Parallella: The $99 Cluster That Needs Your Help

parallella_feat

If its Kickstarter campaign pans out, Parallella from Adapteva promises to deliver a $99 open source development board which contains an incredible 16 independent processors; so that the next generation of parallel processing applications, and indeed, developers, can get off the ground.

But with only 23% of its funding goal met so far, there’s still a long way to go for this revolutionary device.

The Processor Speed Limit

It used to be that, every year, you could get a desktop processor which was clocked at a considerably higher speed than previous models. That is, until the last couple of years or so.

We’ve now hit the point that was predicted by Moore’s Law: we are at the limit of our current technology and manufacturing processes. Until the next major advancement in processor technology becomes practical on a large scale (such as quantum computing), we are more or less stuck with the current “Speed Limit” for single processors.

The only practical way to continue delivering more processing power for our increasingly complex software is to bundle more processors together in the same machine. This concept is known as “parallel processing”, where tasks are spread out over multiple identical processors to cut down the work each individual chip has to do on its own.

We’ve seen this happening on our desktops and laptops for some time, and now even mobile devices like smartphones are coming with dual and even quad core processors onboard. Unfortunately, even this is not enough. It will take more than 4 or even 6 cores per machine to do the kind of advanced computational tasks required by our insatiable technology needs. Computational density needs to go way up to take things such as self-driving cars from prototype to mainstream.

But an even bigger concern is education: developers need to be able to learn and practice the techniques required to create software which adequately scales to dozens or even hundreds of individual processors. To create efficient parallel processing software, you need to develop and test it in a suitable environment, which simply isn’t available to the average developer today.

Adapteva hopes that the answer to both of these problems, computational density and parallel processing education, can be solved by the Parallella.

Parallella

  • Dual-core ARM A9 CPU
  • Epiphany Multicore Accelerator (16 or 64 cores)
  • 1GB RAM
  • MicroSD Card
  • USB 2.0 (two)
  • Two general purpose expansion connectors
  • Ethernet 10/100/1000
  • HDMI connection
  • Ships with Ubuntu OS
  • Ships with free open source Epiphany development tools
  • Dimensions are 3.4” x 2.1”

With the limited volume Adapteva currently manufactures the Parallella in, the cost is simply too high for the average consumer or student. But by manufacturing the components of the Parallella in much higher volume, Adapteva says it can bring the cost of the 16 processor Parallella board down to a mass market friendly $99.

At this price, anyone who wants to develop parallel processing software or do computational intensive tasks will be empowered to do so. Bringing this much processing power to the masses could spark off the next generation of development in fields such as artificial intelligence, graphics rendering, machine vision, and any number of tasks which require processing power which can’t be delivered by traditional consumer computer hardware.

Epiphany Multicore Accelerator

The main Parallella board is nothing more than a dual core ARM machine, not unlike what’s in modern smartphones and tablets. Alone, it doesn’t have any special capabilities, it’s just a way to interface with the real key to Adapteva’s plans: the aptly named Epiphany processor.

With the Epiphany, Adapteva stripped out all of the overhead that comes with running a whole OS and supporting legacy software, and started from scratch on a whole new architecture optimized for parallel processing. This means that the Epiphany processor itself is unable to do things like run an operating system, but when given instructions via the Epiphany SDK, it can execute those instructions which many times the speed and efficiency of a standard processor.

The current model Parallella development boards are using either 16 or 64 Epiphany processors on a small “daughterboard” that plugs into the main computer. But in the future, Adapteva envisions daughterboards with hundreds or even thousands of Epiphany processors.

Open Source Commitment

For the Parallella to really bring about the change in parallel processing education that’s required to get us into the next generation of software development, Adapteva knows they need to get their documentation and software out to as many people as possible. So accordingly, they have decided to open up all the design documents and software for the Parallella to the public at large.

To make parallel computing ubiquitous, developers need access to a platform that is affordableopen, and easy to use. The goal of the Parallella project is to provide such a platform! The Parallella platform will be built on the following principles:

  • Open Access: Absolutely no NDAs or special access needed! All architecture and SDK documents will be published on the web as soon as the kickstarter project is funded.
  • Open Source: The Parallella platform will be based on free open source development tools and libraries. All board design files will be provided as open source once the Parallella boards are released.
  • Affordability: Hardware costs and SDK costs have always been a a huge barrier to entry for developers looking to develop high performance applications. Our goal is to bring the Parallella high performance computer cost below $100, making it an affordable platform for all.

Pledging Support

The concept is sound, the price is right, and opening up the hardware and software is definitely the right course to take for the good of the industry and the developers working in it.

If the idea of being able to afford a cluster which would have only a few years ago cost tens of thousands of dollars appeals to you, or if you just want to help push us into the next generation of computing, pledge your support for the Parallella.


Tom Nardi

Tom is a Network Engineer with focus on GNU/Linux and open source software. He is a frequent submitter to "2600", and maintains a personal site of his projects and areas of research at: www.digifail.com .

Related posts

Top