Now that we have the quarterly numbers in for both Intel (who reported Tuesday) and AMD (who reported Thursday) we’re going to take a quick stroll through the numbers to see what’s up.

AMD

The AMD/Foundry Co. spilt has not been finalized yet, so the foundries are and will continue to stay on AMD’s books through the end of the year. Furthermore any of immediate investments and sales from the transaction are in the 4th quarter, since the plan was announced over a week after AMD’s 3rd quarter ended. So this 3rd quarter marks the last quarter where AMD is entirely whole, while the company’s books will begin to be shaken up in the 4th when Mubadala’s latest investment is counted. Finally, the Foundry Co. spin-off appears to be set in stone to complete in early 2009, AMD has ceased mentioning “late 2008” when discussing it.

Getting down to the numbers, as predicted for some time by AMD this quarter has been another loss. But it’s a loss that was anything but expected. For the quarter AMD lost a paltry $67 million on revenues of $1.78 billion, completely blowing away their own predictions and what analysts were expecting, losing nearly $100 million less than expected. Compared to their loss last quarter of $1.19 billion and their Q3 loss last year of $400 million, AMD may have as well broken even for the quarter. In fact if you switch to the numbers from their non-Generally Accepted Accounting Principles accounting, then they did indeed make a profit of $80mil for the quarter. The quarter also marks the end of AMD’s operating losses, they posted an operating income of $131 million this quarter.

AMD’s better than expected results can be attributed to 3 major sources. The first is their server CPU operations, for which Q3 was the first quarter where the TLB-fixed Barcelona shipped in volume for the entire quarter; in effect it was the first normal quarter for the CPU business in quite some time. The second source is a one-time license fee the company received for its 130nm SOI technology, which added $191 million to the bottom line. Finally, AMD’s graphics division turned a profit for the first time since they were purchased, posting a profit of $47 million on $385 million in revenue.

Looking at the segmented information, AMD’s CPU division hasn’t changed much in the last quarter. Intel continues to dictate CPU pricing, keeping AMD’s ASP for CPUs flat compared to both last quarter and last year. Due in part to the aforementioned TLB fix, the total number of CPUs sold was higher than last quarter, but compared to last year it’s still down – AMD’s CPU business is doing better than expected, but it’s still not doing well. Meanwhile everything is up with GPU sales; prices and units shipped are up compared to last quarter and last year. The Radeon 4000-series has been the smash hit AMD has been looking for from its graphics division, with Q3 marking the first full quarter for the high-end products of that generation. It needs to be kept in mind just how lop-sided AMD’s GPU and CPU business are in terms of revenue, so doing well here isn’t enough, but all-told the business is at a comfortable position. We’ll have more to discuss next month, once NVIDIA posts its Q3 numbers.

Finally, there are the Q4 projections. AMD is being a bit more vague than usual, perhaps because even they aren’t entirely sure how things will turn out (and as we’ll see, Intel is the same way). They’re expecting Q4 revenue to hold flat with Q3’s (minus the license fee revenue), which puts them around $1.58 billion. It seems unlikely that they’ll turn a profit in Q4 as a result, but it is not something AMD is explicitly outlining. AMD has now posted losses for the past 8 quarters, so there’s a lot of interest in finally snapping that streak with a strong Q4 performance. If not Q4, then Q1’09 with the completion of the foundry spin-off will certainly push the company in the black at long last.

Intel

It seems that as long as AMD doesn’t do too well, Intel can do quite well. Intel posted another big quarter, with $2 billion in profits on $10.2 billion in revenue, for a 12% rise in profit and 1% rise in revenue respectively. Interestingly however, the company missed analyst estimates on revenue, where they were expected to bring in $10.25 billion. Intel has been especially proud of their gross margin for the quarter, which was an unexpectedly high 59% (for the sake of comparison, AMD Is 51%). All things considered, the market seems happy as long as Intel’s gross margins stay this high, even if they miss estimates elsewhere.

Looking at the revenue breakdown, processor shipments are up but the ASP was down. Intel attributes this to the Atom, which is now fully shipping and dragging down the ASP of the rest of the CPU business since the Atom is significantly cheaper than Intel’s traditional CPUs. We don’t have a complete breakdown of the Atom versus everything else, but we do know that Intel sold around $200 million in Atom processors and chipsets in the 3rd quarter, and if you remove the Atom from Intel’s CPU sales the ASP of everything else held steady. It’s a bit of a riddle to see Intel’s ASP hold steady when they get to dictate the terms of the market, but it is high-end product launches that tend to drive up the ASP, of which Q3 had little. Furthermore cheap laptops are still driving down the ASP, to which the Atom isn’t helping. Nehalem in Q4 should help drive the ASP back up, but it won’t solve the long-term problem posed by cheap laptops and the Atom.

Finally, we have the future. Intel’s Q4 predictions are unusually reserved, as a result of the recent credit crunch and associated problems. The resolution of these problems is going to have a significant impact on the CPU business for both Intel and AMD, which makes predicting revenues nearly impossible (unless you’re Miss Cleo). Intel’s official predictions are for between $10.1 and $10.9 billion in revenue for Q4, a very weak number for what’s usually a very strong quarter. Intel has a strong product portfolio, but getting customers to buy is a whole other matter. Perhaps as an omen of things to come, sales were down in North America by 8.7%, the only region in which they fell.

