Home    Bloggers    Messages    Webinars    Resources   
Tw  |  Fb  |  In  |  Rss
Warren Miller

The Future of FPGAs: Where Do We Go Now?

Warren Miller
Page 1 / 2 Next >
William Murray
William Murray
10/27/2012 4:14:07 PM
User Rank
Blogger
Re: Radiation-Aware Synthesis
You get that free with Actel/Microsemi  -- Here in the US

50%
50%
JezmoSSL
JezmoSSL
10/27/2012 2:41:21 PM
User Rank
Blogger
Re: Radiation-Aware Synthesis
I did some work on fault tolerant design when working on a control system for radiotherapy in cancer treatments, and the issues which arose where the fact that you can buy a synopsis tool which will do majority voting and fault tolerent coding but it costs £50,000 per seat per year.So you end up having to do the coding by hand which is tedious, or you can use the configuration RAM CRC checking hardware to check for SEUs or a combination of these techniques or you can go for anti-fuse

50%
50%
William Murray
William Murray
10/26/2012 9:25:24 AM
User Rank
Blogger
Re: Radiation-Aware Synthesis
I agree -- There are many other areas where Rad Tollerant comes into play that are not obvious -- Medical during Cancer Treatments --   Downhole Oilfield when passing through Radio-Active Rock layers, or in porous rock with high levels of Radon Gas, -- Aviation at high altitude is another area of challenge.  Mining equipment, and the Nuclear Power, and up and Comming Fusion Power and High Energy Physics all need to consider the effects of Radiation in different ways  -- There also is Non-Destructive test equipment that uses radio-activity as well as Scientific Instrumentation and Mining and Construction Equipement that all can be exposed to Radio-Activity.

50%
50%
Bill White
Bill White
10/25/2012 8:37:16 PM
User Rank
Expert
Re: Radiation-Aware Synthesis
I agree, radiation tolerance specs for FPGAs would be good to know given their increasing use in spacecraft and terrestrial design where radiation becomes a concern.

50%
50%
Paul A. Clayton
Paul A. Clayton
10/25/2012 4:00:19 PM
User Rank
Beginner
Virtualizing FPGA resources
One of the issues constraining the use of FPGA resources as a general MCU extension would seem to be the problem of virtualizing and allocating (including scheduling) such a limited resource.  For a system in which all users of the FPGA resources are known ahead of time and can cooperate in resource use, this would not be a problem; but when any task might want to use any amount of the FPGA allocation becomes more complex.

(One relatively simple way of handling this issue would be to have an alternative software path.  If a task could not run with the extension because the FPGA resources are otherwise occupied, the functionality could be performed by software.)

There may be some non-embedded applications where virtualization of the FPGA resources would not be necessary.  E.g., many servers might have a very predictable utilization of such resources.

It is possible that something like Tabula's Spacetime FPGA's might be more friendly to virtualization by increasing the effective capacity of the FPGA.

On the contrary side, mature computation that is widely used would tend to migrate to hardened logic, though the functionality might be somewhat configurable (rf. "Conservation Cores").  It is not clear (to me) how relatively unused specialized hardware would need to be before FPGA implementation would make more sense, especially considering the desirability of dark silicon.

50%
50%
Max Maxfield
Max Maxfield
10/25/2012 11:15:51 AM
User Rank
Blogger
FPGA Place-and-Route *in* FPGAs
In your blog you mentioned the idea of using the massively parallel processing capabilities of an FPGA to perform place-and-route for FPGA designs.

 

Did you see my recent blog here on All Programmable Planet Wow! A Zynq-Based Supercomputer for Only $100?

 

Well, I posted a related blog over on Programmable Logic Designline (Click Here to see that blog) and someone there made the comment that this Zynq-based supercomputer might be idea for performing FPGA place-and-route

 

"Great minds think alike," as they say (of course, they also say "Fools seldom differ," but I'm sure that doesn't apply in our case :-)

50%
50%
Max Maxfield
Max Maxfield
10/25/2012 11:10:24 AM
User Rank
Blogger
Radiation-Aware Synthesis
You might want to talk to the folks at Xilinx and Mentor abouty their respective capabilities to implement radiation-tolerant FPGA designs

50%
50%
Max Maxfield
Max Maxfield
10/25/2012 11:09:20 AM
User Rank
Blogger
Graph-Based Synthesis
You could also talk to the folks at Synopsys about their Graph-Based Synthesis -- as I recall thsi eases the task of Place-and-Route and it also handles heterogeneous architectures really well

50%
50%
Max Maxfield
Max Maxfield
10/25/2012 11:08:02 AM
User Rank
Blogger
Total Recall
Hi Warren -- I'd love for you to talk to the folks at Synopsys about their Total Recall technology -- I'm not sure if it's on hold at the moment, but it's a brilliant idea.

50%
50%
More Blogs from Warren Miller
When traversing serial links with optics or backplanes, high-speed signals are degraded by impairments in the link, such as insertion loss, reflections, crosstalk, and optical dispersion.
Warren has finally started to write some HDL code to implement his chess-playing FPGA, but he's not a professional coder, so he needs our help and advice.
What might we see in new Ultra Low Density (ULD) CPLD families three-to-five years down the road? Are there new technologies or programmable structures that will find their way into ULD devices?
Following our evaluations, the resources required by a chess-playing FPGA implementation would seem reasonable, even for a small or midsized device.
A number of challenges are faced by the users and manufacturers of ultra-low-density devices (ULDs).
flash poll
follow us on twitter
follow Xilinx on twitter
like us on facebook
like Xilinx on facebook
All Programmable Planet     About Us     Contact Us     Help     Register     Twitter     Facebook     RSS