Sign in with
Sign up | Sign in
Your question

AMD Q3 Report

Last response: in CPUs
Share
October 15, 2010 1:45:13 AM

When is fab light going to start looking good for AMD.
we were all told how great this was for AMD to become
fabless with a 30% stake in GF it's not looking like AMD
can make a profit, maybe AMD should sell the remaining
30%.
m
0
l
a c 127 à CPUs
a b À AMD
October 15, 2010 4:38:22 AM

^AMD will sell of the remaining 30% to ATIC from what I hear over time. In the end, AMD will be fabless much like nVidia and ATI.

But from the report, AMD had a net loss of $118 million while Intel posts 11.1 Billion.

Not sure AMD can survive a price war and its a good thing that Intel doesn't seem to be pushing towards one because if they did, AMD would hurt badly.
m
0
l
Related resources
a b à CPUs
October 15, 2010 1:29:36 PM

What I found interesting is the below info from AMD's earnings conference (obtained from some guy over on AMDZone :D ):

Quote:
Patrick Wang - Wedbush Securities
My first question is just on the 32 nanometer Fusion parts that are coming out here. You talked about some challenges when you updated your product launch schedule last quarter. Just curious how pleased you've been with the yields. How has hit ramped just over the last couple of months here?

Dirk Meyer
The high order answer is that we can't say we're pleased in the sense that we announced last quarter that the yields were not at the level of maturity that we had planned. And for that reason, as well as for a few others we mentioned, we moved the ramp back into next year. If your question is what kind of progress we're making, I'll say that over the last 90 days the GLOBALFOUNDRIES team has made progress, are building momentum, and we need to see that progress continue into next year, and we expect them to do so.

David Wong - Wells Fargo
With your current schedules will Llano appear in systems and ramp before the Bulldozer desktop and server ships, or after the Bulldozer desktop and server ships?

Dirk Meyer
Before.


So if Llano is now Q3 of next year as according to some reports, then Bulldozer is also Q3 or later. And since server is supposed to be before desktop, I'm thinking Fudzilla's Q4 prediction may be true.

I hope JF-AMD can shed more light on this..
m
0
l
October 16, 2010 2:01:38 AM

Quote:
bottom line AMD states they aren't going anywhere and have plan(s) already in motion for the future and competing with Intel.
so if there going to be around and going to try, I'm willing to see what happens..



Well you can't keep losing money and expect to stay in business no matter what AMD say.
m
0
l
October 16, 2010 2:08:23 AM

the government isnt going to let AMD go down.
m
0
l
a b à CPUs
October 16, 2010 1:32:35 PM

earl45 said:
Still not profitable is my point, no fabs and still can't make a profit, what else can they do to make money.


From what I read, it's the ~30% share AMD still owns in GF that cost them $$. If they had completely divested themselves then they would have made a profit.

Also read that AMD reamortized some $800M of senior notes due in 2015 at 6% interest, to $500M of 7.75% interest senior notes due in 2020. Seems to me that perhaps they are sweetening the pot a bit for a potential buyout by pushing debt out even further, plus they did not outright deny that AMD is not for sale, given a proper offer.
m
0
l
a b à CPUs
October 16, 2010 1:34:23 PM

Frizzo said:
the government isnt going to let AMD go down.


I'd believe that Intel wouldn't let AMD go down (and be an outright 100% monopoly instead of just 80%), but the voters are a bit fed up with gov't bailouts right about now :p ..
m
0
l
October 16, 2010 2:02:46 PM

fazers_on_stun said:
From what I read, it's the ~30% share AMD still owns in GF that cost them $$. If they had completely divested themselves then they would have made a profit.

Also read that AMD reamortized some $800M of senior notes due in 2015 at 6% interest, to $500M of 7.75% interest senior notes due in 2020. Seems to me that perhaps they are sweetening the pot a bit for a potential buyout by pushing debt out even further, plus they did not outright deny that AMD is not for sale, given a proper offer.



Well we don't no what would have happen if they sold the fabs outright.
AMD keep blaming their losses on GF don't mean that's actualy where the
losses are come from.
m
0
l
a b à CPUs
October 16, 2010 2:34:17 PM

Well at the time, the x86 license from Intel specified that AMD and ATIC would have to each have a 50% control over GF and AMD to have at least a 30% ownership share, IIRC. So that was the best AMD could do at the time. I'm sure they would have preferred to just sell the fabs outright and be done with it, rather than drag it out for years. I believe the renegotiated license lets AMD divest themselves completely from GF.

