NVIDIA Launches Jetson TK1 Mobile CUDA Development Platform

Subject: General Tech, Mobile | March 25, 2014 - 09:34 PM |
Tagged: GTC 2014, tegra k1, nvidia, CUDA, kepler, jetson tk1, development

NVIDIA recently unified its desktop and mobile GPU lineups by moving to a Kepler-based GPU in its latest Tegra K1 mobile SoC. The move to the Kepler architecture has simplified development and enabled the CUDA programming model to run on mobile devices. One of the main points of the opening keynote earlier today was ‘CUDA everywhere,’ and NVIDIA has officially accomplished that goal by having CUDA compatible hardware from servers to desktops to tablets and embedded devices.

Speaking of embedded devices, NVIDIA showed off a new development board called the Jetson TK1. This tiny new board features a NVIDIA Tegra K1 SoC at its heart along with 2GB RAM and 16GB eMMC storage. The Jetson TK1 supports a plethora of IO options including an internal expansion port (GPIO compatible), SATA, one half-mini PCI-e slot, serial, USB 3.0, micro USB, Gigabit Ethernet, analog audio, and HDMI video outputs.

View Full Size

Of course the Tegra K1 part is a quad core (4+1) ARM CPU and a Kepler-based GPU with 192 CUDA cores. The SoC is rated at 326 GFLOPS which enables some interesting compute workloads including machine vision.

View Full Size

In fact, Audi has been utilizing the Jetson TK1 development board to power its self-driving prototype car (more on that soon). Other intended uses for the new development board include robotics, medical devices, security systems, and perhaps low power compute clusters (such as an improved Pedraforca system).It can also be used as a simple desktop platform for testing and developing mobile applications for other Tegra K1 powered devices, of course.

View Full Size

Beyond the hardware, the Jetson TK1 comes with the CUDA toolkit, OpenGL 4.4 driver, and NVIDIA VisionWorks SDK which includes programming libraries and sample code for getting machine vision applications running on the Tegra K1 SoC.

The Jetson TK1 is available for pre-order now at $192 and is slated to begin shipping in April. Interested developers can find more information on the NVIDIA developer website.

 

March 26, 2014 | 02:21 AM - Posted by trent (not verified)

People don't wanna give pc gaming a chance because they think it's too expensive, even though that's not the case it sure is hard to make that argument when nvidia is doing everything in their power to make the pc platform seem inaccessible. With all their talk about moving the pc platform forward with geforce experience, shadowplay ect... it seems they are doing the exact opposite. If I hear another moron go on about how this is for content creation and it's not marketed at gamers i'll be sick. If this wasn't marketed for gamers it wouldn't be under the geforce brand and they wouldn't specifically state that it's built for 5k multi-monitor gaming. What is the point of this other than to line their pockets with cash and the tears of the titan black owners who just shelled out 1k for the highest end gpu that lasted a couple weeks. Same thing they did to the titan owners last year with the 780, the 780 owners when they released the 780 ti and the ti owners when they released the titan black. This card will not outperform two titan blacks or two 780 ti's and yet will cost 3k, WTF NVIDIA. If you're a content creator buy a titan or two and save a grand. The console fanboys have something to wave in our faces thanks to nvidia and I bet there's a person on the fence about pc gaming that will see the titan z headlines and decide to stick with consoles.

March 26, 2014 | 02:05 PM - Posted by Gregster

This is for content creation and isn't aimed at gamers.

March 26, 2014 | 09:30 AM - Posted by Anonymous (not verified)

This presentation is mostly for stock holders, and cable companies, etc. but that development platform (Jetson) better have some OpenGL libraries and be friendly towards full linux distros. If Gabe wants to gift the Debian developers something usefull, along with other development platforms for mobile he will get them some Jetsons and other companies mobile development platforms. That $3000.00 card better have certified professional graphics drivers, as well as the gaming optimized drivers for that price. It's sure priced like a Quadro. Still no new news about the Denver K1 cores SKU, and I wish someone with deep pockets could get one of the up and coming Mobile GPU companies to Develop a descrete GPU, to give Nvidia and AMD some competition. The PowerVR GR6500 could be scaled up for the Laptop/Chromebook market and with hardware raytracing it would be great for graphics, as well as gaming. There will be more innovation going in in the Mobile market, and a lot of that innovation will filter up to the PC/laptop market, but there really needs to be a third player in the desktop GPU market. Unless AMD finally open sources Mantle, and Nvidia gives other graphics APIs more/the same attention it gives CUDA, it's more about market share, and less about bringing affordable graphics to the market, and a third player in the desktop GPU market would help. Intel just provides graphics to sell its CPUs, and is not really focusing on really competing for the descrete GPU market.

April 1, 2014 | 03:45 AM - Posted by Steve Burke (not verified)

This is not entirely related to the original post, but I'm curious to see whether nVidia's mobile platforms sneak their way up the chain into larger devices at some point. Primarily TEGRA. Since NVLink can only work on IBM Power CPUs in servers right now, and because AMD obviously has no interest in supporting NVLink and Intel likely doesn't want to work with nVidia that much either (ARM alliances; its own IGP), I'm curious to see if a non-x86 CPU in the form of TEGRA could pose an alternative consumer platform.

It'd be far, far off and the architecture isn't built for that right now, but... the market is starting to get a bit fierce again, which could yield interesting changes to competition.

April 1, 2014 | 05:53 PM - Posted by Tim Verry

Well, if we see an uptick of low power, high density servers like Pedaforca (Tegra+discrete graphics), A Tegra chip with NVLink would make sense. NVIDIA could create their parallel of AMD's SeaMicro stuff with a Tegra-powered server cluster that pairs NVIDIA NVLink-enabled GPUs with low power ARM cores for workloads that are highly parallelized and not terribly CPU bound.

 

NVLink is interesting, and I want to see where it goes but I think you are right that Tegra may be the only natural path with AMD and Intel wanting to go their own directions.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.