi is suppose to be released "paper" around first week of March and stock may come by around mid march :shrug:

---

I think you missed what we posted on previous page, but real availability seem to have slipped to May-July.

Quote:

---

Originally Posted by Jamesrt2004

why? it's still wise to wait.. its safe to assume when fermi hits ati will drop prices.. even if you have to wait even 4/5 more months then the price drop will warrant the wait in my and alot of people eyes.. i got burned last time when i brought a 4870 for it to drop like £80 a few weeks later.

---

That's not his point (I think), he's talking about the guys saying: "Don't buy ATI's high end, wait for Nvidia's". (Regardless of price).

Quote:

---

Originally Posted by Piotrsama

I think you missed what we posted on previous page, but real availability seem to have slipped to May-July.



That's not his point (I think), he's talking about the guys saying: "Don't buy ATI's high end, wait for Nvidia's". (Regardless of price).

---

lol i posted that link and i do know that mass availability may come in May-July but very limited availability can be expected around March-April just so Nvidia can say that things are shipping even if it is in limited quantity.

So what do you guys think? Better to get a HD5xxx or wait for Fermi? :confused:

you waited till now probably wont be such problem to wait few more weeks :)

Quote:

---

Originally Posted by iTravis

So what do you guys think? Better to get a HD5xxx or wait for Fermi? :confused:

---

Depends on if you're a PhysX / 3DVision fan...

Quote:

---

Originally Posted by iTravis

So what do you guys think? Better to get a HD5xxx or wait for Fermi? :confused:

---

In the tech world, if you wait for the next best thing, you might as well wait forever :)

Buy it if you need/want something badly enough. Unless a real KILLER feature is around the corner, there's little point in waiting for anything.

Quote:

---

Originally Posted by spursindonesia

All i know is, IF the report is true, that Fermi wide availability would finally come in Q2 '10, any XSys forum proponents/supporters of waiting game should apologize or make a new suggestion/recommendation of waiting for ATi's next gen GPU family. You know who YOU are. :rolleyes:

---

Yeah, I remember that discussion. This was back in early December I believe, and his premise was that since it would be the only card to play Crysis at the highest settings, people should wait for it since their current cards were good enough anyway. Since Fermi would be released in January, it was a no brainer to wait... and wait.... and wait... and wait................
My reason for upgrading is Eyefinity.

Quote:

---

Originally Posted by OutSider

you waited till now probably wont be such problem to wait few more weeks :)

---

lol that has been said since november :ROTF:

Wait till March if you are going with ATi you maybe able to get custon 5870's and that 5870 2GB EyeEdition or else if GTX 470 is good enough get that instead !!

Quote:

---

Originally Posted by spursindonesia

All i know is, IF the report is true, that Fermi wide availability would finally come in Q2 '10, any XSys forum proponents/supporters of waiting game should apologize or make a new suggestion/recommendation of waiting for ATi's next gen GPU family. You know who YOU are. :rolleyes:

---

I hear ya on this one; if this is true I am going to scrap any thoughts of going Fermi; that I can promise.

Quote:

---

Originally Posted by Jamesrt2004

why? it's still wise to wait.. its safe to assume when fermi hits ati will drop prices.. even if you have to wait even 4/5 more months then the price drop will warrant the wait in my and alot of people eyes.. i got burned last time when i brought a 4870 for it to drop like £80 a few weeks later.

---

Quote:

---

Originally Posted by Piotrsama

.............
That's not his point (I think), he's talking about the guys saying: "Don't buy ATI's high end, wait for Nvidia's". (Regardless of price).

---

Sorry, haven't made my intention clear on that post, but Piotrsama explained it well. :)

I'm just so pissed, after NV30 and R600 debacles in the past, some of us still haven't learned any lesson at all. :down:

There is no reason to believe that fermi will be as weak as an R600 or an NV30.

[BSN] nVidia to launch GeForce GTX 470/480 on March 27, 2010?

Twitter GeForce

Quote:

---

One more full quarter delay and suddenly "Fermi" will be competing with AMD's upcoming "Northern Islands" GPU generation...

