AMD Graphics division made $12M on $403M revenue last qrtr

[tviceman]

The article is on anandtech's front page: http://www.anandtech.com/show/5026/amd-q311-results-97mil-on-strong-llano-sales

Finally, the graphics division of AMD once again turned a operating income of $12M on revenue of $403M. As APU revenue is booked separately from GPU revenue, the graphics division often teeters between a profit and loss, so its fate is tied to AMD’s discrete GPU sales. Overall discrete GPU sales were up both sequentially and year-to-year, as was the ASP.

That sounds strange to me. Especially how the graphics division is "often teetering" between profit and loss. I thought graphics was AMD's current saving grace among the company and that it was AMD bringing old ATI's ship down....

 

[RussianSensation]

Could it be that because of certain accounting treatments we are getting such a low #? AMD likely books all graphics expenditures (including marketing and R&D), even attributable to the APU design, but excludes the revenue from APU here, thus lowering the bottom line profitability.

Secondly, keep in mind that cash flow generated by the division need not be equal to its operating income. They probably have a lot of non-cash items that misrepresent the actual cash flows such as Interest Expense, Depreciation/Amortization, etc.

So in reality, the division may just as well be making $40-50M in free cash flow, not $12. Either way, not very impressive in the grand scheme of things, but still likely much better than what the earnings # shows.

All in all, there is no dispute here that NV has a much stronger business model. Let's just hope AMD continues to sell competiting graphics card. We don't want an Intel vs. AMD CPU style situation in the graphics card market (remember GTX280 for $649?).

 

[Silverforce11]

AMD uses ATI R&D for their CPU/APU thus its accounted as a cost. But AMD does not attribute sales of such products to ATI, thus no income. Overall, if ATI can maintain a small profit despite subsidizing AMD, it means their performance is great.

Is that $403M a quarter for discrete graphics? Seems very high and growing.

 

[tviceman]

What struck me particularly odd was Ryan's comment on "the graphics division often teeters between a profit and loss." He makes it sound like there is quite a bit of work involved with very little reward for having a discrete GPU business. If that's the case, why do they even bother?

 

[Silverforce11]

Did you not understand? Its only like that on paper. Its real income is very high and profitable but the division subsidies APU R&D.

IF they didn't bothered, AMD's CPU business would have to fund their own APU development and they would be deep in the red.

 

[notty22]

Where AMD is trying to make inroads is the workstation GPU. Which they have gained some ground. This can push forward a bigger budget business model. From design to die size/complexity. If they can use the same gpu that can get up to 500 from the gaming market and up to 4000+ in the workstation market.
Everyone wins.
http://www.amazon.com/exec/obidos/AS...=ATVPDKIKX0DER

As a whole this was reported

The company reported a third quarter profit of $97 million (13 cents per share), compared to a loss of $118 million (17 cents per share) last yea

 

[iCyborg]

Did you not understand? Its only like that on paper. Its real income is very high and profitable but the division subsidies APU R&D.

IF they didn't bothered, AMD's CPU business would have to fund their own APU development and they would be deep in the red.

Yeah, that's what I was arguing in Q2 results thread as well: http://forums.anandtech.com/showthread.php?p=32183987#post32183987
APUs are skewing the numbers. It would be hard to separate APU devt because a lot of graphics R&D is APU/dGPU agnostic...

 

[SirPauly]

On the graphics division, the gains were from this:

This growth was driven largely by strong seasonal revenue increases in the AIB channel.

http://seekingalpha.com/article/303...sses-q3-2011-results-earnings-call-transcript

 

[SirPauly]

The article is on anandtech's front page: http://www.anandtech.com/show/5026/amd-q311-results-97mil-on-strong-llano-sales



That sounds strange to me. Especially how the graphics division is "often teetering" between profit and loss. I thought graphics was AMD's current saving grace among the company and that it was AMD bringing old ATI's ship down....

That was a nice summary by Ryan -- just read it.

 

[exar333]

Did you not understand? Its only like that on paper. Its real income is very high and profitable but the division subsidies APU R&D.

IF they didn't bothered, AMD's CPU business would have to fund their own APU development and they would be deep in the red.

Good point. A lot of companies are like this where sub-divisions are responsible for the tech (read: R&D costs) while the company 'as a whole' reaps the revenue directly. If a part of company is doing poorly, it is usually reviewed to corrective action. On the other hand, if you can pull a profit, while helping to drive sales elsewhere in the company, it is win/win.

If AMD was still it's own company, this would look pretty bad. But that is not the case here, and more and more sales from 'AMD' will aggregate into APU sales.

 

[SirPauly]

The CEO really talked about improving execution and margins and seems to be a straight shooter -- didn't try to spin the negatives but deal with them and try to improve.

 

[tviceman]

Did you not understand? Its only like that on paper. Its real income is very high and profitable but the division subsidies APU R&D.

