Share |
Login Form
Newsletter



Receive HTML?

Latest Members


Reconfigurable Computing

Community Information
Category:
No community categories created yet.
Name:
Reconfigurable Computing
Tag:
Reconfigurable-Computing
Description:
For anyone interested in FPGA-based computing, platforms, applications, tools, research, products, and related issues
Created:
Tuesday, 11 August 2009
Creator:
Communitie type:
Open
Anyone can join and can view this community.
Private
This community requires approval for new members to join. Anyone can view the community description. Only community members are allowed to see the community content.

Community Leader's Blog

Nothing new today . . .

Community Discussions

Started by Tom VanCourt. Last replied by Brian Bailey on Friday, 30 October 2009
Started by Tom VanCourt. Last replied by Tom VanCourt on Friday, 18 September 2009
Started by Tom VanCourt. Last replied by Tom VanCourt on Friday, 18 September 2009
Started by Tom VanCourt. Last replied by Tom VanCourt on Monday, 17 August 2009

Community Wall

Clive Maxfield, 2009-12-14 10:32:56
Clive Maxfield
Is the time-to-program the main hurdle/challenge in the field of RC?
 
Jaime Cummins, 2009-10-21 10:16:28
Jaime Cummins
So - being a blank wall will start with something that will be controversial. Until reconfigurable computing stops being about using FPGAs to attempt to do computing, there can not be an effective reconfigurable solution. It is similar to using a hammer as a screwdriver. It works (sort of), but it is not what you want to use, and if you try it once, you never do it again.

So what do you really need? As with all new types of computing you need 3 pieces. You have to start with a need. Why do something different then we have been doing? If this is compelling then you can move to the second, what is the new hardware/IC that is required and how do you make it? But the third is the hardest part, where are the software tools that let me use this technology with some ease, and that accelerate my ability to design, implement, test, and deploy the complete solution.

I would argue that 1, and 2 above have been done over and over again. But that the software tools, the design flow, and making this accesible has been the elusive piece. Why one would ask? Well as with most such questions it is because this has been perceived mostly as a hardware problem, and new companies new teams have all been built around hardware. The software tools, making real world OS environments (like Linux an easy one), software libraries and more have always been the after thought. Therefore the new technology has stalled due to lack of access (most engineers are too busy to really learn complex new things, and management aversion to being the first).

So how do we move ahead? I leave this to the next wall scribbler. Next time will talk about what I feel this new form of computing is really about, and why we need to abandon the term reconfigurable computing.
 






Latest Content
User rating
 
0.0 (0)