---

from vrzone : http://vr-zone.com/forums/561413/nvi...yet-again.html

does that sound right? i knoe that fermi is going to be fast but delayed a full 2quarters? i was hoping it was going to be out in march:confused:

It's written by a forum member of VR-Zone, not VR-Zone itself ;)

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

Quote:

---

Originally Posted by ajaidev

+1

But fermi is suppose to be released "paper" around first week of March and stock may come by around mid march :shrug:

I am waiting for fermi because they may drive the prices down quite a bit and i may be able to afford a 5970 with a custom water block :up:

Otherwise i'll get either a 5870 with 2GB ram "Either a eyefinity edition or whatever" or if its good enough a GTX 470!!

---

Yeah that is my thought process, but it just seems they keep stretching this crap. Sometimes I even wonder if they are even telling the truth. Nvidia just better not be lying to us LOL.

Quote:

---

Originally Posted by wez3570

Anyone wanna guess!

---

GT350 :banana:

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

They probably are going to release information on how their next GPU is going to be awesome again LOL

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

I hope this is it, if not im buying two new 5970s.

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

The codename has changed. From now its going to be called Farti: lots of air and no substance.

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

Come on 40nm GT200 for midrange! Or G92 on 40nm.

Quote:

---

Originally Posted by damha

The codename has changed. From now its going to be called Farti: lots of air and no substance.

---

Rrrrooooo ... :ROTF: however ...

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

I know, I know! G92 on 40nm with software DirectX 11 patched in?! :D

Quote:

---

Originally Posted by wez3570

Nvidia have just tweeted....

ARE YOU READY?
Check back here on Monday morning at 9:00 a.m. PST
for a major announcement!

Anyone wanna guess!

---

G92 quad on a single card :yepp: will be called GTX FU and will come in a dark brown color... FTW "Fck the wrld" :D

Doesn't sound like their pre-announcement of a facebook product announcement is going so well. Instead of excitement and anticipation, they are mostly being mocked.

And btw, it just fricking facebook, why pre-announce at all. If you have something to announce, just do it. It's not like some big gathering or meeting where you need time to fly people in, sheesh.

I don't get it either. Well, there's a big fuss about Twitter, Facebook and whatnot. But when a company wants to release something awesome they have to issue a press release in order to get it on the front page of hardware-related sites.

I don't think there will be anything interesting. Remember that "big" announcement EVGA had on halloween?

Dont keep much hope for 22nd whatever will be will be on march 1st week :)

i am betting on tegra cr_p or may be ion 2 or Jens babydoll

Just to clarify, for those lazy to read through 500 posts, nVidia likely will still launch Fermi as is in March as expected. But, very few cards until yield problems fixed.. May-July.


*****************************8

international standard node: 180nm -> 130nm -> 90nm -> 65nm -> 45nm -> 32nm
ie what CPUs are fabbed on.

half-node for 180nm is slight optimizations keeping everything else constant (to make it easy).. ie 150nm (9700pro). 150/180 = ~85%, for shrinks some 40-100 parameters need to be scaled in different ways, but you can get away with only ~15% change (as opposed to whole step of ~30%).
110nm (7800GTX), 80nm (R600), 55nm (RV670), 40nm (RV740). Its used for big chips trying to cram things into smaller die.. as you can imagine at 300mm2+, GPUs are BIG chips.

for TSMC, next node is 32nm.. half node its based on is 28nm.
http://en.wikipedia.org/wiki/32_nanometer
FYI TSMC uses both immersion (wet) and double patterning (55nm/2=28nm) for 32/28nm. It **IS** going to cost even more than 40nm, and be even more risky.

btw found this same ad 'are you ready' guess for which generation it was used

http://www.nvidia.com/content/areyouready/index.html

Quote:

---

Originally Posted by eric66

btw found this same ad 'are you ready' guess for which generation it was used

http://www.nvidia.com/content/areyouready/index.html

---

That's good, it means it will be very likely used as a major GPU announcement.

