The Parallella Board

Overview


The Parallella computer platform is an open source, energy efficient, high performance, credit card sized computer based on the Epiphany multicore chips from Adapteva. This affordable platform is designed for developing and implementing high performance, parallel processing. Applications developed to take advantage of the on-board Epiphany parallel processor can achieve unprecedented performance at the lowest power in the industry.

The Epiphany 16 or 64 core chips consists of a scalable array of simple RISC processors programmable in C/C++ or a parallel programming framework like OpenCL. The mesh of independent cores are connected together with a fast on chip network within a single shared memory architecture.

 

The 66-core version of the Parallella board delivers over 90 GFLOPS while consuming only 5 Watts under typical work loads. For certain applications, this would provide more raw performance than a high end server costing thousands of dollars and consuming 400 Watts.

To get an idea just how powerful this little board is, check out benchmark scores for the Epiphany-IV and Epiphany-III processors at coremark.org or read our blog post.

This slideshow requires JavaScript.

Highlights


  • Credit card size computer
  • Low cost, low power multicore platform
  • HDMI, Ethernet, USB and 48 GPIO ports
  • Dual core ARM and 16/64 core RISC processors on board
  • ANSI C/C++ and OpenCL programmable
  • Up to 90 GFLOPS processing capability

Benefits


  • Accessible solution with Open Source hardware and software
  • Simple and low-cost integration due toimplementation flexibility, and resource availability
  • Powerful platform with
    • Dual Core ARM Processor
    • Programmable FPGA
    • Epiphany 16 or 64 core parallel processor

Features


  • Zynq-Z7010 or Z7020 Dual-core ARM A9 CPU
  • 16 or 64-core Epiphany Multicore Accelerator
  • 1GB RAM
  • MicroSD Card
  • 2x USB 2.0
  • 4 general purpose expansion connectors
  • 10/100/1000 Ethernet
  • HDMI port
  • Linux Operating System
  • 54mm x 87mm  form factor

Product Selection Chart

SKU CPU Coprocessor GPIO Connector Status
A101010 Zynq 7010 Epiphany III No In Production
A101020 Zynq 7010 Epiphany III Yes In Production
A101030 Zynq 7020 Epiphany III No Not Currently Available
A101040 Zynq 7020 Epiphany III Yes In Production
A101050 Zynq 7010 Epiphany IV No Not Currently Available
A101060 Zynq 7010 Epiphany IV Yes Not Currently Available
A101070 Zynq 7020 Epiphany IV No Not Currently Available
A101080 Zynq 7020 Epiphany IV Yes Limited Quantities

Target Applications


Consumer:

  • Smart-phones and tablet app acceleration
  • High end audio
  • Computational photography
  • Speech Recognition
  • Face detection/recognition

Computing Infrastructure:

  • Super Computers
  • Big Data Analytics
  • Software Defined Networking
  • Data-center Appliances
  • High Frequency Trading

Mil/Aero:

  • Radar/Sonar
  • Extremely Large Sensor Imaging
  • Hyperspectral Imaging
  • Communication Jamming
  • Military Radios
  • Munitions/Guidance

Medical:

  • Ultrasound
  • CT

Communication:

  • Communication test-bed
  • Software defined radio
  • Adaptive Pre-distortion

Industrial/Instrumentation:

  • Machine Vision
  • Autonomous Robots/Navigation
  • Automotive Safety
  • High Speed Data Acquisition/Generation

Other:

  • Compression
  • Security Cameras
  • Video Transcoding

 

For detailed information


Source Files:

21 Comments

    • If you’ve got access to WinCE/WinRT sources, you could do a BSP for it and run that. I’d suggest Android, but you’d have to know Linux to get that to run on this. You could run something like FreeRTOS on it since that’s available for the Zync as well as a few other RTOSes- but if you’ve a problem with doing Linux development, you’re going to have “fun” trying to get those to work on it.

      Seriously, you’re going to spend less time and money trying to learn Linux than trying to get “other operative systems” to run on it.

    • It’s an ARM-v7 architecture core set on the Zync. There’s a JVM already available for ARM-v7 class systems.

    • I don’t think it will help you. It runs JVM, but in the CPU (ARM CORTEX A9), but not in the “GPU”. To take advantage from this, you’ll need to rewrite some parts of the game in c to run parallel, or if your game uses OpenGL (almost sure it uses), it could use the Epiphany “GPU” to run the OpenGL calculations, like a CPU. I`m pretty sure it’ll handle it easy, BUT there isn’t a OpenGL implementation for Parallella yet. (Yoy might like to see this: http://forums.parallella.org/viewtopic.php?f=34&t=708)

      So, in short and medium run, it won’t help you gaming, except you actually have the game and port it to take advantage to Parallella.

  1. Currently I have projects that have a space for prallella, but I need to know more about how parallella could help me in some of my scenario, what is your best communication channel?

    Thank you,
    Tommy

  2. I don’t think it will help you. It runs JVM, but in the CPU (ARM CORTEX A9), but not in the “GPU”. To take advantage from this, you’ll need to rewrite some parts of the game in c to run parallel, or if your game uses OpenGL (almost sure it uses), it could use the Epiphany “GPU” to run the OpenGL calculations, like a CPU. I`m pretty sure it’ll handle it easy, BUT there isn’t a OpenGL implementation for Parallella yet. (Yoy might like to see this: http://forums.parallella.org/viewtopic.php?f=34&t=708)

    So, in short and medium run, it won’t help you gaming, except you actually have the game and port it to take advantage to Parallella.

  3. Does anyone know how much resources are left on the ZYNQ FPGA for user custom designs ?
    Is there an utilization chart somewhere ?

  4. Pingback: Custom PC, Issue 137 | Gareth Halfacree

  5. Hi there, well done with finally getting the board to market. I have been following this board development from the sidelines for a while. It’s funny because in the early days some people thought it was a scam, especially when delivery schedules were being changed. But now you have a working board for sale – well done! My only questions are? have you abandoned the 64 core board? If not when will be able to order it? Thanks.

  6. Pingback: Knowm.org | How to Install a Distributed Apache Storm Cluster

  7. Pingback: Knowm.org | How To Synchronize Time Across a Linux Cluster

  8. Hello All,
    I am interested in purchasing a couple of this boards, specifically the A101040 and the A101080. Do you know where I can place an order for tese two specific boards?
    Thanks
    Bob

Leave a Reply Using Facebook or Twitter Account