Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 15 submissions in the queue.
posted by cmn32480 on Wednesday April 06 2016, @01:49AM   Printer-friendly
from the that's-a-lot-of-horsepower dept.

During a keynote at GTC 2016, Nvidia announced the Tesla P100, a 16nm FinFET Pascal graphics processing unit with 15.3 billion transistors intended for high performance and cloud computing customers. The GPU includes 16 GB of High Bandwidth Memory 2.0 with 720 GB/s of memory bandwidth and a unified memory architecture. It also uses the proprietary NVLink, an interconnect with 160 GB/s of bandwidth, rather than the slower PCI-Express.

Nvidia claims the Tesla P100 will reach 5.3 teraflops of FP64 (double precision) performance, along with 10.6 teraflops of FP32 and 21.2 teraflops of FP16. 3584 of a maximum possible 3840 stream processors are enabled on this version of the GP100 die.

At the keynote, Nvidia also announced a 170 teraflops (FP16) "deep learning supercomputer" or "datacenter in a box" called DGX-1. It contains eight Tesla P100s and will cost $129,000. The first units will be going to research institutions such as the Massachusetts General Hospital.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by jdavidb on Wednesday April 06 2016, @02:01AM

    by jdavidb (5690) on Wednesday April 06 2016, @02:01AM (#327893) Homepage Journal

    I never really bothered to learn much about GPUs; what does this mean? For a brief moment there I thought this meant the Pascal programming language!

    --
    ⓋⒶ☮✝🕊 Secession is the right of all sentient beings
    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 5, Informative) by takyon on Wednesday April 06 2016, @02:09AM

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Wednesday April 06 2016, @02:09AM (#327895) Journal

    Brand name for Nvidia's first generation of 16nm GPUs. Previous GPUs have been stuck at 28nm. 20/22nm will be skipped.

    This is also Nvidia's first use of High Bandwidth Memory. AMD used HBM1 in Fury models, Nvidia decided to wait until HBM2.

    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
  • (Score: 0) by Anonymous Coward on Wednesday April 06 2016, @02:48AM

    by Anonymous Coward on Wednesday April 06 2016, @02:48AM (#327906)

    Yeah, my first thought was the submitter was a few days late and a couple paragraphs short.

    Their new GPU probably doesn't have much to do with electric cars either.

    • (Score: 2) by maxwell demon on Wednesday April 06 2016, @08:14AM

      by maxwell demon (1608) on Wednesday April 06 2016, @08:14AM (#327964) Journal

      I bet it doesn't even use Tesla coils.

      --
      The Tao of math: The numbers you can count are not the real numbers.
      • (Score: 2) by bob_super on Wednesday April 06 2016, @04:36PM

        by bob_super (1357) on Wednesday April 06 2016, @04:36PM (#328122)

        You are always free to add them inside your tricked-out case, if there's space left between the watercooling and the LEDs.
        Send us some pics.

  • (Score: 3, Funny) by Anonymous Coward on Wednesday April 06 2016, @03:05AM

    by Anonymous Coward on Wednesday April 06 2016, @03:05AM (#327908)

    What is a Pascal GPU?

    Better than a BASIC GPU, but still not as fast as a GPU with C.

    • (Score: 1, Funny) by Anonymous Coward on Wednesday April 06 2016, @04:49AM

      by Anonymous Coward on Wednesday April 06 2016, @04:49AM (#327928)

      This GPU with C is called a CPU, right?

  • (Score: 3, Informative) by WillR on Wednesday April 06 2016, @12:58PM

    by WillR (2012) on Wednesday April 06 2016, @12:58PM (#328031)
    Nvidia uses famous scientists as codenames, and Pascal is the new GPU family.
    Fahrenheit < Celsius < Kelvin < Rankine < Curie < Tesla < Fermi < Maxwell < Kepler < Pascal
    Then, just to keep things confusing, they reused Tesla as a marketing name for datacenter products.