Quote:

---

Originally Posted by eric66

btw found this same ad 'are you ready' guess for which generation it was used

http://www.nvidia.com/content/areyouready/index.html

---

:rofl::rofl::rofl:

"are you ready" I guess would be used for half of all product launches, but this is still funny :D

Quote:

---

Originally Posted by eric66

btw found this same ad 'are you ready' guess for which generation it was used

http://www.nvidia.com/content/areyouready/index.html

---

http://www.nvidia.com/content/areyou...eyouready.html

That was 3 times the size of a Pentium 4 microprocessor the fermi is suppose to be about 3 billion "3000M and nehalem has what 731M, so that means fermi is more than 4 times the size of a Nehalem microprocessor :)

Prediction: Few cards in march = 480GTX with all the hype for just $999.

There we have a presentation to justify the price.

http://www.nvidia.com/object/fermi_architecture.html

To NVIDIA: we have been ready since Oct 2009.

Come Monday: "SURPRISE SURPRISE!! our cards are going to cost 1000 dollars lol"

Quote:

---

Originally Posted by Metroid

Prediction: Few cards in march = 480GTX with all the hype for just $999.

There we have a presentation to justify the price.

http://www.nvidia.com/object/fermi_architecture.html

---

That presentation says nothing about that price.

They are saying you can use Fermi to turn your PC to a supercomputer. Everybody with a bit respect for PC-technology should actually like it. I don't get why you are trying to make a negative point out of something that can be a great help to those who use their PC for more than just a game-console.

Quote:

---

Originally Posted by Sam_oslo

That presentation says nothing about that price.

They are saying you can use Fermi to turn your PC to a supercomputer. Everybody with a bit respect for PC-technology should actually like it. I don't get why you are trying to make a negative point out of something that can be a great help to those who use their PC for more than just a game-console.

---

That presentation was for Tesla, we already know the price for those, and it is not $999.

Yeah, when fermi was first announced there was all kinds of talk of turning your PC into a supercomputer, but I'm sure I remember reading about nv slashing the consumer card's DP capabilities. Was that just a rumor or?

Quote:

---

Originally Posted by flippin_waffles

Yeah, when fermi was first announced there was all kinds of talk of turning your PC into a supercomputer, but I'm sure I remember reading about nv slashing the consumer card's DP capabilities. Was that just a rumor or?

---

i do believe that's fact, keeps hpc customers from buying geforce products for dp calculation. makes sense, most gamers have no need for dp and it keeps the people with deep pockets buying the high margin stuff.

If nVidia Tesla market gets too big, there wont be any Fermi for gamers. Afterall, why sell for $600 to gamers, when you can charge $3000 in HPC.

While the Tesla line up will have half rate DP compared to SP Flop performance, i believe GeForce will only have one eight DP compared to SP Flop performance, the official news was there backthen. How they block that capability precisely, that hasn't been answered quite clearly.

The whole point is, few cards will be launched yes will be around $599 - $699 and that is being too optimistic. This will be like a super ultra soft launch, probably not even 10 thousand units. This move will be to calm down angry customers and Nvidia adepts, even though you will never find it but it was launched. Ebay prices, expect $1000+. Nvidia's primary business is where the money will be generated. What is the point of selling a product that costs $500 to manufacture at $699 if you can make more money selling Tesla based products at $2500.

Quote:

---

Originally Posted by ***Deimos***

If nVidia Tesla market gets too big, there wont be any Fermi for gamers. Afterall, why sell for $600 to gamers, when you can charge $3000 in HPC.

---

If that happens, you can be your ass that Nvidia will go build their own fab to build their own chips. With that kind of market, no bank would be insane enough to turn them down on a loan.

I hope you realise that building fab is not something cheap and also it does not end with just building it

Quote:

---

Originally Posted by ***Deimos***

If nVidia Tesla market gets too big, there wont be any Fermi for gamers. Afterall, why sell for $600 to gamers, when you can charge $3000 in HPC.

