480 GTX: 275 watt TDP?

[jvroig]

On the forums, he is actually reasonable. Look through his post history. He rarely brings up Nvidia unless the conversation is about it, his arguments are well structured and well written (not personal attacks and informal like articles) and he will respond to most questions asked with explanations and evidence.

I think he deliberately puts on the 'controversial' style for the articles, to get views. It's a shame though, because it makes him look like a raving fanboy and weakens his credibility.

I share your impressions (for both articles and forum posts) and opinion as to why he puts on the controversial style for his articles. I believe I actually have put that opinion forth in a thread a few months ago.

Unfortunately, his controversial style is naturally double-edged - it gets views so it is good for his tabloid-style news site, but it also makes him a clear target of loathing/hate. While I may understand his likely motivations for doing so (he is, after all, making a living), I also understand the reactions of people he likely offends with his controversial headlines. Even his best forum post is not likely to change those people's minds once the damage is done.

No, he isn't perfectly reasonable. He can't even post a positive ATI only blog without cutting Nvidia some sludge. Look at the latest "ATI moves to improve dev relations" or whatever it's called. He needs help. Treatment of some sort. Definitely a bit of sickness there IMHO. And apparently, he will absolutely NEVER get over what Nvidia did to him.

He is trying to sell stories. If he were completely reasonable, I doubt he would stand out from the rest, and I guess he would pretty much be boring - his headlines would not be "shocking" or argumentative, and people will just pass on it. Not the best way to make a living. He is selling stories that way, and it looks to me like it is working. We all know to take it with a grain of salt, and yet a lot of us still give it a peek every now and then. It's not like Charlie is the only one doing such a thing. Pretty much most of the news (newspapers, online, TV) get the "controversial" treatment, to make it more palatable for people, and get views/subscribers.

Well it looks like ATI has won this round.

By default, yes, perhaps that can be said. As Anand mentioned in the Cypress story, you can lose by default if you don't show up to the fight, and that's pretty much what happened to nVidia (Win 7 launch and the holidays). But this is just a 6-month period. Since everything continues on, one round isn't so important as the pattern that emerges. If ATi continues winning round after round, then it will start to look bad for nVidia. But if all that happens is that they alternate rounds, then it is just the same old story and it will take a lot more things happening (good or bad) to either of them before the status quo changes.

 

[Soleron]

NV chips in the last few generation have been big, expensive, power hungry and difficult to manufacture.

Power hungry? GT200 used less power at load and idle than RV770, despite the bigger size.

--

@jvroig

Yes. I ignore the sensationalism and take note of the specific claims. If he's wrong on more than a few counts this time then he won't be worth reading in future. I'm not reading him to confirm some irrational bias I have, I'm reading him to get the truth when no one else is prepared to say it (Bumpgate, fake Fermi, GTS250 review issues).

 

[solofly]

A pig?lol (but that's nothing really new nv being a pig)

 

[SlowSpyder]

No, he isn't perfectly reasonable. He can't even post a positive ATI only blog without cutting Nvidia some sludge. Look at the latest "ATI moves to improve dev relations" or whatever it's called. He needs help. Treatment of some sort. Definitely a bit of sickness there IMHO. And apparently, he will absolutely NEVER get over what Nvidia did to him.

He still gets his jabs in, but he seems to have toned it down a bit lately. I wonder if he's read some comments about him, or his Nvidia-hate style of writing might be hurting his website more than helping it.

 

[ronnn]

Good news for Charlie that Nv pr still finds profit in this game. I guess it is a red herring of some sort.

I am sure the 480 will not use 275 watts - as you could make toast with the left over heat.

 

[Voo]

I wonder if he's read some comments about him, or his Nvidia-hate style of writing might be hurting his website more than helping it.

Not in a hundred years. The discussions on the whole net show how much hits he gets with every sensational headline. And I'd assume that Nvidia partners probably aren't his marketing partners

 

[ronnn]

Well they are certainly acting like partners. Nothing like a bit of internal vitro to encourage exposure.

 

[Soleron]

I am sure the 480 will not use 275 watts - as you could make toast with the left over heat.

TDP is not the same as power consumption. Typical load numbers may be close to TDP, or much lower (think Phenom vs. Phenom II 125W parts). What TDP matters for is OEM design wins, Tesla design wins, and PCIe power limits (can't exceed 300W, so AMD are free to up clocks by 10% on Cypress [a potential '5890'] and Nvidia can't counter).