IF they didn't bothered, AMD's CPU business would have to fund their own APU development and they would be deep in the red.

I started typing my post before you finished yours, so I didn't see your first post until this morning. Do you know for a fact that fusion R&D costs is accounted for on the GPU side?

 

[SirPauly]

Some clarity there would be welcomed.

 

[Dribble]

I remember in the hieght of the 58xx series when fermi was missing/too hot they made $2 million. ATI doesn't seem particularly profitable. AMD massaging the figures to make the cpu's look good sounds like an excuse imo. They just don't sell many high margin graphics cards - barely enough to pay for the RnD anyway. Nvidia makes the big profits selling quadro/telsa - in these markets ATI is weak/missing. I think the bottom line is there simply isn't much money selling cards to people like us.

 

[zlejedi]

What were profits of Ati as a separate company ?

 

[SirPauly]

The key is AMD did make a lot of money with their graphic division:

Q4 -- 2009:

The Graphics segment reported an operating income of $53 million, a substantial improvement from the prior quarter.

http://seekingalpha.com/article/183780-advanced-micro-devices-inc-q4-2009-earnings-call-transcript

Q1 --2010:

Graphic segment operating income was $47 million

http://seekingalpha.com/article/199031-advanced-micro-devices-inc-q1-2010-earnings-call-transcript

Q2 -- 2010:

Operating income was $33 million

http://seekingalpha.com/article/214781-advanced-micro-devices-inc-q2-2010-earnings-conference-call

Q3 -- 2010:

And this is where it started to get spotty, moving forward:

Operating income was $1 million

http://seekingalpha.com/article/230...sses-q3-2010-results-earnings-call-transcript

 

[bryanW1995]

$134 million a year profit isn't too bad, but almost all of that was when they were competing against paper clips, rubber bands, and wooden screws from nvidia. And when you factor in that they paid $5 billion for ATI, that would only take 35 years to make that money back. Assuming that there is no time value of money, etc etc. They'd have almost been better off investing in Greek govt bonds back in 2006.

The really scary part is that's their best 4 quarter performance since the acquisition, too.

 

[SirPauly]

Take the ATI acquisition away from AMD, where would they be? Where would their growth potential be? With a combined ATI/AMD, they can create innovative platforms and grow as a company; it was the best thing the company could of ever done, imho!

 

[exar333]

Take the ATI acquisition away from AMD, where would they be? Where would their growth potential be? With a combined ATI/AMD, they can create innovative platforms and grow as a company; it was the best thing the company could of ever done, imho!

Agreed. The criticism is that AMD overpaid for ATI, when no competing bids were even offered. They paid a premium for it, and could have probably acquired it for much less.

 

[TemjinGold]

$134 million a year profit isn't too bad, but almost all of that was when they were competing against paper clips, rubber bands, and wooden screws from nvidia. And when you factor in that they paid $5 billion for ATI, that would only take 35 years to make that money back. Assuming that there is no time value of money, etc etc. They'd have almost been better off investing in Greek govt bonds back in 2006.

The really scary part is that's their best 4 quarter performance since the acquisition, too.

Can't count it that way. You have to consider if they sold ATI today what they would get for it, then deduct from the $5 billion they paid to find the amount they need to "make back." Because when they've "made $5 billion back" in 35 years, they would have that money AND they would still have ATI.

 

[bryanW1995]

Take the ATI acquisition away from AMD, where would they be? Where would their growth potential be? With a combined ATI/AMD, they can create innovative platforms and grow as a company; it was the best thing the company could of ever done, imho!

They could have worked out a deal to pay ATI for that technology/know-how for MUCH less than $5 billion.

Can't count it that way. You have to consider if they sold ATI today what they would get for it, then deduct from the $5 billion they paid to find the amount they need to "make back." Because when they've "made $5 billion back" in 35 years, they would have that money AND they would still have ATI.

Well, if they were paid 10 x profits that would be around a cool $1 billion, and that number seems incredibly optimistic.

 

[SirPauly]

Agreed. The criticism is that AMD overpaid for ATI, when no competing bids were even offered. They paid a premium for it, and could have probably acquired it for much less.

I think the purchase was about growth potential. ATI needed AMD for growth and AMD needed ATI for growth; a nice fit. There always seems to be criticism no matter what. ATI had leverage to me and AMD paid a bit of a premium.

 

[SirPauly]

They could have worked out a deal to pay ATI for that technology/know-how for MUCH less than $5 billion.

Intel invested over a billion dollars into a failed product up to this point and how many dollars have they invested into GPU's? ATI was a world leader and some could make the point the world leader at that time, with record revenue and the most robust and fastest GPU in the World -- x1900, with chip-set know-how and engineering.

 

[SirPauly]

TO put things in perspective --- nVidia has invested around 2 billion in just Tegra.

 

[OCGuy]

AMD is in trouble.