---

i think they will always sell consumer cards... always... if only to get rid of the busted tr4sh silicon :D
which is kinda what i expect for fermi actually... fully fledged cards will be tesla only, broken ones will be consumer cards... well, except for some limited supply of fully enabled cards for pr purposes maybe...

Quote:

---

Originally Posted by ***Deimos***

If nVidia Tesla market gets too big, there wont be any Fermi for gamers. Afterall, why sell for $600 to gamers, when you can charge $3000 in HPC.

---

please... economies of scale. :shakes:

Fermi gets cancelled

To be tweeted this Monday.

Quote:

---

Originally Posted by BenchZowner

Fermi gets cancelled

To be tweeted this Monday.

---

How crazy that would be :rofl:

Quote:

---

Originally Posted by BenchZowner

Fermi gets cancelled

To be tweeted this Monday.

---

:rofl:

Then 5850 will be around $500 and 5870 will be around $700. The 5970 will be available for $1200.

Quote:

---

Originally Posted by Metroid

The whole point is, few cards will be launched yes will be around $599 - $699 and that is being too optimistic. This will be like a super ultra soft launch, probably not even 10 thousand units. This move will be to calm down angry customers and Nvidia adepts, even though you will never find it but it was launched. Ebay prices, expect $1000+. Nvidia's primary business is where the money will be generated. What is the point of selling a product that costs $500 to manufacture at $699 if you can make more money selling Tesla based products at $2500.

---

The price will depend on the competition. If ATi can't (or don't want) to put up a good performance-fight, then a superior Fermi won't come cheap, for sure. But you are speculating based on what?

Tesla is something else, we are talking about CUDA that runs on all upcoming NV-GPUs. CUDA has already been used for supercomputer-activities, and is accelerating many Gflop-demanding tasks both for personal and business use. But my question (which you are avoiding to answer) was: how can you make a negative point out of something that can be a great help to those who use their PC for more than just a game-console.?

Quote:

---

Originally Posted by Sam_oslo

The price will depend on the competition. If ATi can't (or don't want) to put up a good performance-fight, then a superior Fermi won't come cheap, for sure. But you are speculating based on what?

Tesla is something else, we are talking about CUDA that runs on all upcoming NV-GPUs. CUDA has already been used for supercomputer-activities, and is accelerating many Gflop-demanding tasks both for personal and business use. But my question (which you are avoiding to answer) was: how can you make a negative point out of something that can be a great help to those who use their PC for more than just a game-console.?

---

Negative point where? Avoiding your question?

You better start reading posts slowly. That post I have written my prediction which is not based on anything, prediction can be personal no need to be based on anything, second when you make a product that is going to cost much more than what you were offering you need to justify the price. That is basically what Nvidia is doing, "marketing hype".

It might be a stellar supercomputer and so the price. Is this a negative point? I don't see it as.

Quote:

---

Originally Posted by Metroid

Negative point where? Avoiding your question?

You better start reading posts slowly. That post I have written my prediction which is not based on anything, prediction can be personal no need to be based on anything, second when you make a product that is going to cost much more than what you were offering you need to justify the price. That is basically what Nvidia is doing, "marketing hype".

It might be a stellar supercomputer and so the price. Is this a negative point? I don't see it as.

---

Don't get offensive. Try to hold the focus on the subject. Personal focus is childish and is not going to help you.

You are speculating that CUDA and supercomputer-performance is something negative, because it will make the GPU more expensive. All based on your wild guesses and speculations about the "yet-unknown" price.
Anybody with a couple av days marketing classes wold tell you that the competition between nVidia and ATi will impact the price. But you are bringing inn Tesla into this, and speculating that the competition between nVidia's own products will impact the price?

Quote:

---

Originally Posted by FischOderAal

How crazy that would be :rofl:

---

well NV can't let Intel to be a bigger shocker with their Larrabee canning! :p

The real problem we-the-army-of-cuda-elves face is cooling 4 dual GPU cards dangling off those slow old PCIe slots.

What we REALLY want Nvidia to do is to get GloFo to write quad-Fermis that drop into spare mobo sockets.

