[Beowulf] Nvidia and IBM create GPU interconnect for faster supercomputing

Jörg Saßmannshausen j.sassmannshausen at ucl.ac.uk
Thu Mar 27 09:09:47 PDT 2014

Hi all,

Prentice is right, the picture and the text are not matching. I stumbled over 
that as well but did not have the time to think it through. 
My understanding is also that the NVlink is between GPUs and CPU.

However, the main question is: will that work with other chipsets (Intel, AMD) 
as well or will that only work with the IBM chips?

All the best from a cold London


On Thursday 27 Mar 2014 14:43:57 Prentice Bisbal wrote:
> On 03/25/2014 07:25 PM, Mark Hahn wrote:
> >> Sounds like a memory coherent 80GB/sec link:
> > nah.
> > http://nvidianews.nvidia.com/Releases/NVIDIA-Launches-World-s-First-High-
> > Speed-GPU-Interconnect-Helping-Pave-the-Way-to-Exascale-Computin-ad6.aspx
> > 
> > 
> > shows only gpu-gpu links (just like SLI).
> I think that picture is incorrect. It shows NVlink only between the
> GPUS, but the text accompanying it states
> "plans to integrate a high-speed interconnect, called NVIDIA® NVLink™,
> into its future GPUs, enabling GPUs and CPUs to share data five to 12
> times faster than they can today."
> This is in disagreement with the illustration. I think someone included
> the wrong illustration by accident, or drew it wrong. Is there anyone
> from NVidia on this list that can provide clarity here?
> --
> Prentice
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing
> To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf

Jörg Saßmannshausen
University College London
Department of Chemistry
Gordon Street

email: j.sassmannshausen at ucl.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html

More information about the Beowulf mailing list