Power Consumption

There's a lot of uncertainty around whether or not Kepler is suitable for ultra low power operation, especially given that we've only seen it in relatively high TDP (compared to tablets/smartphones) PCs. NVIDIA hoped to put those concerns to rest with a quick GLBenchmark 2.7 demo at Siggraph. The demo pitted an iPad 4 against a Logan development platform, with Logan's Kepler GPU clocked low enough to equal the performance of the iPad 4. The low clock speed does put Kepler at an advantage as it can run at a lower voltage as well, so the comparison is definitely one you'd expect NVIDIA to win. 

Unlike Tegra 3, Logan includes a single voltage rail that feeds just the GPU. NVIDIA instrumented this voltage rail and measured power consumption while running the offscreen 1080p T-Rex HD test in GLB2.7. Isolating GPU power alone, NVIDIA measured around 900mW for Logan's Kepler implementation running at iPad 4 performance levels (potentially as little as 1/5 of Logan's peak performance). NVIDIA also attempted to find and isolate the GPU power rail going into Apple's A6X (using a similar approach to what we documented here), and came up with an average GPU power value of around 2.6W. 

I won't focus too much on the GPU power comparison as I don't know what else (if anything) Apple hangs off of its GPU power rail, but the most important takeaway here is that Kepler seems capable of scaling down to below 1W. In reality NVIDIA wouldn't ship Logan with a < 1W Kepler implementation, so we'll likely see higher performance (and power consumption) in shipping devices. If these numbers are believable, you could see roughly 2x the performance of an iPad 4 in a Logan based smartphone, and 4 - 5x the performance of an iPad 4 in a Logan tablet - in as little as 12 months from now if NVIDIA can ship this thing on time.

If NVIDIA's A6X power comparison is truly apples-to-apples, then it would be a huge testament to the power efficiency of NVIDIA's mobile Kepler architecture. Given the recent announcement of NVIDIA's willingness to license Kepler IP to any company who wants it, this demo seems very well planned. 

NVIDIA did some work to make Kepler suitable for low power, but it's my understanding that the underlying architecture isn't vastly different from what we have in notebooks and desktops today. Mobile Kepler retains all of the graphics features as its bigger counterparts, although I'm guessing things like FP64 CUDA cores are gone.

Final Words

For the past couple of years we've been talking about a point in the future when it'll be possible to start playing console class games (Xbox 360/PS3) on mobile devices. We're almost there. The move to Kepler with Logan is a big deal for NVIDIA. It finally modernizes NVIDIA's ultra mobile GPU, bringing graphics API partity to everything from smartphones to high-end desktop PCs. This is a huge step for game developers looking to target multiple platforms. It's also a big deal for mobile OS vendors and device makers looking to capitalize on gaming as a way of encouraging future smartphone and tablet upgrades. As smartphone and tablet upgrade cycles slow down, pushing high-end gaming to customers will become a more attractive option for device makers.

Logan is expected to ship in the first half of 2014. With early silicon back now, I think 10 - 12 months from now is a reasonable estimate. There is the unavoidable fact that we haven't even seen Tegra 4 devices on the market yet and NVIDIA is already talking about Logan. Everything I've heard points to Tegra 4 being on the schedule for a bunch of device wins, but delays on NVIDIA's part forced it to be designed out. Other than drumming up IP licensing business, I wonder if that's another reason why we're seeing a very public demo of Logan now - to show the health of early silicon. There's also a concern about process node. Logan will likely ship at 28nm next year, just before the transition to 20nm. If NVIDIA is late with Logan, we could have another Tegra 3 situation where NVIDIA is shipping on an older process technology.

Regardless of process tech however, Kepler's power story in ultra mobile seems great. I really didn't believe the GLBenchmark data when I first saw it. I showed it to Ryan Smith, our Senior GPU Editor, and even he didn't believe it. If NVIDIA is indeed able to get iPad 4 levels of graphics performance at less than 1W (and presumably much more performance in the 2.5 - 5W range) it looks like Kepler will do extremely well in mobile.

Whatever NVIDIA's reasons for showing off Logan now, the result is something that I'm very excited about. A mobile SoC with NVIDIA's latest GPU architecture is exactly what we've been waiting for. 

Introduction
POST A COMMENT

141 Comments

View All Comments

  • yhselp - Thursday, July 25, 2013 - link

    At some point in the near future, if Intel doesn't decide to do it themselves, would it be possible for an OEM to licence NVIDIA IP and integrate it into an Intel ultra mobile design? Reply
  • JlHADJOE - Friday, July 26, 2013 - link

    Now that would be interesting.