I read somewhere that no semi company that spun off its fabs ever became a success - if AMD succeeds then they would be the first.

I also think that the divestiture will mean GF and AMD have started to see their company goals diverge, since GF appears to be going after ARM and other non-SOI customers and AMD is their only SOI customer, plus the fact that the 32nm node is costing billions. Given the problems with Llano causing it to get pushed back (Q3 now apparently), my feeling is that perhaps GF is devoting more time and resources to landing those bulk Si customers who appear to be stronger financially.

For a while last year, I sorta thought the FTC's anti-trust litigation against Intel, if successful, would make them split up like the FTC did to AT&T some 25 years ago. Including maybe making Intel spin off their fabs. In which case we might have seen AMD turn to Intel's fabs to make their 32nm CPUs..
m
0
l
October 16, 2010 2:41:34 PM

Imagine what would it have been with 100% ownership, having to pay for the R&D for the newer nodes, plus HKMG
These expenitures will show a profit later, to me its more an investment, like any good R&D, providing its a winner
m
0
l
a b à CPUs
October 16, 2010 2:51:15 PM

AMD has been making money all year. Between interest, producing and not selling 6xxx, and probably a lot of R&D, id expect a loss this quarter. However, this year has been a very good year, and if Bulldozer and 6xxx/7xxx do well, so should 2011-2012. I doubt they will gain much on Intel, and they already have the lead over Nvidia, but this could get them out of any debt, and really help set them up for a better future. Most games are apparently being designed with Radeon in mind for 2011-2012, so that could be a nice advantage for ATi division. Once they get out of the hole they made for themselves, they will be better off and they actually have made a decent chunk of change this year...until another "Phenom" disaster comes :lol: 
m
0
l
October 16, 2010 2:57:05 PM

The investments into ATI and their fabs havnt seen full potential as yet
Now, to argue will these be enough, or focused properly, or in the right direction, is an all together differing POV
Til these things happen, itll be a smaller recoup on R&D evpenditures.

I also agree, Intels been generous, not really entering into a price war, tho Intels setup with their investors may not be too compelling for this strategy
Whatll be interesting in the future, like what was seen in the settlement, is if AMD does garner more marketshare, Intel declared the gloves will be off, and thats when we, the consumer will see some very nice pricing, if this ever occurs of course
m
0
l
a b à CPUs
October 16, 2010 3:04:37 PM

JAYDEEJOHN said:
The investments into ATI and their fabs havnt seen full potential as yet
Now, to argue will these be enough, or focused properly, or in the right direction, is an all together differing POV
Til these things happen, itll be a smaller recoup on R&D evpenditures.

I also agree, Intels been generous, not really entering into a price war, tho Intels setup with their investors may not be too compelling for this strategy
Whatll be interesting in the future, like what was seen in the settlement, is if AMD does garner more marketshare, Intel declared the gloves will be off, and thats when we, the consumer will see some very nice pricing, if this ever occurs of course


I agree, right now Intel is the big kid holding the little kid back with their hand on their head. If AMD ever got big enough to challenge the big kid, things could really get interesting.
m
0
l
October 16, 2010 3:07:29 PM

Dont have it on hand, but I remember in the settlement, a certain marketshare change changes many things, which I read as the consumer wins
m
0
l
a b à CPUs
October 16, 2010 3:25:21 PM

JAYDEEJOHN said:
The investments into ATI and their fabs havnt seen full potential as yet
Now, to argue will these be enough, or focused properly, or in the right direction, is an all together differing POV
Til these things happen, itll be a smaller recoup on R&D evpenditures.



I agree. The fabs will pay off oneday.
m
0
l
October 16, 2010 3:33:25 PM

If we look at the shrinking competition, and moves towards groups pushing forwards, since fabs exist only to make a good/better product, my thoughts are theres a certain potential profit margin allowed in all this, as its a certain expenditure of the overall product(s), and has to gaurantee a certain stability in the overall costs structure.
Will these profits be huge as pertains to the fabs?
My opinion is no, but adding ATI , thats a whole differing thing, but again, nothing so far but scraps, with lower profit margins, tho even just confined to discrete offerings, shows some decent returns possibly in the near term, other than the IP paid
m
0
l
a b à CPUs
October 16, 2010 3:54:11 PM

Quote:
Rumors are once again circulating on the possibility of another CPU maker joining the ranks of Intel and AMD. This possible new maker isn’t new to the computing scene but is better known for its high end graphics cards. NVidia is rumored to be hiring engineers to work on a new x86 CPU platform. The claims so far are that NVidia is not trying to compete with the top end Intel chips but to open in the lower to mid range market.

