Honored Contributor
Honored ContributorHello Everyone,
I have the same observation, I have 16 core linux based server, 64GB RAM, for the overall time required for compilation, CPU1 is used 100% and rest of the CPUs are used less than 1%. I don't understand why multi-threading is not happening. Anyway my design is very large, kind of an MPSoC. Best regards Waqar --- Quote Start --- Do you see a big speed improvement when using parallel compiling? On my quadcore, I see that the CPU is most of the time at 25% usage, even with parallel compilation enabled (on 11.1 SP2). What is most surprising is that even for tasks that should be very easily multithreaded, such as synthesis, Quartus doesn't seem to use more than one core. I only see the CPU go up to 100% occasionally during fitting, end eve there only for a few seconds each time. To me it doesn't seem that Altera's implementation of multithreading is that optimal. Maybe it has improved since 11.1 Fro a marketing point of view, the web edition is supposed to be used for smaller designs (it doesn't support the bigger Stratix FPGAs) so it could make sense not to have parallel compilation enabled. But I agree it's a rather low move ;) hussainw, what kind of design do you have that takes days to compile? Isn't there anything that you can optimize in there to reduce compile time? Which part takes most time? --- Quote End ---
