Fujitsu
Not logged in » Login
X

Please login

Please log in with your Fujitsu Partner Account.

Login


» Forgot password

Register now

If you do not have a Fujitsu Partner Account, please register for a new account.

» Register now

Recent Discussion:

NickBown | 22.12.2018, 14:18
We have come across this issue as well, and don't seem to have found a way around it (the server is ...
NickBown | 20.12.2018, 18:40
Hi everyoneWe've got an RX2540 M1 which won't boot past the Fujitsu splash screen (which shows the i...
MarkM | 15.10.2018, 11:33
Hi there.I do not believe the Fujitsu policy on this subject has changed.So NO you can not order dri...
Nov 21 2014

NVIDIA Launches New Flagship Dual-GPU Tesla® K80

/data/www/ctec-live/application/public/media/images/blogimages/teslak80.jpg

NVIDIA has properly used this week's SC14 in New Orleans to launch the Tesla® K80, a so-called dual-GPU accelerator which the firm claims to be the "fastest [...] for data analytics and scientific computing."

The Tesla® K80 is the bigger sister of last year's Tesla® K40 and was designed to accommodate a comprehensive range of HPC server applications in scientific, technical and commercial usage scenarios. Examples include CAD/CAE, fluid dynamics, animation, video editing and encoding, energy exploration, and bioinformatics. Altogether, NIVIDIA's list of GPU-accelerated applications counts no less than 280 positions.

Based on the Kepler™ architecture, the Tesla® K80 consists of a dual-GPU board combining a total 24 GB of memory with 480 GB fast memory bandwidth. To achieve this performance level, the Tesla K80 uses two GK210 graphics chips holding 2,496 CUDA cores each, working at 8.74 teraflops single-precision and, if GPU Boost is on, up to 2.91 teraflops double-precision performance – „a performance ten times higher than today's fastest CPUs on leading science and engineering applications, such as AMBER, GROMACS, Quantum Espresso and LSMS", claims NVIDIA. Aside from sheer hardware power, the Tesla® K80 comes with features like Dynamic Parallelism (for spawning new processing threads if required) and GPU Boost (for dynamic scaling of the GPU clock) built in. In NVIDIA's Board Specification (PDF), this latter feature is called "Autoboost," describing the GPU's ability to "raise the core clock to a higher frequency" if there's enough "power headroom" left. Ex-factory, Autoboost is turned on by default, so the Tesla® K80 will start at 560 MHz "base speed" when used for the first time – and eventually scale up to 875 MHz. As may be imagined, this mechanism also drives power consumption: according to NVIDIA, the new GPU will draw 300 watts when running at full steam.

Users keen on trying the Tesla® K80 may sign up for a free test drive on remotely hosted clusters. Shipping already, NVIDIA's new flagship is now available both from NVIDIA resellers and a variety of server OEMs.

For more information about the Tesla K product family check out NVIDIA's library of product literature.

 
SHARE

Comments on this article

No comments yet.

Please Login to leave a comment.