It will be interesting to see the outcome of this rumor and what it brings for the future. NVidia along with ATi have long been known for their graphics abilities in the gamers market so I’d expect a good quality product if they do enter the x86 market. We’ll just have to wait and see what they do next.


Nvidia has the tegra line, but thats always been more of a Ultra mobile smartphone thing, and never very sucessful at that. Id like to see the 3 duke it out on a full scale. Each in CPU, and each in GPU. Now THAT would be interesting! :lol: 
m
0
l
a b à CPUs
October 16, 2010 4:00:22 PM

Intel could really do some damage in GPU market eventually, and Nvidia could really use a CPU division. As much as they dont like each other, a merger would be economical in the event AMD, or more specifically ATi gets stronger. Nvidias massive die thing isnt working. As MP gets smaller, it will be more prone to problems, problems that have and will screw up Nvidia entire lineup. Where as Nvidia is going for another G80 grandslam every time, AMD's small die is more profitable (not necessarily, but in a sense, it is) and easier to mantain. However, If Nvidia really wants to keep going big die, Intel could really help them out with that, and Intel could manipulate Nvidia to keep ATi in check if it grows to become a 800 pound, which it very well might if Nvidia has more fermi's and ATi has more of these successes.
m
0
l
October 16, 2010 4:05:44 PM

Maybe I didn't form my questions the right way.

1. How can GF be such a threat to anyone with all the business and company's
it buying but not equate to a profit for AMD which own 30%.

2. If fabless is the way to go what AMD's problem, Intel has many fabs and making record profits.
m
0
l
October 16, 2010 4:37:10 PM

For AMD, in the scenario they found themselves in, fabless light was the way to go
The investments of keeping a full cutting edge fab costs monies, which hasnr seen the light of day as yet, so in essence, once these improvements do (32nm/28nm/HKMG etc) then profits will arrise, such as more di per wafer, more competitive product in power/perf, thus higher values of components etc.
Thats answering both if I got it right
m
0
l
October 16, 2010 6:51:39 PM

JAYDEEJOHN said:
For AMD, in the scenario they found themselves in, fabless light was the way to go
The investments of keeping a full cutting edge fab costs monies, which hasnr seen the light of day as yet, so in essence, once these improvements do (32nm/28nm/HKMG etc) then profits will arrise, such as more di per wafer, more competitive product in power/perf, thus higher values of components etc.
Thats answering both if I got it right



Well from past statements you insinuated AMD was doing the right thing and Intel may find it self
in trouble for having all the fabs it owns, as of today that's just not true AMD is still having money
problems and Intel is making billions.
m
0
l
a b à CPUs
October 16, 2010 7:07:08 PM

Quote:
Excellent debate very informative.Keep up the good work guys.But one question~ why would intel ever sell x86 license to nvidia?


Well, they have before to others, so its not impossible. Makes a merger seem even more lucrative :lol: 
m
0
l
a c 111 à CPUs
a b À AMD
October 16, 2010 7:17:07 PM

Quote:
Excellent debate very informative.Keep up the good work guys.But one question~ why would intel ever sell x86 license to nvidia?


I don't think they could deny the license - would be big anti-trust problems for Intel if they did so. They question is why nVidia would want to enter into such a cut-throat business (other than the possibility of building back their chipset business).

AMD is doing okay, and Fab Lite is reaping the rewards they anticipated. GloFo will continue to be a drag on the income statement. Saratoga will not begin shipping parts until 2012, Dresden is being upgraded, they sucked up Chartered Semiconductor, took a long look at acquiring UMC before saying no, and the rumor on the street of a planned $7b Fab in Abu Dhabi for 2014-15 won't go away.
m
0
l
October 16, 2010 8:20:08 PM

GF may have to sweeten the pot for its partner (AMD) in order for Abu Dhabi to go thru sooner than later
m
0
l
a b à CPUs
October 16, 2010 9:22:24 PM

I think AMD has played this smart. Right now they should be making quite a bit of money, if they werent paying so much for expansion. They have picked the time were they can afford to spend, and they are solidfying a very solid future. Right now, and id say until 2015, AMD will only really compete with intel if Intel messes up big time and AMD has a succesful series. However after 2015, when all the investments they make now pay off, then they will be a real handful for intel on a full scale, atleast in my opinion.
m
0
l
October 17, 2010 3:51:17 PM

earl45 said:
Well from past statements you insinuated AMD was doing the right thing and Intel may find it self
in trouble for having all the fabs it owns, as of today that's just not true AMD is still having money
problems and Intel is making billions.

