The GT300/Fermi Thread - Part 2!

[570091D]

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...

actually, the gtx285 was easily able to compete with the 48x0 series... they were still selling well after the debut of the 5870 (for what little time they were both on sale at the same time). and as for charlie, someone said it here before "even a dead clock is right twice a day".

a few things in charlie's text that drew my attention:

Our sources have a bit of conflicting data

This speed is the first point the sources conflict on, and it could go either way, since both sources were adamant about theirs being the correct final clock. *sigh*

One said they measured it at 70C at idle on the 2D clock

We will bring you the full spreadsheets when the cards are more widespread

The GTX480 with 512 shaders running at full speed, 600Mhz or 625MHz depending on which source, ran on average 5 percent faster than a Cypress HD5870, plus or minus a little bit.

we have been hearing rumors of the card not having DX11 drivers on launch

GTX480 is too hot, too big, too slow, and lacks anything that would recommend it over a HD5870, much less the vastly faster HD5970. Nvidia is said to be producing only 5,000 to 8,000 of these cards, and it will lose money on each one. The architecture is wrong, and that is unfixable. The physical design is broken, and that is not fixable in any time frame that matters. When you don't have anything to show, spin. Nvidia is spinning faster than it ever has before.

oh god! the irony! IT HURTS MAH BRAIN!!!

in all seriousness, does anyone else feel like they've heard this before? maybe... the gtx280 is too hot, big, power hungry, and broken. maybe... the 4870 will stomp the gtx280 and (because of "bump-gate" + gt200 fail) nvidia will go out of business? this is all beyond that fact that his "sources" can't agree on much... find a new story to spin charlie, we've all heard this one before.


EDIT: the CRAZIEST thing just occurred to me, perhaps charlie writes these stories to spook investors. nvidia prices fall a bit, he buys a lot of nvidia stock, nvidia releases it's real product, stocks go back up, and charlie gets rich. it's crazy, but maybe...

[Piotrsama]

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?

Nvidia "cutted" objective sites from the GTS250 launch.

http://www.anandtech.com/video/showdoc.aspx?i=3523&p=2

Early last week Charlie over at The Inquirer posted a story saying that a number of reviewers were cut out of the GeForce GTS 250 launch. We felt a bit hurt, by the time the story launched we weren't even asked to be briefed about the GTS 250. Cards had already gone out to other reviewers but we weren't on any lists. Oh, pout.

Magically, a couple of days after Charlie's article we got invited to a NVIDIA briefing and we had a GTS 250 to test. Perhaps NVIDIA was simply uncharacteristically late in briefing us about a new GPU launch. Perhaps NVIDIA was afraid we'd point out that it was nothing more than a 9800 GTX+ that ran a little cooler. Or perhaps we haven't been positive enough about CUDA and PhysX and NVIDIA was trying to punish us.

Who knows what went on at NVIDIA prior to the launch, we're here to review the card, but for what it's worth - thank you Charlie

[eric66]

just because he hates nvidia doesn't mean he has crap sources. as proven above Charlie has some info about whats going on if that wasn't the case nobody would cared about him anyway

[mapel110]

just because he hates nvidia doesn't mean he has crap sources. as proven above Charlie has some info about whats going on if that wasn't the case nobody would cared about him anyway

The major source is AMD and he is getting paid by them.

[eric66]

may be or may be not. actually thats even more frightening for nvidia if amd supplies info means that amd has premium info about there actions

[LinkinParkBoy]

NVIDIA
SemiAccurate is sponsored by AMD. Everything they post is not even remotely close to accurate. Even Google doesn't refer to it as a real news site.

[Metroid]

"SemiAccurate is an U.S.-based technology news and opinion web site. It was created by Charlie Demerjian after his departure from The Inquirer. News organization IDG credited SemiAccurate [1] for first reporting the story on Sony warning its customers about faulty nVidia chips in some Vaio laptops.

Tech news site Ubergizmo has criticized a SemiAccurate story [2] which speculated on wheter Apple would ditch nVidia chips calling the news site "half-accurate"."

http://en.wikipedia.org/wiki/SemiAccurate

http://www.theinquirer.net/inquirer/...eapon-revealed

http://charliedemerjianisadouchebag.blogspot.com/

http://www.theinquirer.net/inquirer/...ipsets-history

http://www.theinquirer.net/inquirer/...laystation-gpu

http://www.theinquirer.net/category/...rlie-vs-nvidia

Charlie has just one purpose in life 'criticize Nvidia' and off course get paid for it. Though he rarely gets it right.

