Next Frontier for Volunteer Computing

Another insightful email that came in last week talks about this being the next frontier for volunteer computing (after porting to mobile OS’s).  I think the idea of mobile OS’s is awesome but our focus here is FPGAs. 🙂

This email suggests the OpenCL route is the way to go as Altera and Xilinx are Cheap MLB Jerseys both in the process of developing ARM based SoC-FPGAs.

Currently, a dual core ARM Cortex-A9 coupled to an FPGA fabric via a high speed on chip 되어버린 interconnect.Roblox Robux Hack 2017

In this scenario the BOINC manager would be running on the ARM target while the BOINC application could be rewritten with OpenCL hooks that will be generate soft IP for the FPGA.

Altera has compilers (c to h) that take in high level oakley outlet languages and synthesize HW components based on the needs of the SW. I have heard acceleration values as much cheap nfl jerseys as 100-150x on some applications.

So in Cheap Jerseys From China this case the BOINC manager would load an application such as SETI@Home and then load the pre-compiled soft cheap oakleys IP into the FPGA for acceleration.

These devices (SoC-FPGAs) will be in everything from set top boxes, routers, base stations, and automotive.

Click for more info on the Cortex A-9.

One Response to “Next Frontier for Volunteer Computing”

  1. MarkJ says:

    In the case of a ARM A9 with an FPGA I would think BOINC would be running on the PC and you’d use the A9 + FPGA to compute. BOINC needs some form of local storage plus Internet connection so running it on the A9 would impractical. Best to use all of the FPGA plus whatever CPU support for computing.

    It’s also possible to have an FPGA without an imbedded CPU so those devices would require BOINC running on a PC using the FPGA to compute.

Leave a Reply

Your email address will not be published. Required fields are marked *