AMD is doing the right thing, by them.
Theyve incorporated a world class partner with waaaay deeper pockets than Intel, while at the same time, they found themselves slipping in this regard because of Intels advancements and possible severly damaging effects on them
Now, as Ive said 2 times already in this thread, the returns on the investments thatve kept them from going black havnt been realized yet, but will be
If thats not the way you see it, then tell me how you do?
As fpr Intels profits and their fabs, if they dont get tax write offs for closing fabs down the line, theyll have to eat what I believe is an eventual overabundance of them, which is my opinion
As fabs become more considated, therell be fewer buyers of older fabs, as well as node shrinks bring more per wafer, meaning less overall production needed from more lines/fabs
This is my opinion, nothing more
If you have a differing one, Id like to hear it
m
0
l
a b à CPUs
October 19, 2010 4:55:40 PM

JAYDEEJOHN said:
AMD is doing the right thing, by them.
Theyve incorporated a world class partner with waaaay deeper pockets than Intel, while at the same time, they found themselves slipping in this regard because of Intels advancements and possible severly damaging effects on them
Now, as Ive said 2 times already in this thread, the returns on the investments thatve kept them from going black havnt been realized yet, but will be
If thats not the way you see it, then tell me how you do?
As fpr Intels profits and their fabs, if they dont get tax write offs for closing fabs down the line, theyll have to eat what I believe is an eventual overabundance of them, which is my opinion
As fabs become more considated, therell be fewer buyers of older fabs, as well as node shrinks bring more per wafer, meaning less overall production needed from more lines/fabs
This is my opinion, nothing more
If you have a differing one, Id like to hear it


IMO, GF is more concerned with competing with TSMC and the other contract fabs, instead of Intel. Unless AMD can really pump up their CPU sales - which mainly means taking marketshare from Intel when the global market is flat like it currently seems to be - then most of GF's business prospects would be luring away the contract fab customers. Since ATIC owns over 70% of GF at the moment, I'd imagine that they pretty much steer GF where they think the profits will be.

Also, the same criteria you mention as affecting Intel's fabs also affect GF's, not to mention that Intel's older fabs most likely have already paid for themselves. GF (and AMD before them) seems stuck at 12+ months behind Intel on the last 3-4 nodes now, so by the time they are ramping up on 32nm, Intel will be sampling or shipping 22nm.

So looking at AMD's prospects in the near future, we have Zacate (netbook APU), which at 19 watts according to the AT review is currently a bit too power-hungry for the tablet market (which is where it appears netbooks are headed). We also have an ES sample of Llano (1st Fusion APU) that according to the AT review needs a lot of improvement on the graphics end to dominate SB, as the SB preview was showing similar framerates for some games on medium settings. And of course there is Bulldozer coming down the road, which seems to me to be the most promising of the 3 as it is aimed at the server market.
m
0
l
a b à CPUs
October 19, 2010 5:12:37 PM

From The Street :

Quote:
While many nodded at the company's solid third-quarter gross margin and lower operating expenses (despite the weak computing environment in the third quarter) and found solace in the company's ability to maintain profitablity through finance, manufacturing and operational restructuring -- a closer look at the reports generated concerns.

Such concerns included the company's flat fourth-quarter guidance, its weaker performance than Intel for the same quarter, its ability to follow through on new product execution, and its lag in technology next to Intel. Following the earnings report, Morgan Stanley analyst Mark Lipacis told clients in a note that "we have multiple concerns -- fourth-quarter guidance for flat quarter-on-quarter revenues was about 3 points below Intel's, AMD appears to have lost share in server again, GPU revenues declined by 11% quarter-on-quarter and the company expects the inventory correction to extend into the fourth quarter."