[zalbard]

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.

An idle temperature of 70C is a complete disaster. It either means that cooling is terrible or that the power saving techniques are nearly non-existent.

This is what I would like to see ...

GTX480 Ultra 245W

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

This table is very inaccurate, at least for 285-5870 part. Not to mention 50%+ advantages of GTX480, lol.

[annihilat0r]

lolz. Up to today Charlie was very big about Fermi getting castrated and the high end part getting 448 SPs. Now he says it's gonna be 512 SPs, of course without bothering to say he had been wrong all along.

[blindbox]

lolz. Up to today Charlie was very big about Fermi getting castrated and the high end part getting 448 SPs. Now he says it's gonna be 512 SPs, of course without bothering to say he had been wrong all along.

He did base that 448 SP assumption on nvidia's official Tesla PDFs though.

Time will tell, didn't I hear a May launch?
GPU wars have always been fun to watch.

[saaya]

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.

well, its not like it prevented anandtech or hardocp from bashing the 250

ive said it before tho, the 250 rename was "ok" imo as it was the same DX version and its actually pretty close to a 260 in many benchmarks

in all seriousness, does anyone else feel like they've heard this before? maybe... the gtx280 is too hot, big, power hungry, and broken. maybe... the 4870 will stomp the gtx280 and (because of "bump-gate" + gt200 fail) nvidia will go out of business? this is all beyond that fact that his "sources" can't agree on much... find a new story to spin charlie, we've all heard this one before.

well initially it was, wasnt it?
thats something i just thought about today... its weird how nvidia kept following the same strategy... with G80 it worked well, but even then it was hot and expensive to make (=not great yields)
they continued and with GT200 things got worse... they managed to improve yields to an ok to good level, but they never had great yields... if the yields would be great they wouldnt have needed a 275 but just would have lowered the price of the 285 or offered a part with all sps and lower clocks...
they continued with the same strategy that didnt work perfectly the first time and caused them some headaches the second time... at least some people at nvidia MUST have seen this coming... weird that they didnt change their strategy... i guess they thought they could get GT200 and G92 down to 40nm...

EDIT: the CRAZIEST thing just occurred to me, perhaps charlie writes these stories to spook investors. nvidia prices fall a bit, he buys a lot of nvidia stock, nvidia releases it's real product, stocks go back up, and charlie gets rich. it's crazy, but maybe...

you obviously never met charlie

just because he hates nvidia doesn't mean he has crap sources. as proven above Charlie has some info about whats going on if that wasn't the case nobody would cared about him anyway

he doesnt even hate nvidia... he hates jensen and some other top guys at nvidia... and for a good reason... he didnt always hate nvidia in case you didnt know... and he bashed ati several times in the past...

lolz. Up to today Charlie was very big about Fermi getting castrated and the high end part getting 448 SPs. Now he says it's gonna be 512 SPs, of course without bothering to say he had been wrong all along.

didnt he say that there will be 512sp cards but only very few?

[Marios]

...

This table is very inaccurate, at least for 285-5870 part. Not to mention 50%+ advantages of GTX480, lol.

Wait and see. GTX285 is accurate. HD5870 is with very first drivers just because this will be the case with GTX480ultra drivers.

http://www.tsmc.com/tsmcdotcom/PRLis...581&language=E
No more photo resistive remnants for TSMC? This means no vias problems any more?

[cegras]

An idle temperature of 70C is a complete disaster. It either means that cooling is terrible or that the power saving techniques are nearly non-existent.

This table is very inaccurate, at least for 285-5870 part. Not to mention 50%+ advantages of GTX480, lol.

No.

http://www.techpowerup.com/reviews/P...D_4850/24.html
http://www.anandtech.com/video/showdoc.aspx?i=3341&p=22
http://www.overclock.net/ati/348442-...ur-4850-a.html

You should educate yourself on the difference between heat and temperature.

To repeat myself, if idling at 70 is really true then that means the cooling solution is being throttled for the sake of acoustics. E.g. the maximum fan speed is determined by the minimum noise threshold. This implies that at that certain fan speed, the heatsink only has enough heat removal power to have the temperature at 70. Which further implies that screaming is inevitable when you start loading the card.

