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

  • djboxbaba - Wednesday, July 24, 2013 - link

    This is ALL relative, does anyone realize that this will not be released till 2014? ......2014? the A8/A8X (or equivalent) will be released by then, as well as Qualcomm's latest snapdragon processors. Don't place your expectations too high. Reply
  • Scannall - Wednesday, July 24, 2013 - link

    I am wondering if they will be too late to market again. PowerVR ssries 6 (Rogue) devices should be shipping any time.

    http://en.wikipedia.org/wiki/PowerVR#Series_6_.28R...
    Reply
  • NLPsajeeth - Wednesday, July 24, 2013 - link

    Seems like the most similar desktop part is the GeForce GT 630 OEM
    http://www.geforce.com/hardware/desktop-gpus/gefor...

    And laptop somewhere between 720M and 730M.
    Reply
  • randomhkkid - Wednesday, July 24, 2013 - link

    Think about the implications if they can get just shy of 720m - 730m performance at about 5W what can they do with laptop GPU at around 35W O.o Reply
  • Spunjji - Wednesday, July 24, 2013 - link

    We already have Kepler at that power level, though. So nothing that they haven't done already... Reply
  • randomhkkid - Wednesday, July 24, 2013 - link

    I mean that performance currently hasn't scaled with the power, i'm assuming that the current 720m has a power envelope of around 35W and yet there is a ~5W kepler GPU that has around the same performance. If they could scale it up while keeping the efficiency it would be a sight to behold. Reply
  • DanNeely - Wednesday, July 24, 2013 - link

    I think you're way high on the 720M TDP. It's a GF117 part. The other GF117 parts are 12.5W (710M) or 15W (620M and 625M).

    35W is probably a bit too high for the 730M too, it's a GK208 part and the Quadro 510M/610M (only mobile GK208's I can find TDP for) run at 30W,.
    Reply
  • Jaybus - Wednesday, July 24, 2013 - link

    For the same architecture, it is true that performance scales linearly with clock speed. However it is not true that power consumption scales linearly with clock speed. 5 W at 500 MHz does not equate to 25 W at 2.5 GHz. Reply
  • roberto.tomas - Wednesday, July 24, 2013 - link

    This mostly looks very rosy for Nvidia's future IP in mobile space. If Logan comes at 28nm though, it *does* face the same issue as Tegra 3 had — unless it comes to market not in 2014 but in 2013.

    As a rough estimate, at 28nm and from the power shown, to achieve max performance they need 25w on the gpu power meter. That's a bit of a shame, because that is clearly outside the bounds of minimal cooling.

    On the plus side, that is a tonnage of power, which companies can tune to use as much or little of as they want, all the way up into the absurd levels — rather like the PowerVR Rogue. In that sense it looks good for NVidia.
    Reply
  • sontin - Wednesday, July 24, 2013 - link

    What? If they can get 76GFLOPs/s in less than 1W they do not need 25W for 5x the performance...

    And the remark about 28m: Every company will supply constraint with 20nm in 2014. And Tegra 3 was a huge success for nVidia (50% revenue increase in FY2013) that they can sell it for a lower price (cheaper wafer) and in huge numbers (more wafers, better yields).
    And do not forget: 20nm will only bring a 30% lower power consumption. So if you can archive this with an optimize architectures there is no benefit for using the 20nm from the start for Tegra.
    Reply

Log in

Don't have an account? Sign up now