    Maybe Nokia would do it for an upcoming W8 phone? Silvermont Atom + Logan. Bringing Windows, Intel and Nvidia from your desktop to your phone.
    Reply
  • yhselp - Friday, July 26, 2013 - link

    Sound good, but there might be a few problems. Microsoft wants homogeneity between Windows phones and so they set requirements for the SoC (among other things). Back in the days of WP7 these rules were quite strict which meant an OEM didn't have complete freedom in choosing the SoC. Nowadays, as far as I understand, there's only a set of minimum system requirements that an OEM has to meet. An Intel/NVIDIA SoC would obviously be more than powerful enough, but I wonder whether Microsoft would have anything to say about such an implementation. Furthermore, there's the question of the benefits of all this; while the NT kernel is there, the mobile OS would need some work to make proper use of all that power. Not to mention, having the same architecture and API doesn't immediately translate to running the exact same software from Windows 'PC' on Windows 'mobile'.

    A Silvermont/Logan implementation, while great, is not that exciting. Next-gen Silvermont (hopefully wider) + Maxwell on smaller fab would be quite interesting.
    Reply
  • HighTech4US - Thursday, July 25, 2013 - link

    Anand: NVIDIA got Logan silicon back from the fabs around 3 weeks ago, making it almost certain that we're dealing with some form of 28nm silicon here and not early 20nm samples.

    I believe this is a wrong assumption and that the Logan sample is on 20nm.

    As of April 14, 2013 TSMC has had 20mn risk production available. More than enough time for Logan to be produced on 20nm.

    http://www.cadence.com/Community/blogs/ii/archive/...

    Quote: While TSMC has four "flavors" of its 28nm process, there is one 20nm process, 20SoC. "20nm planar HKMG [high-k metal gate] technology has already passed risk production with a very high yield and we are preparing for a very steep ramp in two GIGAFABs," Sun said.

    Quote: Sun noted that 20SoC uses "second generation," gate-last HMKG technology and uses 64nm interconnect. Compared to the 28HPM process, it can offer a 20% speed improvement and 30% power reduction, in addition to a 1.9X density increase. Nearly 1,000 TSMC engineers are preparing for a "steep ramp" of this technology.
    Reply
  • Refuge - Thursday, July 25, 2013 - link

    I think you have high hopes good sir, but I disagree with you on this one.

    I would be struck silly if this came on a 20nm process when released.
    Reply
  • watersb - Friday, July 26, 2013 - link

    Crikey. Reply
  • wizfactor - Friday, July 26, 2013 - link

    Those are some fantastic numbers! While I'm all for seeing Logan in future SOCs, I'm not a huge fan of seeing them in Tegra. If developers need to rebuild their mobile games (see Riptide GP for Android) just to optimize it for your chip, you're doing something wrong. I have yet to hear anecdotes about how pleasant an experience it is to port an Android game to a proprietary chip such as Tegra.

    With that said, I'd love to see this GPU on other chipsets such as the Exynos, or even Apple's A-series chips. I can't help but think that Nvidia is teasing Apple into a license agreement here. I mean, the very fact that Apple could get more than double the graphics performance on their iPad 4 with a Kepler GPU under the exact same power constraint must ring music to their ears. They could either dramatically increase iPad performance and eliminate any performance woes involved with driving that Retina Display, or they could get a massive boost in battery life while keeping performance levels similar.

    Of course, it's Apple's call if they want to swap out Imagination for Nvidia. Let's hope Cupertino isn't too attached to its investment in the former.
    Reply
  • michael2k - Friday, July 26, 2013 - link

    Why would they want to swap out with NVIDIA? The PowerVR Series 6, of comparable performance, was available for license last January, and expected to be out in production this year. What you miss is that with the Power VR S6 they can get more than four times the graphics performance of an iPad 4 one year earlier than with a Kepler GPU; why would they wait a year then? Reply
  • jipe4153 - Tuesday, July 30, 2013 - link

    Their isn't a single high end PowerVR 6 such as the G6630 planned for production yet. It's just been paper launched.

    - No engineering samples, nothing!
    - Just numbers on a paper!

    Furthermore their peak paper product, the G6630 is slated for a peak performance of ~230 GFLOPS, which is almost half the performance that logan is sporting!

    Logan has the winning recipe:

    - Most powerful hardware
    - Good efficiency
    - Best sofware and driver stack on the market!

    This is going to be a major upset on the mobile market.
    Reply
  • phoenix_rizzen - Friday, December 20, 2013 - link

    And yet, Apple shipped with a series 6 GPU and not an nVidia one. Where's the major upset? Reply

Log in

Don't have an account? Sign up now