 

[Headfoot]

He has maintained for the past 6 months that Fermi was a huge die, expensive, difficult to manufacture, very power hungry, will be late, will be available in extremely limited quantities when it does finally launch. He was the very first to mention all of this. Virtually nobody believed him.

So far he's got a lot more right than wrong. He's 5 watts off on his TDP prediction.

Is that so?

Please show me the link to a 3rd party site like Anandtech, Hardocp, Guru3d, TweakTown etc confirming me all of this info credibly.

Until you can do that it's speculation and we don't know whether he is right or wrong.
==================================================

No one has "won" the round yet. In order to "win" the round the round must be over so you can look back and see what happened. What I'm trying to say is that we wont be able to determine who wins this round until the Radeon 6800 and GTX 500 series are out.

If we use the logic of "first one to the party always wins" then the Xbox 360 won the console war before the Wii even came out, despite the Wii selling massively more until just recently.

Now I doubt that will happen with Fermi, but I'm just saying lets not jump the gun and declare a winner so early.

 

[WelshBloke]

Is that so?

Please show me the link to a 3rd party site like Anandtech, Hardocp, Guru3d, TweakTown etc confirming me all of this info credibly.

Until you can do that it's speculation and we don't know whether he is right or wrong.
==================================================

No one has "won" the round yet. In order to "win" the round the round must be over so you can look back and see what happened. What I'm trying to say is that we wont be able to determine who wins this round until the Radeon 6800 and GTX 500 series are out.

If we use the logic of "first one to the party always wins" then the Xbox 360 won the console war before the Wii even came out, despite the Wii selling massively more until just recently.

Now I doubt that will happen with Fermi, but I'm just saying lets not jump the gun and declare a winner so early.

Are you saying any of those predictions are wrong?

Or are you against him speculating? Because if we arnt allowed to speculate about unrealeased hardware tech forums would be pretty barren.


Consoles have a much longer shelf life than video cards. The problem with being late with video cards is your competitors refresh/next gen cards may clash with your late cards.

Its safe to say AMD has been 'winning' for the last 6 months if you like to look at things that way.

 

[MarcVenice]

Are you saying any of those predictions are wrong?

Or are you against him speculating? Because if we arnt allowed to speculate about unrealeased hardware tech forums would be pretty barren.


Consoles have a much longer shelf life than video cards. The problem with being late with video cards is your competitors refresh/next gen cards may clash with your late cards.

Its safe to say AMD has been 'winning' for the last 6 months if you like to look at things that way.

In all honesty, you would have to explain what winning this round actually means? Selling the most videocards? Making the most profit, as a whole company, or the most profit from just selling discrete videocards? Etc etc. We can't declare a winner, if there is no finish line ...

 

[WelshBloke]

In all honesty, you would have to explain what winning this round actually means? Selling the most videocards? Making the most profit, as a whole company, or the most profit from just selling discrete videocards? Etc etc. We can't declare a winner, if there is no finish line ...

I agree, I think the whole winning/losing metric is bullshit in this context.

I would say that the only valid 'goal post' to end users would be performance or performance/$ if you did want to measure that way.

 

[dguy6789]

Winning for us definitely means who has the most performance per dollar at the most price levels.

 

[SolMiester]

Yeah it is, a TDP of 188W for the 5870 vs a potential 275W for the Gtx 480 is a electricity bill/power consumption caveat.

BUT, you would need twice as many 5870 to get the compute power of fermi!

 

[happy medium]

Winning for us definitely means who has the most performance per dollar at the most price levels.

Is that a company winning or consumer winning?

I think market share and profits are a win for a company.

Price/performance is a win for us.

 

[WelshBloke]

Is that a company winning or consumer winning?

I think market share and profits are a win for a company.

Price/performance is a win for us.

Consumer, as long as they (the company) are around to sell us new shiny shit why should we care how they are?

 

[Keysplayr]

Wait, you don't like what he's saying, don't like some people actually give him some credit, so you try to dismiss him by claiming he's mentally ill? You a shrink?

An unhealthy hatred of anything doesn't necessarily mean mental illness. No, sorry I'm not a shrink. Did you need one?

 

[arkcom]

An unhealthy hatred of anything doesn't necessarily mean mental illness. No, sorry I'm not a shrink. Did you need one?

