GloFo on track for interposers/stacking in 2015

[NTMBK]

http://semiaccurate.com/2013/04/02/global-foundries-talks-about-tsvs-on-20nm/#.UVrd5hzvusY

Should be very nice. Haswell GT3e will show what you can achieve with a fat connection to memory on an interposer, and this will be a big help for other graphics parts (GPUs, APUs).

 

[mrmt]

http://semiaccurate.com/2013/04/02/global-foundries-talks-about-tsvs-on-20nm/#.UVrd5hzvusY

Should be very nice. Haswell GT3e will show what you can achieve with a fat connection to memory on an interposer, and this will be a big help for other graphics parts (GPUs, APUs).

Globalfoundries and "on track" in the same sentence? That's April fool's day joke.

 

[SocketF]

Globalfoundries and "on track" in the same sentence? That's April fool's day joke.

No the news is from the 2nd of April.
For the first moment I shared your idea, but if you read until the end of the article:

full qualification for general use by next year and non-early adopter production in 2015 should bring some amazing device capabilities.

It sounds credible. It is not happening now just "sometime" in the future ^^

 

[mrmt]

No the news is from the 2nd of April.
For the first moment I shared your idea, but if you read until the end of the article:
[/FONT]

It sounds credible. It is not happening now just "sometime" in the future ^^

Charlie has some very good sources for raw info, analysis is his weak point. When talking about Nvidia, he bashes too much the company for their engineering but does not go too deep into marketing, financials, anything, and then he arrives in a much bleaker picture than it is in reality. When talking about AMD, he does the opposite, he doesn't bash enough the company for their engineering and disregard marketing, financials. In the end he paints a much milder picture than in reality.

But when talking about GLF, he goes press release. Given GLF atrocious track record it is funny to read about TSV and finfet in 2015. That's too much for a foundry that is 3 years late with a 28nm bulk node and screwed up their 32nm node. And he knows it. Nobody with the amount of internal info about AMD he has would cheer up for GLF the way he does.

 

[Gideon]

http://semiaccurate.com/2013/04/02/global-foundries-talks-about-tsvs-on-20nm/#.UVrd5hzvusY

Should be very nice. Haswell GT3e will show what you can achieve with a fat connection to memory on an interposer, and this will be a big help for other graphics parts (GPUs, APUs).

They just have to be late to every party, don't they ... Especially if you consider their previous way too optimistic dates for process nodes.

Let's see if by that time AMD is still designing anything more interesting, than VIA Nano was in 2008.

 

[Piroko]

When talking about AMD, he does the opposite, he doesn't bash enough the company for their engineering and disregard marketing, financials. In the end he paints a much milder picture than in reality.

Oh he's started to bash AMD quite a lot about six months back for firing so many engineers. But over the following months he noted that their leaks and press releses got more organized and streamlined. It's an interesting read if you analyze his emotions/reasoning in the articles and forums.

 

[NTMBK]

I should just rename this thread to "GloFo on track for a severe mocking by AT forum users", shouldn't I?

 

[Arachnotronic]

Charlie has some very good sources for raw info, analysis is his weak point. When talking about Nvidia, he bashes too much the company for their engineering but does not go too deep into marketing, financials, anything, and then he arrives in a much bleaker picture than it is in reality. When talking about AMD, he does the opposite, he doesn't bash enough the company for their engineering and disregard marketing, financials. In the end he paints a much milder picture than in reality.

But when talking about GLF, he goes press release. Given GLF atrocious track record it is funny to read about TSV and finfet in 2015. That's too much for a foundry that is 3 years late with a 28nm bulk node and screwed up their 32nm node. And he knows it. Nobody with the amount of internal info about AMD he has would cheer up for GLF the way he does.

Charlie appears to be back on AMD payroll, but not nearly as blatantly as the guys at BSN appear to be.

 

[ViRGE]

I should just rename this thread to "GloFo on track for a severe mocking by AT forum users", shouldn't I?

Probably.

We'd like to see GloFo do well, but they haven't exactly presented a facade of competence over the last half-decade. So whenever they make these half-baked announcements it's hard to take them seriously.

 

[cytg111]

Hope they pull it off

 

[OVerLoRDI]

Globalfoundries and "on track" in the same sentence? That's April fool's day joke.

You beat me too it