"We remain concerned about AMD's 80%-plus revenue exposure to" the consumer PC market, "and execution risks related to its Fusion products," he added. Meanwhile, Goldman Sachs analyst James Covello said, in a client note, that he expects AMD to continue to encounter challenges with its outsourced model because of the complexities of having separate design and manufacturing teams. "This has been evident in the 32nm transition delay."

Stifel Nicolaus analyst Kevin Cassidy said in a client note that AMD did well in the third quarter taking into account a sudden decline in demand and new product launches by rivals. "Returning an operating profit under these conditions is laudable."
m
0
l
a c 127 à CPUs
a b À AMD
October 19, 2010 5:50:28 PM

BadTrip said:
I agree. The fabs will pay off oneday.


I this one day too far off though? Too far away and it could cost AMD a lot.

I wonder how AMD managed to sell 25 million DX11 based ATI GPUs in about 1 year nd they still seem to lose moeny:

http://www.tomshardware.com/news/radeon-hd-dx11-gpu-dir...

Just maks ya wonder really.
m
0
l
October 19, 2010 6:33:31 PM

No mention of new products?
The gfx area has been flat anyways, and its certain theyll gain marketshare there, now with the 6xxx series out
How they come to the assumption of troubles at 32nm is a rarity in discernment, since it isnt out yet
That guy outta stick to the fast track n hug his computer
m
0
l
a c 127 à CPUs
a b À AMD
October 19, 2010 7:21:46 PM

^Troubles in the 32nm would probably go towards the process itself or the ramp. I don't think they mean with the processor itself.

Now Phenom was a example of a bad processor on a decent process. AMDs 65nm, while not great, was decent as seen in the older Athlon X2s. But for some reason Phenom just didn't take with it.

I don't think that will be the case with BD and 32nm. I think it will work just fine on the 32nm process, more that the troubles they are having are in ramping the process and in the process tech itself. As Intel said, HK/MG is not easy to achieve. They researched it for quite a while before going with it. They had plenty of time to work the kinks out.

If GF is trying to push HK/MG out too fast it could be a disaster. I rather they take their time to get all the kinks out and make sure the process works properly or it could hurt Llano and BD.
m
0
l
October 19, 2010 7:35:32 PM

Me too, and thus the delays, which shouldnt be construed as AMDs problems, since its GFs, cant have both, and that guy knows better
32 is right around the corner tho, and wisely, small at first
Of course, he could have blamed everyone else in the consortium as well
m
0
l
a b à CPUs
October 21, 2010 4:50:29 PM

Another story from The Street

Quote:
Sandy Bridge Ships in Fourth-Quarter 2010
Intel is expected to begin shipping its next-generation microprocessor family codenamed Sandy Bridge in the fourth quarter of 2010. The company showcased some design features for this product family at the Intel Developer Forum last month. This includes extended chip performance and battery life along with several in-built visually related features. Intel stated in its recent earnings that Sandy Bridge represents the largest increase in computing performance in the company's history and early demand is very high.
Intel's customers are trimming down their inventories in anticipation of pushing out products based on the new Sandy Bridge processor family. Intel plans to ship these new chips for PCs (notebooks and desktops) first and extend the architecture later to servers as well.

Notebook and Desktop PC Businesses Benefit
The new Sandy Bridge processors can potentially lift Intel's market share in notebook and desktop processors beginning as early as first-quarter 2011. We estimate that the notebook and desktop processor businesses account for about 45% and 11% of Intel's value, respectively.

What if Intel's market share in both notebooks and desktops were to rise to 90% with the help of Sandy Bridge, establishing dominance similar to Intel's share in the server microprocessor market? This could further lift the Trefis price estimate for Intel's stock by about 6%.
Server Processor Business Benefits
Server microprocessor revenue for Intel grew by 30% over third-quarter 2009. This was driven by demand for high-performance Westmere and Nehalem-EX processors.
Cloud computing is also driving demand for server microprocessors. The shipments for this segment in third-quarter 2010 grew by 200% over third-quarter 2009. Moreover, the Xeon product family continues to drive growth in the data center (storage) segment.
We expect Intel to maintain high server processor market share of close to 92% in calendar year 2010. Although we expect AMD to fight back and regain some of its lost share in this segment in coming years, Intel's introduction of server processors based on the Sandy Bridge architecture can potentially moderate AMD's assault. We estimate that Intel's server processor business constitutes 24% of Intel's value and that changes in the company's server processor share can have a significant impact on Intel's stock.

m
0
l
!