Wednesday, December 11, 2013

Parallela low cost desktop supercomputer coming, w/ Python support

By Vasudev Ram

Parallella | Supercomputing for Everyone

In January 2013 I had blogged about The Parallella project - a low cost Linux supercomputer, and mentioned that Python and Erlang were two languages that would be supported by the project. According to the article linked in that post, Parallela units were supposed to start being delivered around May.

Today I saw in an article on an O'Reilly Media site, that the Parallela team seems to have hit some issues and got delayed, but are now supposed to be releasing units of Parallela this mid-December. Here is that article, by Federico Lucifredi - it also has a lot of technical details about the Parallela supercomputer, both the hardware and the software:

Supercomputing on the cheap with Parallella

Some excerpts from the article that I found interesting:

[ - supercomputing power inside a small credit card-sized board running Ubuntu ...

- powered by just a few Watts of juice, delivering 90 GFLOPS of number crunching ...

- a host side powered by a 667MHz Zynq 7020 ARM A9 System-on-Chip manufactured by Xilinx ...

- number-crunching side is powered by a 600MHz, 16-core Adapteva Epiphany-III numerical accelerator

- Out of the box, this build of the board came with the Linaro 12.10 Ubuntu derivative, so it is extremely standard as Linux ARM distributions go ...

- apt-get has access to the full Linaro 12.10 repositories, so the choice of software is pretty much infinite ...

- the key take-away is that this board promises to blow open the doors to low-power, on-demand supercomputing, putting on the table a real, low-cost alternative to current number crunching powerhouses costing thousands of dollars ...

- While much less brawny a number cruncher, the Raspberry PI remains better suited to multimedia applications (than the Parallela) where its on-board GPU can offload rendering tasks ]

Parallela FAQs

The FAQ has a question, What programming languages will be supported?, to which the answer is:

"Any programming language that is actively maintained and supports free and open source development tools is a candidate. C/C++, OpenCL, and MPI will be supported from day one.",

but since the article also says that any Ubuntu package is just an apt-get away, I guess Python can be installed on it, if not there by default, and Ubuntu (on x86 at least) does have Python pre-installed, IIRC.

And a final excerpt from the article linked to in my January blog post about Parallela:

[ The uses to which backers have said that they will put their Parallella computers to use include sound processing, video encoding, 3D scanning, computer visioning, neural networks, physical simulation and, importantly, learning parallel programming!

Software-defined radio is an application that frequently comes up and Parallella is particularly well suited to this since the programmable logic it provides is situated between the ARM host, Epiphany accelerator and GPIO, allowing for digital radio hardware to be more easily integrated. ]

It was interesting to see software-defined radio (SDR) as one of the uses that customers have in mind for Parallela. I had blogged about software-defined radio here a while ago:

Software radio gaining ground?

[ Update: Just noticed that the Parallela blog has this post:

Porting GNU Radio Blocks to The Epiphany ]

GNU Radio is a Software Defined Radio (SDR) framework.

[ Updated the post with the paragraphs below, about the Wikipedia article on SDR, and also for a few typos. Apologies to readers seeing it twice as a result, via feed readers. ]

The Wikipedia article on software-defined radio is interesting. It says that SDR is expected to become the dominant form of radio after some time (this is said by the SDR Forum, though, may be somewhat biased).

Could this become another example of Marc Andreessen's oft-quoted 'Software is eating the world' thesis?

- Vasudev Ram - Dancing Bison Enterprises

Contact Page

No comments: