Optimizing Soft Error Pulse Width for CPU Fault Injection

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
2 replies · 2K views
hoheiho
Messages
43
Reaction score
0

Homework Statement


Hi, I am trying to inject some faults into a CPU manually. But I am not sure what should the soft error (SET) pulse width should be? I know that is different in different case. Therefore, I am going to set up a random generator to generator the SET pulse width. But what range should I set?
I have google the pulse width but it desnot give me any useful information for the range. Most of them are the SET pulse width for inverter.

Best regard
Ivan
 
Physics news on Phys.org
hoheiho said:

Homework Statement


Hi, I am trying to inject some faults into a CPU manually. But I am not sure what should the soft error (SET) pulse width should be? I know that is different in different case. Therefore, I am going to set up a random generator to generator the SET pulse width. But what range should I set?
I have google the pulse width but it desnot give me any useful information for the range. Most of them are the SET pulse width for inverter.

Best regard
Ivan

What computer are you using? Why are you doing this? What if your "soft error" crashes the PC for good?
 
berkeman said:
What computer are you using? Why are you doing this? What if your "soft error" crashes the PC for good?

Thanks for your reply
I am using a simply processor and run it in modelsim using verilog. I want to find out the soft error rate and do the recovery. Before the recovery stage, I need to see which block in the processor cause the most errors rate and then consider how much recovery blocks should I put in. I know ALU and control block normally cause the most fault in CPU but I would like to make the "balance" between each block for recovery.