AMD's restructuring could be fab-less--or at least fab light

AMD's restructuring could be fab-less--or at least fab light

Summary: Advanced Micro Devices (AMD) reported a whopping loss, was crushed by Intel and flopped in the first quarter under any metric you choose. The good news: It's promising a game-changing restructuring.

TOPICS: Processors

Advanced Micro Devices (AMD) reported a whopping loss, was crushed by Intel and flopped in the first quarter under any metric you choose. The good news: It's promising a game-changing restructuring.

AMD, which reported a first quarter net loss of $611 million, laid out a silver lining that looks like this: revenue will be flat with the first quarter tally of $1.23 billion (I don't buy that) and the company is cooking up an aggressive plan to fix things.

What would this restructuring look like? Of course, it'll feature the usual standbys of layoffs and cost cuts. But beyond that AMD talked a good bit about an "asset light" model. In English, that means AMD is going to outsource more of its manufacturing to firms with better fabrication plants (fabs). The move makes sense in many ways. For starters, fabs are a cash drain, suck up research and development and may be a futile effort considering Intel has more resources. Why not outsource key manufacturing processes?

On a conference call with analysts, AMD CEO Hector Ruiz outlined some of the restructuring parameters.

"The level of restructuring that I am envisioning is very significant, and so it would be difficult to outline it on a phone call. But let me give it a try as to what we intend to discuss later on in the year.

For example, four years ago, we had one product in the company that, through tweaking and maneuvering, we could actually make it serve various segments of the market. Today, we have proliferated on a product line towards goodness. We have a broad array of products, but now we are serving segments in the industry that each of these segments are large in size. For example, an entry level segment in an emerging market is a very large segment. All on its own, it actually requires its own separate business model in how to address that segment. Compare that to a workstation or super computer segment, which all on its own requires a completely different business model."

Ruiz also noted that AMD used to make chips and put them in the channel, but as its products have increased the company is more about partnering with OEMs. In the end, Ruiz said each product line will have a different management nuances.

"As part of that, one of the things that has become pretty clear in our experience, we have had now for a number of years had some experience in partnering with people such as, for example, IBM in joint development programs. That has started a lot about how you can do some asset light strategies, since we did not have to build an R&D laboratory to do that.

We have had also an experience now for a number of years with some of our friends in the foundry business, and in particular Charter Semiconductor. We have learned a lot of that.

Through the acquisition of ATI, we now have a perspective into a very asset light model that we are pretty excited about learning more and more about it. When we look at all these things, we see a tremendous opportunity for us to really do something different going forward, but it’s unique for us."

Jeffries analyst John Lau said in a research note that a manufacturing retooling may offer the biggest returns out of AMD's turnaround initiatives. "A hybrid or a fab-lite model should enable AMD to cut down on its R&D and capital expenditure and improve its return on capital. Given AMD's scale and the challenges in growing its scale, a fab-lite model is likely to create most value," said Lau.

Not all agree. CIBC analyst Richard Schafer said in a research note that relying on outside manufacturers may ensure Intel's generation lead indefinitely. "AMD's cap-ex reduction plan and 'asset-lite' strategy portend increased reliance on foundry partners. We believe such a move would perpetuate Intel's generation lead indefinitely," said Schafer. 

Lau, however, noted that foundries such as TSMC are moving in lock-step with Intel as it moves to smaller manufacturing processes.

In the end though, AMD has little choice but to try an asset light strategy. It can't keep losing north of $600 million a quarter when it only has $1.2 billion in cash as of March 30. Bottom line: AMD will need more capital. The big question is where the capital comes from. 

If there were ever a company that should go private, restructure and re-emerge like Seagate did it's AMD. Ruiz didn't shoot the going private idea down either. "We have absolutely no prejudice or bias towards the source of capital as long as it makes sense for us, and we are very open to any of those ideas," said Ruiz.

Topic: Processors

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • "Today, we have proliferated on a product line towards goodness."

    AMD is not going to recover by relying on IBM for R&D, nor on other companies to make its products. To me, when Apple began to purchase all its hardware components, that was an admission that the design a user sees is more important than the design he doesn't. Control over the manufacturing process can be the easiest, fastest way to implement a good idea.