POST A COMMENT

17 Comments

View All Comments

  • BSMonitor - Monday, October 20, 2008 - link

    "the company missed analyst estimates on revenue, where they were expected to bring in $10.25 billion"

    10.25 - 10.20 = $50 Million.

    Wow, rough quarter Intel... Missed "Average" revenue estimates by .5%..... $50 Million might be a lot to AMD, but it is nothiing compared to $10 billion...

    "even if they miss estimates elsewhere" HUH? Again.. $50 Million out of $10.2 Billion...

    "Intel’s official predictions are for between $10.1 and $10.9 billion in revenue for Q4, a very weak number for what’s usually a very strong quarter."

    Weak? $10.1 Billion in a collapsing economy? $2 Billion in Profit? Most companies would kill for those numbers.

    ====

    Now on the AMD side of the BS....
    "For the quarter AMD lost a paltry $67 million on revenues of $1.78 billion, completely blowing away their own predictions and what analysts were expecting, losing nearly $100 million less than expected."

    "The second source is a one-time license fee the company received for its 130nm SOI technology, which added $191 million to the bottom line. Finally, AMD’s graphics division turned a profit for the first time since they were purchased, posting a profit of $47 million on $385 million in revenue."

    So,

    Overall Company: -$67 Million
    ATI Profit: +$47 Million
    ==============
    AMD minus ATI -$114 Million
    License fee Profit +$191 Million
    ==============
    Old AMD -$305 Million!!!!!!!!

    So, how is AMD the processor company doing?

    -$305 Million.... And you are scoffing at $50 million ($30 million profit) in missed Intel revenue... What a joke, no wonder the idiots on Wall Street are losing their butts...
    Reply
  • Shadowmaster625 - Monday, October 20, 2008 - link

    Seems like NVDA is the real story to me. Seems like a steal at $7.50 a share. Looks like its heading up to the 50 dma, for a quick and easy 30% gain. Reply
  • InternetGeek - Monday, October 20, 2008 - link

    Ok, what I understand from this article is that AMD sold its foundry so it can get rid of the ATI debt. Hector must think that even if the company as a whole does well interests and debt will kill them. Which is why he sold the foundry and not the rest. It seems like a smart move if it pays well.

    Maybe in the future AMD will own its own fab again?

    Reply
  • Xavitar - Tuesday, October 21, 2008 - link

    There is a big difference between "selling off" and "spinning off". The fabs being spun off are in Germany. AMD also has a fab in Texas and one that is due to be built in New York which is not a part of the Foundry initiative, so they are not truly going fabless. Additionally, they will still retain something like a 48% ownership stake in the new Foundry company that operates the German fabs, and the management team currently running the German fabs is supposedly staying in place.

    In other words, this represents more of a shift in paperwork and legal entities than it does a major sell off of the company as you seem to be implying.
    Reply
  • Ryan Smith - Wednesday, October 22, 2008 - link

    The Texas facilities are also going to the Foundry Co and the New York project is being shifted to them too. AMD is getting rid of all of its CPU fabrication capacity. Reply
  • Justin Case - Sunday, October 19, 2008 - link

    [quote]The AMD/Foundry Co. spilt [/quote]

    It spilt? I had no idea AMD was that full.
    Reply
  • perzy - Sunday, October 19, 2008 - link

    This is great for my personal economy. Since there is almost zero progress there is no need for me to upgrade my cpu or, well anything. I bought one of those dual-cores last year, but it seems one of the cores hase'nt been used yet. None of my programs go any faster anyway.
    And since the graphics cards are so expensive, the kids can't afford them and the games follow the common nominator, so no progress there either. Besides, all the kids has moved to consoles because, yepp cards are more expensive than whole console.
    Reply
  • gvaley - Monday, October 20, 2008 - link

    The most sensible thing I've heard for a long time.
    You'd hear that competition is good for your pocket, repeated over and over again. And you'd believe it. But that's not true. The right statement is that competition is good for the technical progress.

    Think of it this way: with no competition you'll be spending $1000 every couple of years to have an up to date system.

    With no competition you'd be spending $2000 every four years to have an up to date system.

    Do the math yourselves.

    So the only difference is that you'd be able to play this game at 60 fps instead of 30.

    And yet, the latter is not entirely true, as game developers would have a much more predictable situation to deal with, and consequently the games would be far better suited to everyone's hardware than now. We've seen this happening before. And we were happier hardware users.
    Reply
  • AnnonymousCoward - Monday, October 20, 2008 - link

    > And since the graphics cards are so expensive

    What are you talking about--they're cheaper than ever. Get an 8800 off eBay for 100 bucks.
    Reply
  • Nfarce - Sunday, October 19, 2008 - link

    Well, it's now Q4 2008 and not a whisp of when Nehalem and X58 will be available by anyone, let alone when a full review will be available. I just blew a cool grand and some change replacing an older E6600/8800GTS rig with an E8500/260GTX rig. I got tired of waiting and bet that Nehalem would not be worth the wait. Hope I made the right choice.

    "Kids" have always been on the console, however. My online gaming with the PS3 vs. PC is like night and day, or adults and pre-pubescent punks. PC gaming always has been and always will be more expensive than console gaming, which keeps the juvenile zit faced kids away.
    Reply

Log in

Don't have an account? Sign up now