IHS iSuppli: Intel gained CPU revenue share last quarter

It ain’t easy being green. Although recent figures suggest AMD grew its share of x86 CPU shipments last quarter, IHS iSuppli now reports that the underdog saw its slice of overall CPU revenue shrink in Q2—at least compared to the same quarter a year back. Here are the research firm’s numbers:

  Q2 2010 Q1 2011 Q2 2011
Intel 80.7% 82.6% 81.8%
AMD 11.6% 10.1% 10.4%
Others 7.7% 7.3% 7.7%

(For the record, these figures include revenue from all types of CPUs, including "RISC and other types of general-purpose microprocessors," not just x86 chips.)

To AMD’s credit, its sequential performance wasn’t too bad. The chipmaker’s revenue share still tumbled by over one percentage point compared to last year, though.

IHS iSuppli says Intel performed particularly well in the second quarter because of a "combination of a recovery in PC demand and strong shipment growth for its new Sandy Bridge line of microprocessors." In particular, Intel is said to have benefited from strong sales in the corporate PC market, which is outpacing the consumer PC market.

Comments closed
    • Buzzard44
    • 8 years ago

    Is this article just based on really outdated information, or does it really take 2 MONTHS AFTER you’ve calculated market share to calculate revenue share? (see link in article) What gives? This makes no sense. Someone please explain. Seems like that information would be calculated at about the same time.

      • dpaus
      • 8 years ago

      If I’m not mistaken, the new/current accounting guidelines require the revenue to be [i<]received[/i<] before you can declare it. And if they're actually collecting all that revenue in less than 60 days, hell, I'm impressed.

    • codedivine
    • 8 years ago

    Lots of interesting questions and discussions about this:

    1. Assuming this includes revenue from ARM shipments as well, its interesting to see that even AMD made more revenue than all non-x86 CPU vendors combined.

    2. How do they calculate the CPU revenue for chips with more-than-a-cpu on the same die? For example, if you take a Tegra 2, Llano or sandy bridge, how do you calculate its “CPU revenue”?

      • Game_boy
      • 8 years ago

      It doesn’t include ARM.

      Even if it did, ARM’s revenues on chips are only pennies per unit because the implementation and manufacture is done by ARM licencees.

      And they include the GPU part in the revenue, no special treatment. They also count SB and the others in graphics shipments which is why Intel dominates those figures.

        • DavidC1
        • 8 years ago

        Pennies? Why would they have to include ARM Holdings share of the pie? They just have to get the licencees like TI/Nvidia/Qualcomm etc.

        And for #2 of codedivine’s question: It doesn’t matter. When you integrate something into a CPU, it becomes part of a CPU. Just like when FPUs went from a seperate chip into a processor, they did not count revenue seperately.

    • lycium
    • 8 years ago

    This is why they don’t need to put out IVB yet :'(

      • NeelyCam
      • 8 years ago

      Philosophical question: if you’re a monopoly because your opponent screws up so bad they have to quit, does it mean you’re ‘evil’ because all monopolies are ‘evil’? Or are you still ‘good’ because it wasn’t your fault?

        • lycium
        • 8 years ago

        Intel definitely did some dirty things (apparently going so far as to destroy evidence) during the Pentium 4 days.

          • NeelyCam
          • 8 years ago

          Intel [i<]allegedly[/i<] did some bad things back then. But I'm talking now. If AMD implodes because of its own mistakes, will people call Intel evil and claim Intel sabotaged AMD somehow?

    • Game_boy
    • 8 years ago

    Unless CPU revenue share is seasonal, why are you comparing it to the same quarter a year ago rather than last quarter in your headline?

      • NeelyCam
      • 8 years ago

      I think the numbers in the table might be unit share – not revenue share…

        • Peldor
        • 8 years ago

        No, way. Unit share is ~19% for AMD. Their chips are cheap and their revenue share is much lower.

        And yes, CPU sales are seasonal to some extent (along with practically everything else). It’s virtually universal for companies and analysts to do quarterly and yearly comparisons.

          • Game_boy
          • 8 years ago

          CPU /sales/ are seasonal, yes. I can’t see a reason why revenue share would be though.

            • Phr3dly
            • 8 years ago

            Different customers have different buying patterns. Corporations spend a lot in Q1, and less in Q2. Intel has a higher market share for that sector, so it’s revenue share will be higher in Q1 and lower in Q2.

            • Game_boy
            • 8 years ago

            Good point. Thanks.

          • NeelyCam
          • 8 years ago

          Ah – you’re right. I should’ve clicked on that link…

      • sschaem
      • 8 years ago

      ….

      • esterhasz
      • 8 years ago

      You’re of course completely right but there may even be a little bit of seasonality in the Intel/AMD split. Consumer products and enterprise tech certainly have different seasonal patterns and with AMD more or less out of the picture on the server side for the moment, these may shine through a little bit. With the market being so global now (for PCs the national markets now rank China>US>Brazil>Germany) I guess that seasonality is less and less significant.

      • HighTech4US2
      • 8 years ago

      > why are you comparing it to the same quarter a year ago rather than last quarter in your headline?

      Because it clearly shows if a company is growing or contracting.

      Revenue share is a better measure of a companies worth and Unit share by itself is meaning less.

      You can grow Unit share and still lose in Revenue share if you have to sell your CPU’s at a much lower ASP.

      AMD’s problem is that they have to sell at a low price because they can’t compete with Intel in performance.

      And with today’s other headline it looks like the next quarter will be even worse.

        • Game_boy
        • 8 years ago

        Did you actually read the question?

          • HighTech4US2
          • 8 years ago

          Yes I did and your question that only sequential quarterly results should be counted and Y-on-Y quarterly results should be ignored is absolutely wrong and I stated why it is wrong.

          The financial results of all tech companies have always shown Q-to-Q and Y-to-Y Quarterly results. You can’t just cheery pick the one you like whenever one looks better than the other.

          The Y-on-Y Quarterly results is actually a better measure exactly because it removes seasonality.

      • Palek
      • 8 years ago

      I did a double-take, too. Unless explicitly stated, I think the average person will interpret the title as meaning that intel gained revenue in Q2 over Q1, that is, sequentially. Just a minor quibble, though, and the table clarifies it all.

Pin It on Pinterest

Share This