OS Driven Power Management

When Intel introduced Nehalem and the Core i7, we saw a new generation of power engineering in microprocessors. In the past, the OS would request a particular performance state from the CPU and the chip would respond by changing its clock speed. Nehalem’s Power Control Unit (PCU) instead dedicated enough transistors to build a 486 to monitoring the power and performance demands on the chip. Based on those demands and what the OS was doing, the PCU would power up or down individual cores, as well as move clock speeds up or down. The PCU would guess at what the OS was trying to do and respond accordingly.

Nehalem and its successors were massive chips, eating up to 130W of power under load and idling down in the 6 - 10W range. Lincroft has to be sub-1W under load and 6mW at idle. With even more stringent power demands and a much smaller die, Intel couldn’t blow a sizeable percentage of the Lincroft transistor budget on power management.

Instead of guessing at what the OS wants, the Moorestown platform uses OS Driven Power Management (OSPM) to tell Lincroft and Langwell what to do. OSPM is supported in Moblin and presumably the Wind River build of Android.

The OSPM process tells the hardware what apps it’s running and to shut down what it doesn’t need. There are well defined operating modes - standby, internet browsing, MP3 playback, video playback, voice call, video capture, etc... Based on the profile, the hardware doesn’t have to guess at what it should turn off/on, it just does it right away.

The OSPM driver communicates directly with the two power management units in Moorestown - one in Lincroft and one in Langwell. It instructs those PMUs to shut off various blocks, and in turn they tell Briertown to gate and cut voltage to the parts of chip that aren’t needed.

I wondered if this couldn’t be done in hardware, but it seems that given current die constraints and the sort of accuracy of information it needs Intel must implement at least some of the power management control in software. Toolkits will be available for developers to control the OSPM.

Power Gating Putting Power in Perspective: Estimated Battery Life of a Moorestown Phone
Comments Locked

67 Comments

View All Comments

  • DanNeely - Wednesday, May 5, 2010 - link

    I think you're misunderstanding the slide. It's not saying 1024x600 to 1366x768, it's saying upto 1366x768 on interface A, upto 1024x600 on interface B.
  • Mike1111 - Wednesday, May 5, 2010 - link

    Thanks for the clarification. Looks like I really misunderstood this sentence:
    "Lincroft only supports two display interfaces: 1024 x 600 over MIPI (lower power display interface) or 1366 x 768 over LVDS (for tablets/smartbooks/netbooks)."
  • uibo - Wednesday, May 5, 2010 - link

    I wonder how many transistors are there in a Cortex A9 core? Just the core nothing else.
    For me it seems that ARM could just double or quadruple their core count against the Intel solution while still maintaining lower transistor count.
    Also they could just increase the CPU clock speed, if there is a market for the more power-hungry Intel solution the there is one for the ARM also.
  • strikeback03 - Wednesday, May 5, 2010 - link

    I would imagine even less smartphone software is written for multi-core now than was for desktop when dual-core CPUs started appearing in desktops. So going beyond 2 cores at this time is probably not a great move. Plus the dual core A9 isn't out to see power consumption yet, but even at 45nm I doubt it will be much below the current 65nm single-core chips if at all, so if Intel is already competitive then ARM doesn't exactly have the power budget to add cores.
  • uibo - Thursday, May 6, 2010 - link

    That actually makes sense. Nobody is going to write multi-threaded apps for a single thread CPU. I'd imagine that the number of apps, which experience is hindered by performance, is not that great at the moment. Games, browsers, UI, database for the info stored in your device - I'm not expecting these to scale perfectly across many cores but do expect a x0% performance increase.
  • DanNeely - Thursday, May 6, 2010 - link

    The real benefit for the 2nd core is probably multi-tasking. Your streaming music app can run in the background on the second core while your browser still has a full core to render web pages.
  • Shadowmaster625 - Wednesday, May 5, 2010 - link

    Mooresetown has to support a desktop OS. Intel is clearly moving towards wireless computing. They are bringing wireless video. With wireless video you can turn your phone into a desktop pc instantly by adding a wireless monitor and keyboard. What is the point of moving in that direction if you're moving towards a crippled OS? (Not that windows isnt crippled, if you consider obesity a form of cripple.)

    If it needs a pci bus, then emulate one!
  • Caddish - Wednesday, May 5, 2010 - link

    Just registered to say keep up the good work. Since the SSD antology I have red all of your article like that one and they are awesome
  • legoman666 - Wednesday, May 5, 2010 - link

    Excellent article, very well written.
  • jasperjones - Wednesday, May 5, 2010 - link

    Anand,

    You mention twice in the article that Apple and Google dominate the smartphone market. This is utter nonsense. The numbers from IDC as well as the numbers from Canalys clearly show that Nokia is the worldwide leader in the smartphone market. RIM is number 2. Apple is in the third place, the first company that produces Android devices, HTC, has the number 4 spot.

    I realize that Nokia's market share in the U.S. is smaller than its global market share. However, even if we restrict ourselves to the U.S. market, RIM smartphone sales are bigger than those of Apple. They are also bigger than the sales of all Android smartphones combined.

Log in

Don't have an account? Sign up now