But looks like AMD may do this first... what's an apprentice elf to do :shrug:

Quote:

---

Originally Posted by nnunn

The real problem we-the-army-of-cuda-elves face is cooling 4 dual GPU cards dangling off those slow old PCIe slots.

What we REALLY want Nvidia to do is to get GloFo to write quad-Fermis that drop into spare mobo sockets.

But looks like AMD may do this first... what's an apprentice elf to do :shrug:

---

quad fermi? that wouldnt physically fit on a package of any cpu socket, even G34 :lol:
besides it being impossible to pay for, impossible to cool, and starving from memory bandwidth... what would be the point of this? more system bandwidth... but it wouldnt be able to use it since it would be completely starved memory bandwidth wise... :D

what we need is hyper transport instead of pciE... if qpi would be open then that would work too...

but just out of curiousity... why do you think those gpus are system bandwidth limited exactly?
silverstone had a quad 9800gx2 gpgpu demo at computex and they said it scaled linearly with every card they added... which means pciE bandwidth wasnt an issue at all as adding more cards results in some slots switching from 16x to 8x...

Quote:

---

Originally Posted by saaya

what we need is hyper transport instead of pciE...
if qpi would be open then that would work too...

---

Ok, that'll be fine for a next step--our stuff is still computationally bound and scales nicely on 1-8 GPUs. But adding double-width, 12" cards is sooo like using (12") cards to add a MB of extended ram to a 286. Just wondering what might be down the road... quad Fermis and NO intel inside? :shrug:

Quote:

---

Originally Posted by nnunn

Ok, that'll be fine for a next step--our stuff is still computationally bound and scales nicely on 1-8 GPUs. But adding double-width, 12" cards is sooo like using (12") cards to add a MB of extended ram to a 286. Just wondering what might be down the road... quad Fermis and NO intel inside? :shrug:

---

thats what nvidia wants... at least for tesla... :D
i dont think fermi successor will be there yet... it might do well with a slow basic cpu, but itll still need a cpu i think... but the one after that... in 2013+ might do just fine without a cpu altogether...

if you look at a tesla server today the cpu is not much more of one of the many chips on the mainboard that powers and connects the vgas, which are the server blades pretty much :)

Quote:

---

Originally Posted by Sam_oslo

Don't get offensive. Try to hold the focus on the subject. Personal focus is childish and is not going to help you.

You are speculating that CUDA and supercomputer-performance is something negative, because it will make the GPU more expensive. All based on your wild guesses and speculations about the "yet-unknown" price.
Anybody with a couple av days marketing classes wold tell you that the competition between nVidia and ATi will impact the price. But you are bringing inn Tesla into this, and speculating that the competition between nVidia's own products will impact the price?

---

I already answered all the questions.

Regards,

Quote:

---

Originally Posted by saaya

silverstone had a quad 9800gx2 gpgpu demo at computex and they said it scaled linearly with every card they added... which means pciE bandwidth wasnt an issue at all as adding more cards results in some slots switching from 16x to 8x...

---

I'm not so sure about that.

There are multiple constraints, and bandwidth is just one. #GPUs has to be >= #frames being rendered ahead. Hypothetically, if you had 100 GPUs running at 50fps, those 100 frames would be 2 sec of lag. And D3D and OpenGL have limitations on #chain/buffers.

There is also diminishing returns due to cost of partitioning work, and how uniform the work units are. Most games I play dont have exact same pixel/texture/vertex workload on all parts of the screen all the time.

Bottom line:
If you cant get anywhere near 50% let alone 100% scaling in many (not all) games with just 2 GPU, how do expect it to be possible for quad?

http://semiaccurate.com/2010/02/20/s...gtx480-scores/ :shocked:

"SemiAccurate gets some GTX480 scores"
http://www.semiaccurate.com/2010/02/...gtx480-scores/

EDIT: Whoops, already posted, shoulda refreshed the page :)

Quote:

---

The GTX480 with 512 shaders running at full speed, 600 or 625MHz depending on which source, ran on average 5% faster than a Cypress/HD5870, plus or minus a little bit. The sources were not allowed to test a GTX470 which is likely an admission that it will be slower than Cypress/5870.

