AMD Reports Record First Quarter Revenue, Profit Of $257 Million.

[god_43]

SUNNYVALE, Calif. —4/15/2010

AMD (NYSE:AMD) today announced revenue for the first quarter of 2010 of $1.57 billion, net income of $257 million, or $0.35 per share, and operating income of $182 million. The company reported non-GAAP net income of $63 million, or $0.09 per share, and non-GAAP operating income of $130 million.

“Strong product offerings and solid operating performance resulted in record first quarter revenue,” said Dirk Meyer, AMD President and CEO. “We continue to strengthen our product offerings. We launched our latest generation of server platforms, expanded our family of DirectX 11-compatible graphics offerings, and commenced shipments of our next-generation notebook platforms to customers.”

http://www.amd.com/us/press-releases...2010apr15.aspx

[Manicdan]

saw this, kinda related and explains a little too.
http://finance.yahoo.com/news/AMDs-1....html?x=0&.v=1

looking good now that they dont have that fab

[terrace215]

Selling off 5% A/H.

Basically, operating income declined in both GPU (-$3M) and CPU (-$15M), and the reason they are profitable vs last Q is all in the GloFo deconsolidation. (Not the $325 one-time, but no $100M of shared GloFo losses this time.)

Q2 guidance was typically weak.

The real problem seems to be this:

Analyst: "What can you do to improve microprocessor profit margins from the current 7% other than growing the top line revenue?"

Answer: It's a topline story.

Meaning... now that deconsolidation has happened, there doesn't look like any room for profit growth other than gaining share or overall market growth.

And the other story about Magny-Cours pricing bringing 2s pricing to the 4s market may sound good to customers, but investors/analysts are like, um, how will you make up for the loss in profits? Didn't have a great answer there.

Just confirmed no 32nm & no llano in available systems until H1'11.

Late to the call, didn't hear if anything on Bulldozer was asked/answered.

[Hans de Vries]

From the CC:

- AMD has samples of the Bobcat (28nm) fusion design.
- Llano is sampled to selected customers.


Regards, Hans

[informal]

Hans beat me to it .
Bobcat is less likely to be sampled to select partners so it must be Llano. Since they have Bobcat they indirectly hinted that BD cores are ready too(similar to Bobcat only done @ 32nm SOI not 28nm bulk).

[ajaidev]

The 28nm risk for ARMs was warm up/testing i knew it

[informal]

Dirk confirmed(35:24 into Q&A session): volume ramp for 32nm is H2 2010 with systems OEM systems avalaibility in H1 2011 (both notebook and desktop). I really have no idea what CC did terrace listen to,it surely wasn't this one ...

[terrace215]

Dirk confirmed(35:24 into Q&A session): volume ramp for 32nm is H2 2010 with systems OEM systems avalaibility in H1 2011 (both notebook and desktop). I really have no idea what CC did terrace listen to,it surely wasn't this one ...

AMDs "volume ramp" timing is never interesting. Dirk was asked 2 questions at different times.

1. No 32nm products in systems until H1'11

2. No Llano in systems until H1'11

The CC transcript will be up soon, you're welcome to check.

AMD always goes on and on about their internal ramps of processes to make people think the actual products will be available sooner than they actually will be. At least Dirk was clear in the CC:

32nm products? H1'11. Llano specifically? H1'11.

And let me get this right... by mentioning Bobcat, he was indirectly referencing Bulldozer?

That's, erm, quite some level of indirection, wouldn't you say?

[informal]

You can turn it around all you want,the man clearly said and I quote "OEM systems availability in H1 2011 ,both notebook and desktop". This has no hidden meaning man,deal with it already.
You are boring us to death with your no 32nm in 2011 talk.You have been beating that dead horse for a while now,let the "dream" die and move on already...

[terrace215]

Since they have Bobcat they indirectly hinted that BD cores are ready too(similar to Bobcat only done @ 32nm SOI not 28nm bulk).

Bobcat is NOTHING AT ALL like Bulldozer.

Good grief.

Don't you trust Chuck Moore?

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

APU presentation, slide 8.