    Any "restructuring" can be a way to solve problems with products and sales by getting rid of the people who can solve problems with products and sales. If AMD follows that route, the CEO can expect a large bonus and the company deserves what will inevitably happen to it.

    Maybe a place to start a real restructuring is for AMD to examine whether it wants to proliferate products. The quotes in the article did indicate that different products have different business plans. Maybe the company should eliminate the products with hopeless business plans, and keep the promising lines. The result may not be as big a company, but profitabilit does a lot to make up for limiting growth.
    Anton Philidor
  • Beginning of the end.

    Sorry, but if you are going to be in the CPU business the ONLY real competitive tool is R&D. With AMD struggling to get to 65nm and Intel well on the way to 45nm the writing is on the wall. Intel will produce faster, more energy efficent, and cheaper CPUs.

    Perhaps AMD can corner the market on the low end CPUs where speed and power use are not primary issues. But even then, if Intel is making the big bucks on the high end they can afford to sell low cost CPUs into that same market and AMD will slowly fade away.
    • RE: Beginning of the end.

      AMD is not talking about getting rid of R&D. They are talking about doing it cheaper.
      • Sorry but no.

        They are trying to farm out the R&D needed to manufacure CPUs. Yes, the manufacturing requires huge amounts of R&D AFTER you have the chip designed. That is why they have failed to move to 65nm and can't even think about 45nm right now.

        That alone will prevent them form lowering the cost of the products. (Waffers are VERY expensive.)
        • We'll see about that

          Time will tell, I suppose, but just because you say the restructuring won't work doesn't make it so. I think AMD is still very interested in staying in business and will find a way to keep afloat.
          • What do you call a manufacture that makes nothing?

            Answer - Broke and out of business.
          • So AMD makes nothing?! Humm...

            The still offer Athlon 64 X2, Opteron, Turion 64 and Sempron processors last time I looked. And that's just their cpu market.

            They also own ATI, which not only produces some great video cards for desktop computers but is more often found in laptops with better graphics solutions than rival NVIDIA.

            Anyway, you may think that's just laptops. Well, the more we go mobile, and we can get TV on cell phones now, the greater need for laptop graphics solutions.

            Now don't get me wrong. I like both Intel and AMD. I have computers in the house with both companies? processors. My machine runs a nice (and it is fast) Athlon 64 X2 4800+ cpu. This cpu is fast enough for me do ANYTHING on my computer. I don't have to wait for anything to finish. Our other computers are older slow P4 systems that will still run Vista! Sure, an Intel Core 2 Duo might be faster than what I have now but will I notice the difference, except in the price?

            Each of my children by age 17 get a laptop for their birthday. I plan on buying one for my daughter's August birthday with a fast AMD Turion 64 processor, which will save me at least $200 over any Intel Core 2 Duo system. What's more important for this new laptop is that I get 2GB of RAM since it will come with Vista Home Premium. I know the computer will need that RAM!

            Now, it should obvious, and go without saying, AMD does manufacture and sell products. Consumers still buy them, just not in the numbers when AMD was leading. Sure, this has hurt AMD's business but with their acquisition of ATI, they will have their foot in the laptop market with or without their laptop processors.

            Too, I understand there are plans to develop a cpu with graphics integrated in the chip. With the trend to downsize media solutions, do you see any market for such a product? Say, cell phones and other handhelds? Business is very much into small in size but with great function.

            AMD is not dead nor do they manufacture nothing. You must be without vision - don't worry, your nose is at the end of face. I would image you can see that far. :)
    • looks that way ...

      Different CPUs serve different needs. 30 years ago there was the Zilog Z80 which was a very popular chip for the consumer market (1MHz - 8MHz models). It has since mutated beyond recognition in chips such as the "Rabbit4000" (50MHz) but its low power consumption makes it great for many embedded applications. One step up from that we have the XScale and ARM processors which run to about 200MHz; 'ColdFire' and 'Dragon' go in that class of capabilities. Then you've got the AMD Geode which can run up to 500MHz but still consumes very little power. Then you get to the VIA processors and Intel Pentium M/Celeron at 500-1600MHz which are getting power-hungry but are still extremely useful for some things (and much better than my desktop machine 10 years ago), and finally to the huge power-hungry desktop and server CPUs.

