AMD Analyst Day. No 32nm products til 2011.

[terrace215]

Samples in 2010, but 2011 for 32nm products.

http://phx.corporate-ir.net/phoenix....rol-analystday

Look at Rick Bergman's pdf presentation.


Still vague about Bulldozer timing, nothing more than "2011" (products) and "2010" (samples).

Possibly more during afternoon "breakout sessions"

[Helloworld_98]

lets hope AMD didn't plan on using GF to manufacture the 6800's.

[Oliverda]

Samples in 2010, but 2011 for 32nm products

It's not a fresh news...

[informal]

Samples in 2010, but 2011 for 32nm products.

http://phx.corporate-ir.net/phoenix....rol-analystday

Look at Rick Bergman's pdf presentation.


Still vague about Bulldozer timing, nothing more than "2011" (products) and "2010" (samples).

Possibly more during afternoon "breakout sessions"

It was 2011 since last analyst day,so nothing changed apparently.
BTW,bulldozer is now confirmed to be AM3 compatible ,a huge plus.

[FlanK3r]

And he said about Bulldozer "very highend" platform

[FlanK3r]

[Nedjo]

informations are dispersing over few threads (http://www.xtremesystems.org/forums/...=232784&page=6) (http://www.xtremesystems.org/forums/...d.php?t=238693)

I urge mods to create dedicated "AMD Analyst day 2009" thread and merg all other posts and threads in it!

[terrace215]

It's not a fresh news...

32nm products only in 2011 is a bit of news, at least if all this year you've been listening to AMD's misleading talk about their 32nm "ramping" in 2010. Of course, what they left deliberately vague was the fact that this "ramp" was not a production ramp, hoping folks would think they'd have 32nm PRODUCTS in 2010.

Then, at the last CC we got "deep in the back half of 2010", i.e. Q4, for production STARTS, which fits with 2011 32nm products.

So, at this point, 1 to 1.25 years behind Intel on 32nm revenue shipments, despite the bluster about "closing the gap" that they provide every couple years.

[informal]

32nm products only in 2011 is a bit of news, at least if all this year you've been listening to AMD's misleading talk about their 32nm "ramping" in 2010. Of course, what they left deliberately vague was the fact that this "ramp" was not a production ramp, hoping folks would think they'd have 32nm PRODUCTS in 2010.

Then, at the last CC we got "deep in the back half of 2010", i.e. Q4, for production STARTS, which fits with 2011 32nm products.

So, at this point, 1 to 1.25 years behind Intel on 32nm revenue shipments, despite the bluster about "closing the gap" that they provide every couple years.

Q3 production starts and by the end of 2010 shipment of products.

[informal]

Thuban six core(Leo platform) is now Q2 2010,a pull in again(they say H1 so it can be Q1 i guess,but unlikely).It was Q3 in last product roadmap 2 months ago.


They stated that 32nm SOI SHP yields will be ~50% by the year's end

Update on the BD cores,now i read it is 4x4way cores ,not 2x2way as dresdenboy speculated(shame on him /joke ).

[Oliverda]

32nm products only in 2011 is a bit of news, at least if all this year you've been listening to AMD's misleading talk about their 32nm "ramping" in 2010. Of course, what they left deliberately vague was the fact that this "ramp" was not a production ramp, hoping folks would think they'd have 32nm PRODUCTS in 2010.

Then, at the last CC we got "deep in the back half of 2010", i.e. Q4, for production STARTS, which fits with 2011 32nm products.

So, at this point, 1 to 1.25 years behind Intel on 32nm revenue shipments, despite the bluster about "closing the gap" that they provide every couple years.

Misleading? What kind of misleading are you talking about? The AMD has never told that they will have 32nm CPU in the market in 2010.

A slide from the previous year:



Good morning...

[Nedjo]

32nm products only in 2011 is a bit of news, at least if all this year you've been listening to AMD's misleading talk about their 32nm "ramping" in 2010. Of course, what they left deliberately vague was the fact that this "ramp" was not a production ramp, hoping folks would think they'd have 32nm PRODUCTS in 2010.

Then, at the last CC we got "deep in the back half of 2010", i.e. Q4, for production STARTS, which fits with 2011 32nm products.

So, at this point, 1 to 1.25 years behind Intel on 32nm revenue shipments, despite the bluster about "closing the gap" that they provide every couple years.

when you consider that some of the "legends" of this forum like Shintai firmly believed that Bulldozer will not be on market before the end of 2012... yeah guess you're right than, this "piece of info" that you've put on thread title is important...