-----

If anything, mentioning LLano & Bobcat, yet not being able to mention BD suggests no working BD silicon yet, perhaps not even tape-out.

[terrace215]

You can turn it around all you want,the man clearly said and I quote "OEM systems availability in H1 2011 ,both notebook and desktop".

Yes, that's LLano. I... never disputed this, in fact, I posted about it. LOL

[informal]

Nobody asked about Bulldozer rofl.... You are clearly wishing that BD hasn't taped out yet,but all I can say is "running material" is there(even your "beloved" Chuck Moore stated so in Q&A session,back in November man-that's 5 months ago).AMD always stated 2011 as a target for BD launch,looking at how they executed starting with Deneb(pull-ins all the way with the roadmaps) you can bet they will execute equally if not better.


PS Bobcat is very similar to BD,except it was made by a separate team with different goals in mind(hence half the int pipes and no AVX and FMAC support in FPU,different cache organization etc.).

[god_43]

pfft, w/e all i care about is market parity, and intel playing fair. i wish there were more than just 2 companies (via does not count....you cant even go into a BB and get one for gods sake!).if amds new stuff is on par or better than Intel's, and intel plays it fast and loose.........i will declare war!

[terrace215]

PS Bobcat is very similar to BD,except it

is NOTHING AT ALL like BD. I mean, really. It's a synthesizable part...

It's AMD's "Atom". If it were anything like BD, then BD is in big trouble, let's put it that way.

[informal]

The synthesize-able part means it's done by a set of tools and the process is much faster then the custom way. It doesn't represent the underlying uarchitecture and the way chip works. Bobcat lacks the frontend BD will have and has some other limitations I previously said that AMD made public. Synthesize-able means much greater flexibility in design and production (much less custom work) and the downside as Mr Moore said a loss of ~20% clock frequency.
BD is targeted at high end and it's 1) developed by a separate team 2) not using the Bobcat design process approach since it's a high speed and throughput uarch.

PS You skipped commenting (again ) on Chuck Moore's confirmation they have running material in the end of Q&A .This happened 5 months ago btw.

edit: My mistake ,Rick Bergman stated this ,not Chuck Moore.

[JF-AMD]

And the other story about Magny-Cours pricing bringing 2s pricing to the 4s market may sound good to customers, but investors/analysts are like, um, how will you make up for the loss in profits? Didn't have a great answer there.

Obviously, I can't get into the specifics on this because it digs deeper into the financials than we release, but here is the long and short of it.

2P makes up ~75% of the market. 4P makes up ~4% of the market (IDC server tracker numbers).

That means that roughly ~6% of the business is 4P and if you just look at 2P and 4P for now (because the 6000 series can do both).

Here is a post that I did on the subject:

http://www.xtremesystems.org/forums/...5&postcount=36

Now, the goal with shifting the pricing on 4P was the following:

1. Actually grow the market (it needs to get bigger or go away, it was on a pretty good clip to do the latter...)
2. Drive more revenue
3. Drive more absolute profit dollars
4. Drive more market share

None of that should surprise people.

The net is, yes, ASPs will be lower, but overall, the ASPs for 2P and 4P combined went up because, surprise, when the prices are lower, the 4P people buy towards the top of the stack ($1386) not the bottom).

If you sell more processors proportionally (because the 4P eats away at 2P), you shift business to higher margins.

[Russian]

Obviously, I can't get into the specifics on this because it digs deeper into the financials than we release, but here is the long and short of it.

2P makes up ~75% of the market. 4P makes up ~4% of the market (IDC server tracker numbers).

That means that roughly ~6% of the business is 4P and if you just look at 2P and 4P for now (because the 6000 series can do both).

Here is a post that I did on the subject:

http://www.xtremesystems.org/forums/...5&postcount=36

Now, the goal with shifting the pricing on 4P was the following:

1. Actually grow the market (it needs to get bigger or go away, it was on a pretty good clip to do the latter...)
2. Drive more revenue
3. Drive more absolute profit dollars
4. Drive more market share

None of that should surprise people.

The net is, yes, ASPs will be lower, but overall, the ASPs for 2P and 4P combined went up because, surprise, when the prices are lower, the 4P people buy towards the top of the stack ($1386) not the bottom).

