Update #53: The 64-core Parallella is alive!

Dear backers,

I am extremely happy to announce that the 64-core Parallella board is alive and well!! The 64-core and 16-core Parallella computers share the same PCB design and on paper “it should have been” a very simple bringup process. The picture you see above is a 64-core Parallella board with a socket soldered onto the board to allow for in-place testing of the 64-core Epiphany-IV chips. We powered up the board last Thursday and by Friday evening we had a program running through to completion on the Epiphany, blinking an LED for good measure. Getting the elink read path flushed out took a little longer, but by Wednesday this week we had the whole board up and running perfectly. Having debugged 4 different Epiphany chips and 6 different Epiphany based boards to date, there is definitely a familiar pattern emerging:-)

  • Chip/board arrives
  • Nothings works–>panic
  • First “doh” moment–>signs of life, relief
  • Everything seems to work–>euphoria
  • A weird behavior manifests itself–>depression
  • Subtle “doh” moment and fix–>calm

We will start building and shipping the 64-core boards next week!

Despite all the delays and challenges along the way, the 64-core Parallella is still setting the standard in terms of energy efficiency. In fact, it could be argued that it’s the most efficient computer in the world today! Amazing to think that none of this would have happened if YOU hadn’t taken a chance on us back in 2012!

At one point during the Kickstarter campaign, life was a nightmare…now it feels more like a dream.

On the 16-core production front, we have now shipped up to backer #2,999 and all the remaining boards have made it through assemly. We expect to get ALL of the Kickstarter boards to our shipping partner by the end of next week.

If you received your board already, PLEASE do power up your board and tell us about your experience. The Parallella project long term survival hinges on the strength of the community and we need a steady stream of feedback to make the platform better.

Sincerely,

Andreas

Posted in Parallella Kickstarter Blog.

17 Comments

  1. I’ve got four of the 16-core boards. They’re beautiful! (Though I only got them a couple of days ago, so haven’t really had a good opportunity to kick the tyres.)

    When do you expect the 64-core parallella to be generally available?

  2. I was an early backer for the 16-core board and received my board about 2 months ago. Being new to c and all things resembling embedded programming, I’m just now starting to do something useful with the board, and I expect that in another two months or so I’ll start wanting another.

    Do you have any idea when the 64-core board will be generally available? (Great job, by the way!!) Next quarter? Next year?

    Now I’ll really push my luck– how about the chips with core counts in the thousands?

    Keep up the good work. I’m really excited about these chips.

  3. Hello Andreas,
    I’m a visual effects designer specifically interested in Parallella for rendering purposes. It seems like something that could speed up render times substantially, a competitive edge in my industry. However, most of the applications that i use do not support linux. Is there plans to add support for other operating systems in the future?
    Thanks,
    Ben

    • Ben, you would not necessarily need to have the end-user application run on a linux platform. Instead you would network your spplication machine with a parallella machine that runs a render server.

      What applications do you use?

      Cheers

  4. Actually I have another question – What is the performance difference between a GPU card with many more cores (eg. 3000 CUDA) and your 64-core architecture? Will you get the price of a 64-core board below $100? Cheers

  5. Everyone is interested to purchase the product (64 core) …. Is this possible to upgrade my 16-desktop to 64-desktop as well for embedded and server

  6. Um… wouldn’t an operating system and subsequently all applications running under it have to be optimized to use all 64 cores in order for them to really take advantage of the parallel processing power that thing has? I don’t want it ending up like a sort of Redundant Array of Inexpensive Processors instead of the Bad Ass Parallel Processor it’s supposed to be. Do you?

Leave a Reply Using Facebook or Twitter Account