Imperial College London > Talks@ee.imperial > CAS Talks > Exploiting Algorithm-based Fault Tolerance for Reliability in Parallel Accelerators
Log inImperial users Other users No account?Information onFinding a talk Adding a talk Syndicating talks Who we are Everything else |
Exploiting Algorithm-based Fault Tolerance for Reliability in Parallel AcceleratorsAdd to your list(s) Download to your calendar using vCal
If you have a question about this talk, please contact Grigorios Mingas. This talk will serve as a summary of my recent work on leveraging low-overhead, algorithm-level error detection for the purposes of fault detection, location and avoidance/correction. The current benchmark application, a parameterisable matrix multiplication accelerator running on the Xilinx Zynq platform, will be presented, followed by an explanation of the adopted error detection scheme. Hardware modifications to support both error detection and runtime algorithm modification will be explained, with exploration into the overheads and potential gains of several candidate correction strategies. This talk is part of the CAS Talks series. This talk is included in these lists:
Note that ex-directory lists are not shown. |
Other listsWind Farms Circuits and Systems Group: Internal Seminars Featured talksOther talksDegree-Based Network Models Market-Oriented Cloud Computing and the Aneka Platform FCCM '13 Debrief Mismatched Decoding Next-Generation Smart Grid: Completely Autonomous Power Systems (CAPS) FPGA fast prototyping with LabVIEW and Vivado HLS |