If you sell more processors proportionally (because the 4P eats away at 2P), you shift business to higher margins.

Yes, but how in the hell can AMD survive with lower ASPs considering that they've already got lower ASPs than Intel to begin with? how do they expect to attain profitability if they can't retain high ASP?

My view is that the crux of the issue is that AMD simply cannot provide enough value to both consumers and enterprise class customers in order to validate a higher ASP. This, to me, has been the story for AMD since Intel's core 2 rolled out. Please correct me if i'm wrong.

[god_43]

has that not always been the case?

[cky2k6]

Yes, but how in the hell can AMD survive with lower ASPs considering that they've already got lower ASPs than Intel to begin with? how do they expect to attain profitability if they can't retain high ASP?

My view is that the crux of the issue is that AMD simply cannot provide enough value to both consumers and enterprise class customers in order to validate a higher ASP. This, to me, has been the story for AMD since Intel's core 2 rolled out. Please correct me if i'm wrong.

Way to misread and go off on a tangent... He said 4s asp will go down, but its still higher than 2s asp, and the 4s volume is expected to increase so overall (2s + 4s) asp will go up.

[terrace215]

Way to misread and go off on a tangent... He said 4s asp will go down, but its still higher than 2s asp, and the 4s volume is expected to increase so overall (2s + 4s) asp will go up.

He was spin-spin-spinning... and his bosses would be interested in his message-board commentary about 2p+4p asps.

[terrace215]

The synthesize-able part means it's done by a set of tools and the process is much faster then the custom way. It doesn't represent the underlying uarchitecture and the way chip works. Bobcat lacks the frontend BD will have and has some other limitations I previously said that AMD made public. Synthesize-able means much greater flexibility in design and production (much less custom work) and the downside as Mr Moore said a loss of ~20% clock frequency.
BD is targeted at high end and it's 1) developed by a separate team 2) not using the Bobcat design process approach since it's a high speed and throughput uarch.

Good grief, why don't you ask Hans, or someone else you trust, to set you straight on this point. Bobcat IS NOTHING LIKE Bullldozer. Period. End of story. Bobcat is an order of magnitude less complex.

[Russian]

Way to misread and go off on a tangent... He said 4s asp will go down, but its still higher than 2s asp, and the 4s volume is expected to increase so overall (2s + 4s) asp will go up.

Honestly, i feel like thats been the story for AMD for a long time. I admit, i did misread it... but nevertheless AMD has suffered in ASPs ever since conroe and seeing them have to lower 2S and 4S ASPs. Also, how will increased 4S volume affect ASPs? Most of the time increased volume of something like a 4S would drive down ASPs not up... unless demand outstrips supply. Which i think is doubtful.

[terrace215]

From the CC:

- AMD has samples of the Bobcat (28nm) fusion design.
- Llano is sampled to selected customers.


Regards, Hans

The only mention of "Bobcat" in

http://seekingalpha.com/article/1990...script?page=-1

is:



Derrick Meyer

We’ve talked in the past about our Bobcat core. That’s the brand new City 6 micro architecture which first appears in the Ontario product which we talked about in the analyst conference. The Ontario product is really focused on I’ll call it value PC’s and netbooks, but the Bobcat technology and other technologies that we have in house are appropriate to lower power envelopes such as are appropriate for Pads, and you’ll see Bobcat based products show up the following year appropriate to that market segment.



And there is no such mention of a 28nm design.

Methinks you may be misreading (spinning?) the following:



We do now have internal samples of both of our initial Fusion designs, are learning quite a lot, and are quite happy with what we see, and we started sampling to select customers, one of those two designs.



I think he refers to desktop + mobile LLano here, not Bobcat.

[ajaidev]

AMD reports $257 million profit

http://vr-zone.com/articles/amd-repo...ofit/8881.html

god_43 pls put $257 million and profit in the title

[Kuroimaho]

AMD always stated 2011 as a target for BD launch,

AMD will start including SSE5 support in the Bulldozer core, which with an expected launch date of 2009 is still two years off.

Source