You said it:

He needs help. Treatment of some sort. Definitely a bit of sickness there IMHO.

 

[Keysplayr]

An unhealthy hatred of anything doesn't necessarily mean mental illness. No, sorry I'm not a shrink. Did you need one?

You said it:

He needs help. Treatment of some sort. Definitely a bit of sickness there IMHO.

Point? Maybe he is nuts, I have no idea. But I know this hate campaign isn't really what you would call "normal". Yoga might help.

I'm pretty sure that you see that he runs amok with contempt in his blogs. He could get his points across perfectly well without all that crap. But, since it is "semi-accurate", he needs to go for shock value to get any hits. So externally, he seems to have quite a large problem with Nvidia. Actually, Large is an understatement.

Understand something folks, whether Charlie is right or wrong isn't the problem I have with him, it's the way he comes across. He is 100% biased against NV and 100% pro AMD. If he ever has anything negative to say about AMD, he does it very mildly and throws in a comparison of Nvidia and how many levels of magnitude more evil they are.
He calls Jen Hsun Huang "Dear Leader". What's up with that? I'll tell you. It's a total lack of respect and a total presence of contempt.

If Charlie hadn't blown his NDA privileges just for the sake of getting web hits, I think he would be a much different person right now. IMHO.

 

[BenSkywalker]

He has maintained for the past 6 months that Fermi was a huge die, expensive, difficult to manufacture, very power hungry

Four years from now nVidia will be releasing a chip with a huge die that is expensive, difficult to manufacture and very power hungry. Four years ago nVidia was making chips with a huge die, that were expensive and difficult to manufacture along with being very power hungry. There is a bit of a trend.

will be late, will be available in extremely limited quantities when it does finally launch.

The limited quantities at launch would be shocking, I mean, everyone knows that 5850 had mass availability for everyone from launch day... oh wait, what's that? They have been so supply constrained the SRP has gone up by ~$40? Clearly there is absolutely nothing alike between ATi buying 40nm parts from TSMC and nVidia buying 40nm parts from TSMC. It must be something 'wrong' with nV's part, anyone that can agree to that must certainly agree to something being 'wrong' with ATi's part as they obviously fell way short of demand also.

Pretty much, Charlie has said that water is wet and Fermi will be late, nothing else he has stated has held any merit from all information available to date.

What he has gotten wrong about Fermi- initial tape out, projected clock speeds, Oak Ridge Labs and shader configuration(relating to him being wrong about yields) that we know for certain. The rest time will tell, but so far out of everything that required any sort of projection to get accurate Charlie is 1:4.

What TDP matters for is OEM design wins, Tesla design wins

As a general comparison, the market Tesla competes in even if it were 375Watts TDP that would pale next to some of the KWatt chips in the segment- it would still be very low on a watt/flop basis, likely still the best available for the HPC market.

 

[dguy6789]

Four years from now nVidia will be releasing a chip with a huge die that is expensive, difficult to manufacture and very power hungry. Four years ago nVidia was making chips with a huge die, that were expensive and difficult to manufacture along with being very power hungry. There is a bit of a trend.

Because Fermi is normal in terms of high expense, power consumption, and manufacturing difficulty with what is a regular day for Nvidia

Fermi isn't a trend, it's a turning point. Nvidia is running into the exact same problem Intel ran into with Prescott.

 

[tweakboy]

Awwww man that is not good. I yall sure its 245watts it eats up.

Looks like a lot of people will jump on the Fermi bandwagon only to find out their 550 watt PSU is not enough and is giving problems. hmmmmmmmm

 

[frostedflakes]

Prescott was bad arch. I don't think that's the case with Fermi. Seems like a good architecture, nVidia just got too ambitious and crammed way too many transistors on a 40nm die. Performance is going to be insane because of this, but the downsides are higher cost, power consumption, etc. Should be great once they shrink it to 28nm, though.

 

[ronnn]

BUT, you would need twice as many 5870 to get the compute power of fermi!

Yep I am pretty sure that all those stating games didn't need the power of the 5870 will suddenly find a need for compute power.

 

[SolMiester]

Yep I am pretty sure that all those stating games didn't need the power of the 5870 will suddenly find a need for compute power.

Did you read my post out of context?...the poster I was replying to was stating twice the wattage usage, whereas I was simply stating for the same compute power, you would need twice the ati cards...nothing to do with gamers!