Twelve days ago, an amazing thing happened.

Project Kal-El came into the building, shepherded by a team of engineers from around the world, who had been working around the clock for months.  For the uninitiated, the entrance that I’m referring to was by none other than the next generation of our Tegra super chip.

Today at Mobile World Congress (MWC), we demonstrated this little beauty running in an Android tablet. We not only showed that it was alive. We showed it browsing the Web, running games and streaming amazing video. This wasn’t your average amazing video. It was 1440p video content running on a 2560×1600 panel. That will enable mobile devices to output to the highest resolution monitors or tablets equipped with a 10.1-inch display with 300 DPI.

I should mention, in passing, that this is the first mobile quad-core processor. And that it contains a new 12 core GeForce GPU. Our customers are getting samples now, and they’re planning production in August.

While demonstrating Project Kal-El was exciting, we also gave a glimpse of our roadmap here at MWC. It includes Projects codenamed Wayne, Logan, and Stark, coming out in a steady one-year cadence over the next three years. You might well ask, What on earth can be done with nearly 75x improvement in performance over Tegra 2 that Stark will provide in 2014? Our customers and partners have already indicated that they’re confident they can use everything we give them.

If you’ve followed NVIDIA, you know the relentless velocity of innovation that we’ve  brought to the PC space with our GeForce processors. Prepare for that same intensity to play out in the mobile space.

While we demonstrated this to our customers and press today, you can come by and see it live in our booth at MWC in Barcelona tomorrow.

Quad-Core CPU performance of Project Kal-El

I left CES thinking this was going to be the most exciting year of my career. Now, I know it will be.

I can’t wait to see the next wave of superphones and super tablets.

  • http://www.brightsideofnews.com ANshel

    Loving the code names :)

  • http://mobile-device.biz Scott B

    Comical

  • CHRISS MARTIN

    i think nvidia’s KAL-EL chip will be in Sony’s NGP……….cause it says a quad core chip…..the only quad core SoC known now is Kal-El………

  • Michael Simmons

    OpenCL? CUDA?

  • http://ficforlife.blogspot.com Hok

    Waiting GT 540 equivalent aka entry level or value version of this “KAL-EL” so it can be sold at $80 or less than $100 handphone (not chip only).

  • http://none JOSUE

    sorry Intel. Nvidia You Rocks!!!!!!!!!!!! Now the serius question and when on the PC market!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!?????????????????

  • http://www.cnx-software.com cnxsoft

    I can’t believe you plan a processor with 75x the performance of Tegra 2 by 2014. That is MAD, I don’t know what they will do with that much power!

  • Sizz

    Awesome code names….

  • http://ficforlife.blogspot.com Hok

    Btw to all monitor Vendor, please cooperate with Nvidia to make monitor beyond 2500 x 1600 like in this article state. What is the benefit of having Kal-El able to provide 2560 x 1600 with no monitor that support that resolution.

  • Lexusprada

    Sorry HOK apple cinema display 27″ and 30″ can

  • Igor S

    Unexpected!, excellent!, stunning!, awesome!, powerful! …
    Please bring it to market as quickly as you will be able.

  • Bob

    Look very closely at the Coremark slide – they say which compiler version and what switches are being used. The Nvidia SoC’s use GCC 4.4.1 with -O3 and Intel uses GCC 3.4.4 with -O2 etc…

    Why in the world use such radically different environments?

    What are the results if the same compiler and switches are used?

    This really is an unfair comparison it would seem.

    -Bob

  • Snus

    I wonder what will be the average power consumption. No one doubts that nVidia can give performance, but they don’t say anything about battery life of such tablet.

    About unfair comparison: all marketing comparisons are unfair. Choose one particular benchmark, compile it as you wish, take competitors product released in 2006 (yes, wikipedia says T7200 is august 2006) and you can get any results that you want.

  • Michael Rayfield

    Wow, what a week. Appreciate all the great responses. Once I land in the US I will get back to you on your comments. Mjr

  • http://itextsoftware.com Mark Storer

    I’d like to echo SNUS. My Evo 4G burns through its battery too fast for my taste already. I cannot imagine that a x75 improvement in performance will come without some sort of increase in power usage.

    NVidia’s history of GPUs (which I’ve used for many generations) does nothing to help me believe that these chips will be anything but juice sucking monsters. Extraordinarily powerful monsters, but all that power won’t be of much help with a dead battery.

