search

amy-huang / pim_power

Brown University summer research fellowship 2018: using McPAT and CACTI to measure power consumption of concurrent applications on PIM and non-PIM systems simulated with SMCSim
2 stars 0 forks source link

How to run 3D memory correctly in CACTI #1

Open yinlee opened 6 years ago

yinlee commented 6 years ago

First I want to thank you for sharing your code and showing how to fix the bug in the configure file. I follow your README, moving the line about tsv_length and adding the mem_data_with&num_clk. My CACTI start the calculation, but it shows that there still a segmentation fault. I see your cacti-out.txt and I find that my result shows the same thing, except that segmentation fault. Could you show me some guides?

amy-huang commented 6 years ago

Mine actually also still has a segmentation fault, except it happens after all the important information is already printed out - so after that much is printed in cacti-out.txt, the segfault happens and prints to terminal. I'm not sure why that happens, and haven't looked into it since I don't need the recommendations that would be given if there weren't a segfault. Also, my bad for not replying sooner - I didn't realize anyone looked at my repo! Email me at amy_huang1 at brown dot edu if you'd like to talk more - I check that the most.