I don't mind posting here and saying how wrong I was. Because on the original post, I said for sure it wouldn't be funded.
To show that I have sufficiently eaten crow, I felt I should make this post and pledge $99 + $20 (for a case) to them. And so I have.
EDIT: Clever use of the ?foolhn tag. I was going to post the Kickstarter link to HN as penance for my doubting Thomas ways but I've never submitted anything, and when I tried, I got a dupe check, so I didn't bother.
Congratulations. I was a doubter in the first thread on Parallela, but I'm very impressed and happy that you guys achieved your goal.
I'd love to hear more about your plans to build a community around the boards now. I'm working on building a community around a yet unreleased but announced open source project myself (http://famo.us/) and I'm in the early stages of planning everything.
Check out The Art of Community by Jono Bacon, the guy who was responsible for the Ubuntu community. That book has a lot of good stuff.
Building the community will be the success or failure of this project. We've seen it with the Raspberry Pi and (to some extent) Ubuntu. They aren't the best products in their field (IMHO; I've not made anything better), but they have the best community. These days that means more than a mail list; it's the forum and Q&A (stack exchange) sites as well.
Dear Founders, please please set up a Q&A site. They are much more effective than forums.
Do you see anything wrong with using StackOverflow to manage community Q&A over self hosted?
Also, are there any good self-hosted versions of StackOverflow? Is there a way to use SO on your company site in a way that works as a portal to SO, but with all the Q&A visibility limited to tags/queries relevant to your product?
It's a shame they didn't hit the 3m goal. The device they were going to build them seemed more competative. I think the 16 core device will probably be an interesting toy though. Wrote my thought up here:
Oh good! I've posted about the project on HN a couple of times, but being $150k short of the goal yesterday with only 24 hours to go I wasn't optimistic.
My thoughts were exactly the same. In mere 8 hours yesterday they went up by more than $100k. Is there any sorts of stats on Kickstarter to see which sizes of pledges were contributed during a given period of time? Seems a little dodgy to me - could they actually fund the rest of the campaign out of their own pocket(they have a 2.5M investement after all)? But anyway I already pledged $120 myself to get the board, they guys seem trustworthy.
Update2: So they raised $192k yesterday, October 26 with 903 backers which averages to $212 per pledge. So it is all looks fine, perhaps that's how campaigns generally behave.
It seems it is the norm - lots of people were going back and upping their pledges. They did a ton of updates, and offered cases (for an extra $20 per unit) and a book (at $25) to pull in existing people. And a lot of those of us who had already backed were also far more actively tweeting or otherwise promoting it. Add on coverage here and several other places, and the time pressure. Others pointed out yesterday that it's fairly common for Kickstarter campaigns to get most of their funding in the first 72 and last 24 hours.
Do we know the specifics of the final funding? In the discussion yesterday there were good comments flying around about founders maxing out credit cards, general trends of most funding being in the first 72 and last 24 hours of a Kickstarter campaign, and various other interesting arguments (not all optimistic). I'd be curious what it came down to in the end.
Personally I see a killer application of (clusters of) parallella boards in machine perception for e.g. robotics and information retrieval/websearch. When I talk about Machine perception it could, for instance, refer to advanced visual object detection, recognition and tracking combined with pattern recognition on other sensory modalities such audio and haptics.
To make high computing power portable, as could be needed for machine perception applications in robotics, high power efficiency is needed. This is something the Parallella is aiming for.
Machine perception algorithms using "brain-like" computation structures are not only highly parallel by nature but also need fast access to memory. Instead of having a thin pipe to a central memory, it needs to be distributed, build close to/in to the computation cores.
The long-term vision of Parallella’s Epiphany multi-core processors indicate a 1000 cores per processor, with 128KB memory per core by 2014 a possibility(!) This is exactly the order of parallelism and distributed memory Machine perception applications need.
As visionscaper says, a lot of applications in vision and robotics, but I believe the incredible low power consumption offers broad potential in portable electronics, and also the datacenter. Imagine running an Erlang app on a single box crammed with thousands of these cores and a few TB of RAM.
They understand that the key is getting developers writing software in a new way, and the trick there is ubiquitous hardware. They also understand that it has to start as a co-processor, too, so they've married it to an existing ARM chip to provide peripherals and the host operating system. These decisions are very, very good ones, especially as compared to other multicore processor manufacturers (XMOS and GreenArrays are the two I'm thinking of).
Signal processing immediately comes to mind, and that is the use I am planning to put the chip towards. An associate of mine has developed a new method for signal decomposition on non-linear signals which FFT is not suitable for, and which is able to split the processing into parallel streams of processing.
Because the chip consume so little power, you will be able to build high performance processing into a small portable box that can run on batteries or USB (depending on how power efficient the final chip is, but they are currently talking about 2 watts).
I could see that. I pledged at the "cluster" level ($1000) when the project looked likely to fail, then when it crossed the finish line I backed off to the "personal cloud" level ($500).
To show that I have sufficiently eaten crow, I felt I should make this post and pledge $99 + $20 (for a case) to them. And so I have.
EDIT: Clever use of the ?foolhn tag. I was going to post the Kickstarter link to HN as penance for my doubting Thomas ways but I've never submitted anything, and when I tried, I got a dupe check, so I didn't bother.