There is one bright spot, and it is a very bright spot indeed. No, not the thermal cap of the chip, but the tessellation performance in Heaven. On that synthetic benchmark, the numbers were more than twice as fast as the HD5870/Cypress, and will likely beat a dual chip HD5970/Hemlock. The sources said that this lead was most definitely not reflected in any game or test they ran, it was only in tessellation limited situations where the shaders don't need to be used for 'real work'.

---

Well, this better not be true. I mean, great performance in 'over-the-top-tessellation-using-games' is nice, but when are these coming out? In 2012? Waiting till Monday. But it would be a shame since these cards wouldn't compete with Cypress well just because of the price...

I just read up to, "Small numbers of final cards have started to trickle in to Nvidia, and they are only showing it to people they consider very friendly for reasons that we will make clear in a bit."

And I'm already ROFL'ing here.

Oh man, Nvidia had better pray that these guys are pulling all of this out of there arse's. :rofl:

Not that I expect much after the whole delaying for all of eternity and then holding up a blatantly fake card in order to lie to investors.

It's just sad if this is true, if only for the fact that I wanted a monster card. Crysis with 16X AA @ 60+ FPS on a single card would probably look so nice. :shakes:

ran on average 5% faster than a Cypress/HD5870....

hooo boy. The 5870's will overclock well beyond 480 then

Quote:

---

Originally Posted by SemiAccurate

Now you know the raw clocks, how does it perform? It is a mixed bag, but basically it is much below their original expectations publicly stated as 60% faster than Cypress. The numbers that SemiAccurate were told span a variety of current games, all running at very high resolutions. Here is where we can't list specifics or the Nvidia Keystone Kops might find their first SemiAccurate mole. We will bring you the full spreadsheets when the cards are more widespread.

The GTX480 with 512 shaders running at full speed, 600 or 625MHz depending on which source, ran on average 5% faster than a Cypress/HD5870, plus or minus a little bit. The sources were not allowed to test a GTX470 which is likely an admission that it will be slower than Cypress/5870.

---

Highly doubt things are that bad.

Oh snap if true...... :(

Even the clocks are wrong. There are 4 Clockdomains. Chip/half hot clock/hot clock/vram

yea we believe you buddy...... lol

<CONSPIRACY_THEORY>there's always possibility that NVIDIA is feeding CD with false info to make him ridiculous and irrelevant news source after final launch?</CONSPIRACY_THEORY>

Quote:

---

Originally Posted by Nedjo

<CONSPIRACY_THEORY>there's always possibility that NVIDIA is feeding CD with false info to make him ridiculous and irrelevant news source after final launch?</CONSPIRACY_THEORY>

---

Haha it would certainly be an amusing turn of events.

No matter what I do though I just can't get the feeling out of my gut that fermi is the GeForce FX series all over again. :eek:

Quote:

---

Originally Posted by Aaronage

"SemiAccurate gets some GTX480 scores"
http://www.semiaccurate.com/2010/02/...gtx480-scores/

EDIT: Whoops, already posted, shoulda refreshed the page :)

---

Oh wow. SNAP.

I can kind of see how it makes sense though.
Its like nVidia fx vs 9700pro - tied in performance.. in DX8 games.

Vertex processing power in current GPUs has increased dramatically over last 10 years... but all games made nowadays are pretty much console ports. And XBOX360 and PS3 use GF7/X1900 type graphics...

Even "high end" GPUs selling today like GT240 and G92 score just "adequately" in Ozone3D 18million triangle, geometry instancing demo.

Code:

---

18 million triangles
8800GTX - 13fps
HD3870 - 25fps

8 million triangles
8800GTX - 28 fps
HD3870 - 33 fps

---

Perhaps this is why games keep the poly count down, and how/why Fermi is so much faster with vertex processing... GF8 was just slow at it to begin with.

Quote:

---

Both of our sources said that their cards were smoking hot, one said they measured it at 70C at idle on the 2D clock.

