- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
This is my acl_quartus_report.txt
///////////////////////////////////////////////////////// ALUTs: 7794 Registers: 9,641 Logic utilization: 5,368 / 32,070 ( 17 % ) ( 16 % ) I/O pins: 103 / 457 ( 23 % ) DSP blocks: 0 / 87 ( 0 % ) Memory bits: 348,224 / 4,065,280 ( 9 % ) M10K blocks: 63 / 397 ( 16 % ) Actual clock freq: 135.639999807 Kernel fmax: 135.64 1x clock fmax: 135.64 2x clock fmax: 10000 Highest non-global fanout: 2723 ////////////////////////////////////////////////////////// 1) I can see that Kernel fmax is 135.64. Does this mean that fmax is 135.64MHz??? 2) my kernel make 350036 elements and work. if I caculate kernel ideal execution time except memory load/write delay. That is 1/135.64MHz * 350056 = 2.58ms??? Thanks,Link Copied
3 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
The fmax is the highest frequency the design could be run at. The actual used frequency depends on your hardware, on the oscillator you are using and on your pll settings, if you use plls.
I don't understand "my kernel make 350036 elements and work.". If your design uses 350036 clock cycles to compute a result, then yes the execution time will be 1/f*350036.- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
2)
If you mean that you run the kernel on 350056 work-items, then to know the ideal execution time you would need to know how many work-items come out of the pipeline per clock. It might not necessarily be one depending on the kernel structure and instructions, in fact it will most probably be lower. I don't have any report on hand right now, but I believe with the profiler you are able to get that information. Also, you would need to know the generated pipeline depth and calculate the time it takes for the first work-unit to get processed. All in all, maybe something like this would do: 1/135.64MHz * 350056 / clockcycles_per_work-unit + pipeline_processing_time- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I am grateful for your replys.
Yes, 350056 means 350036 work-items. And I don't know my pipeline depth of my kernel. Can you teach me how do I know pipeline depth??? thanks.
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page