Parallella: A Supercomputer For Everyone
-
http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone[^] "The Parallella project will make parallel computing accessible to everyone." When your Arduino or RaspberryPi just won't cut it....it is time to go parallel...big style! Maybe this is the Epiphany you have been waiting for :)
Dave Find Me On: Web|Facebook|Twitter|LinkedIn
Folding Stats: Team CodeProject
-
http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone[^] "The Parallella project will make parallel computing accessible to everyone." When your Arduino or RaspberryPi just won't cut it....it is time to go parallel...big style! Maybe this is the Epiphany you have been waiting for :)
Dave Find Me On: Web|Facebook|Twitter|LinkedIn
Folding Stats: Team CodeProject
With only 1GB memory, why would anyone need that much processing power? It is a good idea but it needs to be thought through with an eye toward potential applications. The POC's claim that "big industry" does not want to buy into parallel processing is clearly false. What the POC means is that he is unable to convince established companies to support the project. What are the reasons for that?
-
http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone[^] "The Parallella project will make parallel computing accessible to everyone." When your Arduino or RaspberryPi just won't cut it....it is time to go parallel...big style! Maybe this is the Epiphany you have been waiting for :)
Dave Find Me On: Web|Facebook|Twitter|LinkedIn
Folding Stats: Team CodeProject
Now why does this remind me of the Transputer? ;P
-
With only 1GB memory, why would anyone need that much processing power? It is a good idea but it needs to be thought through with an eye toward potential applications. The POC's claim that "big industry" does not want to buy into parallel processing is clearly false. What the POC means is that he is unable to convince established companies to support the project. What are the reasons for that?
If you had scrolled further, you would have seen the FAQ, with a ram question. "The current board configuration only supports up to 1GB of SDRAM. This is a limit our current host ARM CPU. If the Parallella project gets funded, there will be more boards coming that have significantly more RAM." Also, each of the Epiphany cores has its own memory cache, which is shared amongst all Epiphany cores. So the 1GB is for the ARM cores, and the Epiphany chip appears to not even use it. Also, Big Industry doesn't want to buy into it, because there is 0 demand for this product and no known purpose, and even though I would love to have one of these computers, I have no idea what I would do with it. Benchmarks... all day, erryday.
-
If you had scrolled further, you would have seen the FAQ, with a ram question. "The current board configuration only supports up to 1GB of SDRAM. This is a limit our current host ARM CPU. If the Parallella project gets funded, there will be more boards coming that have significantly more RAM." Also, each of the Epiphany cores has its own memory cache, which is shared amongst all Epiphany cores. So the 1GB is for the ARM cores, and the Epiphany chip appears to not even use it. Also, Big Industry doesn't want to buy into it, because there is 0 demand for this product and no known purpose, and even though I would love to have one of these computers, I have no idea what I would do with it. Benchmarks... all day, erryday.
Jak Decidus wrote:
Big Industry doesn't want to buy into it, because there is 0 demand for this product and no known purpose
Most likely they didn't know how to sell this to the "big guys", as they don't mention to who or how they tried to sell this, i can't tell, but taking this product to the graphics industry (Nvidia, AMD, etc.) or to the cell phone makers (an insanely powerful cellphone, anyone?), probably would have given them better results. I can think of several outlandish uses for an insanely powerful pocket computer: realtime (put what you want here) recognition, neural networks, an emulated brain, run the Matrix, etc. :)
CEO at: - Rafaga Systems - Para Facturas - Modern Components for the moment...
-
If you had scrolled further, you would have seen the FAQ, with a ram question. "The current board configuration only supports up to 1GB of SDRAM. This is a limit our current host ARM CPU. If the Parallella project gets funded, there will be more boards coming that have significantly more RAM." Also, each of the Epiphany cores has its own memory cache, which is shared amongst all Epiphany cores. So the 1GB is for the ARM cores, and the Epiphany chip appears to not even use it. Also, Big Industry doesn't want to buy into it, because there is 0 demand for this product and no known purpose, and even though I would love to have one of these computers, I have no idea what I would do with it. Benchmarks... all day, erryday.
Thanks for this reply Jack. When I clicked on that FAQ link all I got was another Home page with no information on the memory question. Being a computational scientist in an industrial HPC center, I can imagine all sorts of things I could do with something like this board, if it will run generic Ubuntu or some other Linux distribution, and all the associated tools, programming tools included.
-
With only 1GB memory, why would anyone need that much processing power? It is a good idea but it needs to be thought through with an eye toward potential applications. The POC's claim that "big industry" does not want to buy into parallel processing is clearly false. What the POC means is that he is unable to convince established companies to support the project. What are the reasons for that?
More than enough for Mandelbrot sets and a lot of other CPU intensive fractal generation.
-
More than enough for Mandelbrot sets and a lot of other CPU intensive fractal generation.
At $100, the computer could be used for a lot of mundane tasks too. The project shows, overall, that it is possible to build an inexpensive computer that can be used for things like micro-tasking. To date though, such a machine has not appeared in a form that runs a standard OS and is mass-purchasable.