---

i think they are measuring in kelvin.:p: i dont see how idle power could be worse by this much when it consumes in worst case scenario 50% more power at peak.

Quote:

---

Originally Posted by Nedjo

<CONSPIRACY_THEORY>there's always possibility that NVIDIA is feeding CD with false info to make him ridiculous and irrelevant news source after final launch?</CONSPIRACY_THEORY>

---

maybe this is why charlie writes libelous articles about nvidia.

With this article, CD puts A LOT of his credibility, well, basically all of it, on jeopardy. Not just to Green camp fanbois, or the general mass, but even to people who has been skeptic toward nVidia's Fermi, like me. In my skepticism, i still see 512 SP's Fermi to be around 15-20 % faster overall than plain jane HD 5870, so this claim is just rather unbelievable to me at this moment. :shrug:

he has been right about a lot of things so far, EOL of GTX 200 series before fermi, no high end offerings, not able to drop prices to compete with the 48x0 series, late in production, problems with yields, problems with heat, etc...

Not an ATI fanboy here, I have a GTX 275, but fermi looks to be full of fail for gamers. the GPGPU will subsidize any losses in the gaming segment though.

Quote:

---

Originally Posted by ***Deimos***

Oh wow. SNAP.

I can kind of see how it makes sense though.
Its like nVidia fx vs 9700pro - tied in performance.. in DX8 games.

Vertex processing power in current GPUs has increased dramatically over last 10 years... but all games made nowadays are pretty much console ports. And XBOX360 and PS3 use GF7/X1900 type graphics...

Even "high end" GPUs selling today like GT240 and G92 score just "adequately" in Ozone3D 18million triangle, geometry instancing demo.

Code:

---

18 million triangles
8800GTX - 13fps
HD3870 - 25fps

8 million triangles
8800GTX - 28 fps
HD3870 - 33 fps

---

Perhaps this is why games keep the poly count down, and how/why Fermi is so much faster with vertex processing... GF8 was just slow at it to begin with.

---

vertex processing is cheap and is bound by vram and tri-setup. the 3870 is faster because the core clock is higher. most games have ~1million polys/frame so its not that important. triangle setup/rasterization/vertex culling is much faster on fermi than any gpu. if clocks on fermi are 600MHz then setup rate is 2.4 billion tris/sec. 5870 is 850 million tris/sec. this boost was added for tessellation. most triangles today cover 30 pixels most of the time. with tessellation micropolys can cover half a pixel.

Quote:

---

Originally Posted by zalbard

Well, this better not be true. I mean, great performance in 'over-the-top-tessellation-using-games' is nice, but when are these coming out? In 2012? Waiting till Monday. But it would be a shame since these cards wouldn't compete with Cypress well just because of the price...

---

it doesnt take time to get good at adding a lot of tessellation which is one of its advantages. all game devs have to do is remodel characters and there are already software tools for that. if you have a displacement map that has very small intricate detail it will require higher tessellation levels to be visible.

So now nvidia realizes tessellation is a good idea when ATI has been saying this for the past, what, 10 years?!

I don't know, Charlie hates a lot, but his GTS 250 hoopla back at the inquirer from a while ago really was an 'aha' moment.

Quote:

---

Originally Posted by Chumbucket843

i think they are measuring in kelvin.:p: i dont see how idle power could be worse by this much when it consumes in worst case scenario 50% more power at peak.


maybe this is why charlie writes libelous articles about nvidia.

---

People around here don't seem to be able to differentiate between power and temperature. An idle temperature of 70 may not mean much, since the 4850 idled really hot too. If this is all true, likely this is done to keep acoustics in check, with the screaming coming during load.

there's a good chance that "most games" means cpu limited games
it's sorta a convenient way to ignore crysis and the like.

Quote:

---

Originally Posted by grimREEFER

there's a good chance that "most games" means cpu limited games
it's sorta a convenient way to ignore crysis and the like.

---

That would be a creative use of the word most, especially considering most games are not CPU limited.

This is what I would like to see ...