What you can infer from this is in fact the cooling solution have to be loud and heavy duty for when the card really starts eating juice. Which means the fan will not have the best acoustics, again implying (gosh, so much extrapolation) that it does have good low speed acoustics.

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

Acoustics, my friend.

"The problem is that this creates heat, and a lot of it. Both of our sources said that their cards were smoking hot. One said they measured it at 70C at idle on the 2D clock.

The fans were reported to be running at 70 percent of maximum when idling, a number that is far, far too high for normal use. Lets hope that this is just a BIOS tweaking issue, and the fans don't need to be run that fast. It would mean GF100 basically can't downvolt at all on idle. On the upside, if it's any comfort, the noise from the fans at that speed was said to be noticeable, but not annoying."

This is the dustbuster threshold for idle.

[BrowncoatGR]

http://www.tsmc.com/tsmcdotcom/PRListingNewsAction.do?action=detail&newsid=4581&l anguage=E[/url]
No more photo resistive remnants for TSMC? This means no vias problems any more?

Over the past several months TSMC has expanded its Maskless Lithography team and has been working with MAPPER engineers at Fab 12 to integrate electron beam direct write capabilities into manufacturing processes for development of future technology node

It's not about current technologies

[FischOderAal]

Wait and see. GTX285 is accurate. HD5870 is with very first drivers just because this will be the case with GTX480ultra drivers.

And what's the use of that? Fermi will have to compete with a matured HD 5870.

[Badfastbusa]

I am so ready to see what this thing is really going to do.

[Sushi Warrior]

Anything Charlie says take with a cup or 5 of salt - he speaks as his "sources" say and he never fabricates stuff but he quotes bad sources when they're knowingly wrong just to hurt Nvidia more. Again, he speaks as his sources say, but he ignores anything that is good about Nvidia cards.

Besides, Nvidia doesn't care much about Fermi - they made over $150 million last quarter IIRC. They're not a sinking ship, and they may have realized Gaming GPU's are not where all the money is at.

[Zed_X]

Tomorrow will be exciting date, for fans!

[saaya]

Anything Charlie says take with a cup or 5 of salt - he speaks as his "sources" say and he never fabricates stuff but he quotes bad sources when they're knowingly wrong just to hurt Nvidia more. Again, he speaks as his sources say, but he ignores anything that is good about Nvidia cards.

Besides, Nvidia doesn't care much about Fermi - they made over $150 million last quarter IIRC. They're not a sinking ship, and they may have realized Gaming GPU's are not where all the money is at.

so they will go the way of 3dlabs?

[zalbard]

No.
You should educate yourself on the difference between heat and temperature.

My 5870 idles at around 30C while being completely inaudible. Power saving features FTW (downclocking to 300MHz, 0.9V vCore, 21% fan speed). What's your point? lol.
4850 just isn't the best card in terms of PowerPlay management. The heatsink isn't the best one around either.
There is no excuse for a card to idle at such high temps.

[NaMcO]

And who told you those figures are final anyway. My 4870x2 also idles at supid temps and previous 4890 did the same. What's new

[Chumbucket843]

i think it is safe to assume that idle power is better than gt200 and the cooler is better too. the temperature of the card should not be 70C at idle. especially if you consider this article.
http://www.fudzilla.com/content/view/17723/1/

[zalbard]

I dunno who came up with 70C idle temps anyway. This is just such a ridiculous BS rumour it's not even funny.
I can see Fermi not being much faster than Cypress, but idling at 70C? LOL. What's next? 3 slot cooler? Nuclear reactor to power it? 40cm length?

[kadozer]

so they will go the way of 3dlabs?

nice sig

[xVeinx]

My 5870 idles at around 30C while being completely inaudible. Power saving features FTW (downclocking to 300MHz, 0.9V vCore, 21% fan speed). What's your point? lol.
4850 just isn't the best card in terms of PowerPlay management. The heatsink isn't the best one around either.
There is no excuse for a card to idle at such high temps.

My GTX 260 55nm idles at around 50C. Even with a new node, 3 billion transistors are going to produce heat. There are plenty of factors that play into how that heat is dissipated, resulting in the perceived temperature at any one point; even if true, it's not the end of the world. Assuming these figures will be true of released units, then the only real impact this will have is on overclockability (and perhaps lifespan, if the proper precautions haven't been taken).