"The chip is unworkable, unmanufacturable, and unfixable. "
unfixable:
a. cannot be repaired, uncorrectable, incorrigible
It's pretty clear what he meant by that.
unworkable:
a. Not workable, especially not capable of being put into practice successfully; not practicable
It's being put into practice, despite Charlie's objections.
Your argument is flawed.Who gives a crap what he says.
Let the benchmark show how it is.
No one buys a high end card and gives a rats ass about any of this blabbering about. No one cares about power consumption, price, etc when buying a high end card. The want performance.
The one that really is funny is people crying about power consumption. You plan on putting it into a outdated computer? Worried about spending a extra couple bucks out of a paycheck? sheesh.
Your argument is flawed.
High power consumption often means lots of heat.
Cooling a hot chip means potentially noisy fan profiles.
Try listening to an HD5850 on 50% fan...its loud and that is a cool running chip.Not how you want your "high end" PC to sound lets face it.
Your argument is flawed. You plan on having some noise when buying a high end card.
Same goes for power requirements, you plan on it.
Again, you don't buy a $500+ card caring much for sound.
Meh, not trying to be his savior but that's still not the way I read it. By unmanufacturable I interpreted as being too costly to mass produce efficiently, unfixable and unworkable meant there was no way to correct the chip to be more manufacturable/increase yields significantly.
I don't know if you're just trying to hate Charlie or I'm trying to like him, but I do think most of his doom and gloom is overblown Nvidia spite, that's just not the way I read that article. There
Yeah, you're right... no one has ever spent hundreds of dollars liquid cooling an expensive computer because they care about the sound....
If this is what passes for logic these days... stay in school kids..
Right, cause most those people do that for the sound and not the overclocking the plan on doing on the expensive rigs. /rolls eyes.
Right, cause most those people do that for the sound and not the overclocking the plan on doing on the expensive rigs. /rolls eyes.
No one buys a high end card and gives a rats ass about any of this blabbering about. No one cares about power consumption, price, etc when buying a high end card. The want performance.
You want top-end performance, not caring about price or power consumption? Buy a 5970. The 480 is not going to be faster.
But I believe most people do care about price, and a significant number about power consumption or extra features (DX11, PhysX, Eyefinity, 3D Vision etc.).
Anyway this thread is not to discuss whether the 480 is worth buying, but only whether he is credible enough to believe the next time he writes an article. For example, the GF108 article said August. If he's credible on GF100, then we can believe that date. If not, then not. It allows us to know Nvidia's plans a few months to a year in advance.
This is the important one, and he was wrong.
"Nvidia GF100 pulls 280W and is unmanufacturable (January 17, 2010)"
To me he was saying there will be no Fermi. After this statement, the rest don't matter.
Your argument is flawed. You plan on having some noise when buying a high end card.
Same goes for power requirements, you plan on it.
Again, you don't buy a $500+ card caring much for sound.
C'mon, be real, he never said that. He obviously said there will be no Fermi in its current form (512 SP, that is, in January) which turned out to be true, as usual.
This guy always had plenty of spooks at every level (TSMC, OEM, distributors etc) and lately he might even got someone inside NV as well, some of his hints suggest.
You know better than that. He parrots the FUD that is sent to him from certain quarters, come on, you were at B3D long enough to know who feeds him and why :\
You know better than that. He parrots the FUD that is sent to him from certain quarters, come on, you were at B3D long enough to know who feeds him and why :\
Even if he's iffy on the details he was pretty much spot on with release from the beginning. His first articla in may stated at least two respins likely which would put a Q1 '10 release hard to reach
Step away from the KB and go think about what you're postingRight, cause most those people do that for the sound and not the overclocking the plan on doing on the expensive rigs. /rolls eyes.
CONCLUSION
He got a lot of it right, a lot more than you would expect an uninformed journalist to get at the time he wrote.
Performance figures were off, possibly because his samples had different specs to launch, but he should have revised them if he knew that.
Power requirements were overstated by him. GF100 was hot, but not disastrously so.
And yields/availibility won't be known until launch or even a few weeks after when initial supply runs out.
There no information about the delays in articles, so his claims of 3 steppings and die sizes couldn't be verified.