On the other side, fact that there are bunch of old posts on this forum with God-knows-when leaked AMD roadmaps stating that Zambezi is out in 2011... then your's "add-on" in thread title is really pointless!

But then again apparently you were the one who didn't folow those posts and roadmaps

Also when AMD spoke about ramping up 32nm in 2010 AMD meant production 'cos at that time AMD owned production....

Bottom line Bulldozer EVT and ES products are coming in 2010, so your thread title IS pointless!

[terrace215]

Q3 production starts and by the end of 2010 shipment of products.

Mmm, no.

I direct you to:

This time around I’d expect as I said the ramp to start deep in the back half of next year.

That's the CEO, Dirk, saying this last month:

http://seekingalpha.com/article/1668...e=-1&find=deep

And, "back half" = H2, "deep in the back half" = Q4, not Q3.

And, of course, no products til 2011. AMD claims about 10 weeks from start to out, so 2011 products fits with starts in Q4 2010 or later.

Remember we're talking about AMD, not the GloFo foundry, whatever that bar means.

[informal]

EVT samples are Q4 2010 for BD cores ,so it doesn't matter what date it is exactly.Products are 2011 like it was known for a year now.
BD cores look like a game changer,2 shared int units,256b wide FPU,APU support,advanced clock gating and Turbo functionality,platform longevity(G34 and MA3 compatible).

[terrace215]

Bottom line Bulldozer EVT and ES products are coming in 2010, so your thread title IS pointless!

EVT and ES samples are not products, nor were they that specific about WHAT level of sample they hope to provide in 2010.

I wouldn't be surprised to see Bulldozer products slip to Q1 2012, actually. Best case is probably Q3 2011. There's a lot of verification work that needs to happen for a new server part, even once they've debugged and tuned the silicon.

[informal]

Yeah they could have said it's Q3 2013 and when they launch it in H1 2011 they can say we overdelivered by 2 years!

[Nedjo]

Yeah they could have said it's Q3 2013 and when they launch it in H1 2011 they can say we overdelivered by 2 years!

informal, don't go in to this pointless arguing with some one who is obviously misinterpreting pure facts, and hoping that AMD will slip Bulldozer to an "Apocalypse year"

come here and give us more info from the analyst day: http://www.xtremesystems.org/forums/...d.php?t=238702

[terrace215]

Misleading? What kind of misleading are you talking about?

This kind of misleading:

http://www.informationweek.com/news/...SSfeed_IWK_All

There's Dirk telling folks they'll be ramping up 32nm production in mid-2010 with VOLUME in Q4!

All the fans went wild when that was announced... but, no, sorry, just the usual AMD spin labeling their "sampling period" as "ramping up", but leaving it vague on purpose.

[informal]

Q&A,Bergam hints that they are doing the BD tape-out at the start of the next year and having silicon in H1 2010 for testing purposes.
I highly suspect they already taped out BD samples at 45nm and verified the design works as it should so the job now is to "just" shrink it to a new node.Remember BD design was 2009 and 45nm without AVX. The delay is logical looking at things from this POV.

[kl0012]

Yeah they could have said it's Q3 2013 and when they launch it in H1 2011 they can say we overdelivered by 2 years!

No, they can't. New arch in 2013 isn't something they can tell to shareholders. In fact theirs predictions lately were a bit optimistic (like a 30% of market share in 2009).

[highoctane]

Roadmaps are not law/set in stone, if things go good and they can pull things in if things go bad they extend the roadmap out to a later date, its part of planning.

[terrace215]

BD cores look like a game changer

I see you swallowed AMD's marketing meme for this Analyst Day. "Changing the game"... "game changer"...

Lately they've also been saying that performance benchmarks don't really matter....

Uh Oh.

http://blogs.amd.com/work/2009/10/09...st-benchmarks/

[terrace215]

I highly suspect they already taped out BD samples at 45nm and verified the design works as it should so the job now is to "just" shrink it to a new node.

I highly suspect you are wrong, but that does make for some nice wishful thinking.

[informal]

Ok.Relax now,listen to the webcast and don't be so negative

[w0mbat]

Lets move the talk to here -> http://www.xtremesystems.org/forums/...d.php?t=238702

Thanks!