Cell yields 'horrible'

[nonamer]

I agree but i don't know why anyone would characterize yields on the 65nm process for any chip as 'horrible', given how early it is in the migration to that process. So while they may be sensationalizing the news, i dont see where they could have even gotten such a quote in regards to 65nm?

The quote could be completely out of context.

 

[Shifty Geezer]

So while they may be sensationalizing the news, i dont see where they could have even gotten such a quote in regards to 65nm?

I can't say for sure, but one example might be an 'insider' said they were getting 20% yields when talking about the 65nm process, at which point Inq say Cell yields are horrible and Sony are doomed, without appreciating that's the normal yield this early in development of the process (if it is. I don't know squat about yields!). Without the source and context, and with the Inq track record, it's hard for me to think this applies to current 90nm Cell production.

 

[macabre]

Or it is about 90nm and they dug out a really old quote somewhere and posted it as news, not unusual for that site.

 

[ShootMyMonkey]

Or it is about 90nm and they dug out a really old quote somewhere and posted it as news, not unusual for that site.

Not out of the question considering they had a much older article about Cell yields being not so great very early on... back when we were seeing the earliest die photos and package photos. But usually Groo isn't the type to make a mistake like that. Can't speak for the others on the site, though.

 

[NavNucST3]

Traditionally schmaditionally.

IBM is a huge semiconductor manufacturer, what you're doing is exaggerating and extrapolating a few examples that have made the rounds in computing media and trying to establish some kind of pattern from it. That's not valid reasoning.

Or it could be from my own experience, take your issues somewhere else. I don't need to rely on the computing media.

 

[nonamer]

The bigger issue with this story is it's reasoning. Exactly what is causing the "horrible" yields? If it's the SPEs then they would have known for well over a year. If it's the PPE then apparently Xbox 360 is powered by magic as they got three of them running at the same speed with little problem. If its anything like a memory controller then it would be some timing or signaling issue that should be easily solved. The process itself is not agressive at all, but rather the same stuff used in the Xbox 360's CPU, AMD's Opterons, etc. so the given justification is unbelievable too. There's absolutely no plausible reason why this story is true and occam razor's states that its probably the article itself that is wrong. The only mystery is how they got it wrong.

 

[rounin]

The bigger issue with this story is it's reasoning.

No, the biggest issue with this story is the source. /thread.

 

[NavNucST3]

The bigger issue with this story is it's reasoning. Exactly what is causing the "horrible" yields? If it's the SPEs then they would have known for well over a year. If it's the PPE then apparently Xbox 360 is powered by magic as they got three of them running at the same speed with little problem. If its anything like a memory controller then it would be some timing or signaling issue that should be easily solved. The process itself is not agressive at all, but rather the same stuff used in the Xbox 360's CPU, AMD's Opterons, etc. so the given justification is unbelievable too. There's absolutely no plausible reason why this story is true and occam razor's states that its probably the article itself that is wrong. The only mystery is how they got it wrong.

While I can't say I disagree it should be noted that Ockham is typically as misrepresented as Moores "Law"

 

[predicate]

The bigger issue with this story is it's reasoning. Exactly what is causing the "horrible" yields? If it's the SPEs then they would have known for well over a year. If it's the PPE then apparently Xbox 360 is powered by magic as they got three of them running at the same speed with little problem. If its anything like a memory controller then it would be some timing or signaling issue that should be easily solved. The process itself is not agressive at all, but rather the same stuff used in the Xbox 360's CPU, AMD's Opterons, etc. so the given justification is unbelievable too. There's absolutely no plausible reason why this story is true and occam razor's states that its probably the article itself that is wrong. The only mystery is how they got it wrong.

That's a pretty serious misuse of Occam's Razor. It's for shaving off unnecessary parts of theories, not determining the legitimacy of something :S

[edit]Also, you're not dispelling assumptions, you're discounting the source.

 

[nonamer]

That's a pretty serious misuse of Occam's Razor. It's for shaving off unnecessary parts of theories, not determining the legitimacy of something :S

[edit]Also, you're not dispelling assumptions, you're discounting the source.

I'm trying to distill the most likely reason for the story. If it is true, then a series of implausible events at Sony/IBM would need to occur, whereas if the story is false, then only the writer made a mistake. I see this as a perfectly reasonable application of Occam's razor.

 

[Carl B]

I'm trying to distill the most likely reason for the story. If it is true, then a series of implausible events at Sony/IBM would need to occur, whereas if the story is false, then only the writer made a mistake. I see this as a perfectly reasonable application of Occam's razor.

It can be true without being unreasonable if the article is about 65nm vs 90nm production. The article certainly falls short in fleshing out the story, but at the same tiem the 'new methodologies' and 'test wafers' would point to the line they're getting ready to rooll out rather than the one they roled out a couple of years ago.

 

[PSman]

http://www.theinquirer.net/?article=31443

LOL!, i just stop reading right there

Guys, theinquirer is the same site that started the whole "RSX is weaker than the G70", and that Gears of War will come to the PS3

 

[rounin]

http://www.theinquirer.net/?article=31443

LOL!, i just stop reading right there

Guys, theinquirer is the same site that started the whole "RSX is weaker than the G70", and that Gears of War will come to the PS3

In fact maybe we should take this as a sign that CELL WILL in fact be launching with very very good yields at 65nm

 

[overclocked]

The first thing i thought of when reading that "article" was, "yeahh!" they go with 65nm and dont care about getting bad yields.

Ok back to reality my inital response without thinking was BS, and that is nothing i going to defend

 

[_xxx_]

Or it could be from my own experience, take your issues somewhere else. I don't need to rely on the computing media.

And your experience would be...? You ordered batches of chips from IBM and got to supervise the yields during the production?

 

[London Geezer]

Ok, not that i'm taking the article seriously, but i take it this is about 65nm... If they're having trouble with 90nm, then something's gone seriously wrong down the line...