{"id":3906,"date":"2018-09-25T14:48:09","date_gmt":"2018-09-25T14:48:09","guid":{"rendered":"https:\/\/support.plunify.com\/en\/?p=3906"},"modified":"2020-04-06T04:32:44","modified_gmt":"2020-04-06T04:32:44","slug":"speed-up-your-fpga-regressions-with-the-cloud","status":"publish","type":"post","link":"https:\/\/support.plunify.com\/en\/2018\/09\/25\/speed-up-your-fpga-regressions-with-the-cloud\/","title":{"rendered":"How to speed up your FPGA builds with the cloud using a single Tcl command"},"content":{"rendered":"

Update: 03 Apr 2020
\nMany of us are presently working remotely and there has been a surge in compute requirements. Some of us brought the FPGA boards home but the compute servers are in office. This is a rehash of on how you can take advantage of the cloud to complete your builds faster.<\/p>\n

---------------------------------------------------------------------------<\/p>\n

Any serious hardware or software development effort will have a regression test suite. This article focuses on those tests that require constant (re-)compilation of FPGA designs and on how to speed up the turnaround times in order to keep tests, e.g. regression, running steadily.<\/p>\n

Imagine if you have 5 designs or variations of the same overall design to build for every regression.<\/p>\n

set design_list [list \"A\/A.xpr\" \"B\/B.xpr\" \"C\/C.xpr\" \"D\/D.xpr\" \"E\/E.xpr\"]\nforeach each_design $design_list {\n\u00a0   open_project \"$each_design\"\n    reset_run synth_1\n    launch_runs synth_1 impl_1\n    wait_on_run impl_1\n    close_project\n}<\/pre>\n
If each design takes 4 hours, then off-loading each build to a different machine can save a huge chunk of time.<\/p>\n
With these considerations, we developed additional Tcl capabilities into FPGA Expansion Pack, a\u00a0plugin to Vivado<\/a> that enables the use of high-performance build machines in the Amazon cloud.<\/p>\n
What is the FPGA Expansion Pack (FEP)<\/h2>\n
Comprising of a set of automation scripts and cloud utilities, FEP presents simple graphical and Tcl command-line interfaces to users to compile FPGA designs in the cloud. Subsequently download results back to their local environment as if the builds were done on-premise. The FEP can be\u00a0downloaded<\/a>\u00a0here and you will need a Plunify Cloud account to get started.<\/p>\n
\"\"
Cloud-enabled buttons in the Vivado toolbar<\/p><\/div>\n
How to use the cloud-enabled Tcl API<\/h2>\n
To use the previous example, this is how it would look like if you use FEP's Tcl API to compile in the cloud:<\/p>\n
foreach each_design $design_list {\n    fep::runCloudCompile -project \"$each_design\" -serverclass RA1\n}<\/pre>\n
Each design will be off-loaded to a separate cloud server so the builds will happen concurrently.<\/p>\n
You can also change the active run, e.g. impl_5, in a loop for each design so that you can fire off different design runs at the same time.<\/p>\n
# To choose a design run, make it active\ncurrent_run impl_5\n<\/pre>\n
The FEP comes with a Tcl API that aims to introduce as few changes to your existing scripts as possible, for example:<\/p>\n
To compile your project:<\/p>\n
fep::runCloudCompile -project \"\/home\/user2002\/bft_2017_2\/bft_2017_2.xpr\" -serverclass \"RA1\"<\/pre>\n
where
\n-project\u00a0<\/em>points to your project
\n-serverclass<\/em> specifies the type of cloud build machine to use.<\/p>\n
(Here is a list<\/a> of server class type. A \"RA2\" would have around 30Gb of RAM and costs only 76 cents per hour, including tools)<\/p>\n
This command returns a \"Job ID\", which is a number representing the build running in the cloud.\u00a0 To download the results, simply use<\/p>\n
fep::runCloudDownload -jobid <jobid, e.g. 12345><\/pre>\n
If the job is completed, the results will be downloaded and sync with your Vivado project as a new design run.<\/p>\n
Supporting Non-project Modes<\/h2>\n
For non-project mode, the current approach is to convert the non-project mode into a project mode using the \"save_project_as\".\u00a0 While this may not be as intuitive to non-project users, this ensures all necessary files are captured and uploaded to the cloud.<\/p>\n
How to get started<\/h2>\n
1. Go to https:\/\/support.plunify.com\/en\/documentation\/install-and-run-fpga-expansion-pack\/<\/a> for the instructions on where to download and install the plugin.<\/p>\n
2. Download sample scripts from our github<\/a>.<\/p>\n
3. Refer to this page<\/a> to get started.<\/p>\n
<\/h2>\n            
\n\t\t\t\t                    
If you like this article, please subscribe to our blog. Enter your email address and have the latest insights on FPGA, cloud and Machine Learning delivered straight to your inbox.<\/p>\n<\/div>                    
\n                        
\n                        \n                        \n                        \n                        \n\t\t\t\t\t\t