      So AMD have very strong competition at all levels of the CPU market, with perhaps the Geode having the strongest bet in that gap between XScale and Pentium M (but how big is that market).

      The big question I think is: What market is AMD trying to capture? No R&D = no AMD so I think they need to maintain that, but for some of its product lines they might be better off selling the information (or even the division/unit) to a competitor. Ruiz is giving nothing but "management speak" though - he certainly hasn't enlightened anyone on what's likely to happen. The CEOs get incredible payouts for mergers/spinoffs - you name it, they've got a ridiculous cut. Over the past 15 years I've seen most US company mergers/buyouts/etc as nothing more than a means to get some CEO an insane payout and this is really not good for the industry.

      Yeah, maybe AMD will become just another video card manufacturer.
  • DEC/Compaq tried this

    DEC tried going fabless during their collapse. We didn't recognise it as a collapse, then, and whether this was a cause or not it's hard to say... but one does wonder whether AMD will end up just making video cards in a few years.
    • I miss the Alpha...

      Intel and AMD developed technology to the point where people said "why am I paying so much for an Alpha?" For the typical processing I do it takes a 1.6GHz Pentium to outdo the Alpha on calculations and I often wish desktop CPUs had a more capable 64-bit math unit. But DEC and other still (barely) existing companies such as SGI were victims of the advance of technology. The "minicomputer" had no more role in corporations and even the "supercomputers" could now be built up from CPUs aimed at the desktop. DEC did not improve the Alpha to match those other advances so they had nothing.

      Sun still survives but they're having a tough time as well - they may have to move out of the chip business and into improving Solaris to run on more architectures etc. (although in some ways they have a lot of catching up to do with Linux).
      • Alpha isn't a mini

        Alpha was a microprocessor, there's nothing of the "minicomputer" inherent in the Alpha model, and in fact all the way up to the DS and ES series the difference between an Alpha based system and a high end x86 based system were minimal. The GS was a different kettle of fish, but then so are the clustered high-availability x86-based boxes.

        Alpha didn't fall behind until after Compaq bought DEC, and the Compaq-HP deal was under way. They killed EV8 development right before that was announced, barely letting the ink dry on the obituary before "coincidentally" deciding to dump the Alpha and Tru64 and take a giant leap into the past on the Itanic.

        And now the EV8 multicore model ... more work per clock at a lower clock speed ... shows up in Intel's court. Years later than it would have been if DEC hadn't been dumped in the Compaqter.
  • hope not

    If they go B/K. Intel will rape everyone! Fact they did it before AMD introduced the Athlon & gave them a run for their money. Maybe stock up on X2-6000+ like Newegg.
    • Why stock up on

      old tech when the Intel chips are cheaper, use less power, and perform better.

      That's like saying lets stock up on early 70s corvetts with 454 big blocks that produced 175 HP (smog controls) and got 12 miles to the gallon...
      • Hummm...

        You just can't see past your nose, can you? ;)
    • AMD Athlon X2's are great for home users!

      I ran Intel Pentiums all the way to the P4 line. Then I bought my first AMD Athlon 64 X2 4800+ system and have never looked back!

      Sure, I have 4GB RAM. I have 2x74GB WD Raptor 10,000rpm hard drives running stipped in RAID 0 setup. And I have a ONLY one GeForce 7800GTX v-card but could have 2. My mo-board permits all this. But I have never been so pleased with the processor.

      I don't have to wait for anything except maybe downloads. But that wait with my cable Internet service is almost always short.

      AMD can certainly play in the low cost home computer user market. I would certainly buy one again, and I plan to purchase my daughter a laptop with a fast AMD Turion 64 processor in it, which will save me between $200 and $300 compared to an Intel Core 2 Duo system with ALL the same components.
  • asset light, but not for everything

    Don't swing the argument to the absurd just to score a punchline, please. Did AMD
    say it's going 100% asset-light? Did you listen to the call? They're going to look at
    ways to optimize for each business segment, not asset-light as an entirety.
    (Comparison was made to TI abandoning its mfg after 45nm, though not entirely an
    accurate one)
    • was reply to NoAxeToGrind

      Sorry, wrong clicky..that was meant to respond to NoAxeToGrind's post about "what
      do you call a manufacture [sic] that makes nothing"