GTX480 Ultra 245W

Code:

---

GTX285        HD5870        GΤΧ480        %Faster        Game
65,7        83,8        109,72        30,93        Far Cry 2
115        111,2        192,05        72,71        Batman Arkham Asylum
85,5        88,4        142,79        61,52        Tom Clancy's H.A.W.X
88,5        89,3        147,8        65,5        Resident Evil 5
60,3        65,4        100,7        53,98        World in Conflict
6376        7953        10649        33,89        3DMark Vantage Extreme GPU

---

Quote:

---

Originally Posted by cegras

People around here don't seem to be able to differentiate between power and temperature. An idle temperature of 70 may not mean much, since the 4850 idled really hot too. If this is all true, likely this is done to keep acoustics in check, with the screaming coming during load.

---

But when you read ideling @ 70C with Fan at 70% makes you think that his source is smoking something or sitting in a 50C ambient :rolleyes:

Quote:

---

Originally Posted by Marios

This is what I would like to see ...

GTX480 Ultra 245W

Code:

---

GTX285        HD5870        GΤΧ480        %Faster        Game
65,7        83,8        109,72        30,93        Far Cry 2
115        111,2        192,05        72,71        Batman Arkham Asylum
85,5        88,4        142,79        61,52        Tom Clancy's H.A.W.X
88,5        89,3        147,8        65,5        Resident Evil 5
60,3        65,4        100,7        53,98        World in Conflict
6376        7953        10649        33,89        3DMark Vantage Extreme GPU

---

---

I pray this is the real result, I want Fermi to be good so bad so we can get rid of this onesided nonsense.

Quote:

---

Originally Posted by ***Deimos***

I'm not so sure about that.

There are multiple constraints, and bandwidth is just one. #GPUs has to be >= #frames being rendered ahead. Hypothetically, if you had 100 GPUs running at 50fps, those 100 frames would be 2 sec of lag. And D3D and OpenGL have limitations on #chain/buffers.

There is also diminishing returns due to cost of partitioning work, and how uniform the work units are. Most games I play dont have exact same pixel/texture/vertex workload on all parts of the screen all the time.

Bottom line:
If you cant get anywhere near 50% let alone 100% scaling in many (not all) games with just 2 GPU, how do expect it to be possible for quad?

---

we were talking about gpgpu... the limitations for graphics are different... especially for sfr the system bus definately plays a big role, but for gpgpu each gpu works on independent tasks (i thought?)

Quote:

---

Originally Posted by cegras

I don't know, Charlie hates a lot, but his GTS 250 hoopla back at the inquirer from a while ago really was an 'aha' moment.

---

hmmm i must have missed that... what was it about?

Quote:

---

Originally Posted by Marios

This is what I would like to see ...

GTX480 Ultra 245W

Code:

---

GTX285        HD5870        GΤΧ480        %Faster        Game
65,7        83,8        109,72        30,93        Far Cry 2
115        111,2        192,05        72,71        Batman Arkham Asylum
85,5        88,4        142,79        61,52        Tom Clancy's H.A.W.X
88,5        89,3        147,8        65,5        Resident Evil 5
60,3        65,4        100,7        53,98        World in Conflict
6376        7953        10649        33,89        3DMark Vantage Extreme GPU

---

---

why is a 285 as fast as a 5870 in this table? 0_o

Quote:

---

Originally Posted by saaya

hmmm i must have missed that... what was it about?


why is a 285 as fast as a 5870 in this table? 0_o

---

When nvidia renamed the 9800GTX to the 250 GTS, but didn't send parts for preview/review to anandtech or hardocp as they knew they'd say something bad about it.

And the 285 is that speed because the table has been pulled directly out of the posters arse.

Quote:

---

Originally Posted by Iconyu

When nvidia renamed the 9800GTX to the 250 GTS, but didn't send parts for preview/review to anandtech or hardocp as they knew they'd say something bad about it.

And the 285 is that speed because the table has been pulled directly out of the posters arse.

---

Ah, I didn't realize that is what all the fuss was about...Thanks for filling me in :) .