  • atlana

    First one ? hm, I question that, since a GERMAN I.T. news-site posted http://winfuture.de/news,61417.html on 14th FEB 2011 09:42am THEIR time (thats 8:42am UTC or 2:42am EastCoast Time) that QUALCOMM has revealed their APQ8064 quad-core 28-Nanometer SOC running at 2.5 Ghz (with 12x performance and 75% less power-usage compared to first Snapdragon SoC) with a four-core 320 Adreno GPU that runs supposedly 15x faster than the Adreno 200.

    So who’s lying here? Nvidia or Qualcomm ?

  • Hesam Aslan

    hi Michael,
    Thanks for this post on the Tegra Roadmap. About a month back Netflix commented on Tegra 2:
    http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&prid=704926&releasejsp=release_157&xhtml=true
    “Netflix is using the Tegra 2 development platform to bring the Netflix experience to Android super phones and tablets. We’re working closely with NVIDIA to ensure Netflix takes full advantage of Tegra’s outstanding acceleration and security capabilities.” — Greg Peters, Vice President, Product Development at Netflix
    I was wondering if you can update on that, if Tegra 2 is ever going to receive Netflix certification. Because Tegra 2 is in many coming tablets I think it’s very important for consumers to know if Netflix will ever be available on their Tegra 2 Tablets like Motorola Xoom.

    Thanks again for your response.

  • Michael Rayfield

    @Bob We did not mean to be unfair. I ran the Kal-El on the latest compiler I had and pulled the Intel data from the Coremark web site. To be fair I also tested my 1.0GHz part and not the 1.5GHz part. Thanks for point out my oversight, mjr

    @SNUS & Mark These are architectures on the roadmap. That said, I believe that at any given generation I must have an offering that will be able to do any given task that the previous generation does but at a lower power. Kal-El will accomplish that, mjr

    @Atlanta Big difference, I demonstrated a complete system. Chip, OS, applications and industrial design. QCOM demonstrated their ability to type a press release, just like TI did on OMAP5. Customers can’t build systems out of press releases (-: thanks mjr

  • Michael Rayfield

    @Hesam Yes we are still on track. this is a 3 way cooperation. I will supply the technology in the Tegra and secure OS, the OEM must integrate it and work a deal with Netflix. Looking forward to the complete solution. mjr

  • atlana

    @MICHAEL RAYFIELD
    Touché!

    ;-)
    I was under the impression that they actually had a WORKING SYSTEM to show off. It seems that they’re 2012 sending samples to ODMs/OEMs, and not now. Seems Qualcomm wants to milk the market a bit more/longer, claiming there’s no demand for such “powerful systems/SoC’s” *snort* yeah right…
    anyhoo – my apologies.

  • TechU

    MICHAEL RAYFIELD said “To be fair I also tested my 1.0GHz part and not the 1.5GHz part.” “Big difference, I demonstrated a complete system.”

    sure, Ok but can it run x264 well today :D

    simple i want to see a real life checkasm –bench (the only one that matters really today for real life cross platform app comparison AFAIC) on both this nvidia quad ARM 1.0GHz part AND your 1.5GHz part.

    simply git pull and compile x264 from https://github.com/DarkShikari/x264-devel for instance
    and do a simple
    make checkasm;./checkasm
    checkasm –bench

    we will make allowances for the fact a lot of ARM assembly needs writing yet OC and so there’s a lot of potential to massively increase these initial checkasm –bench results on the quad chips OC.

    will you do that simple thing for both the1.0GHz part AND your 1.5GHz part and put the results on the #x264dev IRC channel and a web page somewhere ASAP.

    as a side note you could also get you’re nvidia assembly dev’s to also write that missing ARM assembly and submit a patch to the same #x264dev IRC channel and work with the devs there to get ssh shell access to your kit and so improve your scores compared to other vendors later :D

  • Micky Badgero

    Hi Michael,

    Does the new chip have OpenCL or CUDA support? “secure OS” implies no, but can you clarify? Also, I understand that you tested a 1.0GHz (8W?), not the 1.5, but are we looking at about 12W max for the final version?

    Thanks,
    Micky

  • miabeats

    I am actually writing this on my xoom. I must say nividia you are doing a great job. This processor is flawless been running lots of tests on it. Cant wait until tegra 3 comes out.

  • Nate

    i knew it was coming something faster than fast looking forward to getting the tegra 3

  • Michael Rayfield

    @Micky, we will support OpenCL and Cuda in time, nothing announce yet. There is a “secure OS” offering. On the speeds and power not sure where those came from, 8W would not be a mobile SOC, we are well below 1W for nearly all use cases in mobile. Flat out it would burn more but never approaching 8 or 12 watts, thanks mjr

  • http://pdf.my-addr.com/ gottwrene

    Multumesc pentru informatii interesante