Design Space Explorer of QII 7.2 has serious bug!

Go to this link to download older Quartus versions, hope this helps.

https://www.altera.com/support/software/download/altera_design/quartus_archive/dnl-quartus_archive.jsp

--- Quote Start ---

Go to this link to download older Quartus versions, hope this helps.

https://www.altera.com/support/software/download/altera_design/quartus_archive/dnl-quartus_archive.jsp

--- Quote End ---

That's a great help! Thanks a lot!

--- Quote Start ---

After I finished exploring and found a better timing result, I saved it as a new revision, opened it in QuartusII 7.2 and compile it, but the compiled timing result is not as the DSE reported! This problem makes the new v7.2's DSE almost useless, Now I'd rather roll back to v7.1, although v7.1 is full of bugs and crashes frequently, but at least it's DSE can give me a reliable result as it reported... But the most miserable thing is I can't found v7.1 web version's download link any more....

--- Quote End ---

Hi, I just tried a simple design and I couldn't reproduce this problem. I ran a 5 seed sweep in DSE searching for performance using highest effort level. DSE reported the best results at a point that wasn't base and I turned those best results into a new revision in the project using DSE's create-new-revision dialog. When I opened that new revision up in the Quartus II 7.2 GUI and ran it I saw the exact same timing results for the circuit, same worst-case slack for the clock, same failing paths, same utilization numbers, etc.

Sounds like this might be a problem specific to your circuit or use scenario. If you want to discuss it further just drop me an email. I'd be happy to help you try and figure out what's going on here.

--- Quote Start ---

Hi, I just tried a simple design and I couldn't reproduce this problem. I ran a 5 seed sweep in DSE searching for performance using highest effort level. DSE reported the best results at a point that wasn't base and I turned those best results into a new revision in the project using DSE's create-new-revision dialog. When I opened that new revision up in the Quartus II 7.2 GUI and ran it I saw the exact same timing results for the circuit, same worst-case slack for the clock, same failing paths, same utilization numbers, etc.

Sounds like this might be a problem specific to your circuit or use scenario. If you want to discuss it further just drop me an email. I'd be happy to help you try and figure out what's going on here.

--- Quote End ---

Yes I've just done the same thing as you, the compiled result in QII 7.2 is not same as in DSE. But I could get perfect matched results in QII 7.1 and it's DSE on the same project, so that make me think it's a new problem the new version software brings... Now I'm doing more test, thanks for your help.

--- Quote Start ---

Yes I've just done the same thing as you, the compiled result in QII 7.2 is not same as in DSE. But I could get perfect matched results in QII 7.1 and it's DSE on the same project, so that make me think it's a new problem the new version software brings... Now I'm doing more test, thanks for your help.

--- Quote End ---

No problem. Can I ask what platform you're running Quartus II under? Windows? Linux? If you hit play twice in the Quartus II GUI do you get the same results? If you just do a seed sweep in DSE (set the effort level to Low) is the best seed result found by DSE reproducible in the Quartus II GUI ?

--- Quote Start ---

No problem. Can I ask what platform you're running Quartus II under? Windows? Linux? If you hit play twice in the Quartus II GUI do you get the same results? If you just do a seed sweep in DSE (see the effort level to Low) is the best seed result found by DSE reproducible in the Quartus II GUI ?

--- Quote End ---

Yes I'm running QII in Windows XP, QII surely get the same results if I compile my project again and again cause it won't really compile anything when I click the play the second time:p In fact I think the QII itself is much more stable compare to v7.1, seem the only problem is there are some misunderstanding when it's DSE talk to the compiler... I haven't try the low effort level, I'll try now.