| *** tpb has joined #symbiflow | 00:00 | |
| *** Degi_ has joined #symbiflow | 00:11 | |
| *** Degi has quit IRC | 00:13 | |
| *** Degi_ is now known as Degi | 00:13 | |
| daniellimws | mithro: Where? | 00:14 |
|---|---|---|
| mithro | @daniellimws: Bunch of people submitted pull requests to help fix https://github.com/SymbiFlow/ideas/issues/47 -- always good to have a second pair of eyes look at that everything in the checklist is covered | 00:16 |
| tpb | Title: Updated the repositories to compile with the current copyright best practices · Issue #47 · SymbiFlow/ideas · GitHub (at github.com) | 00:16 |
| daniellimws | mithro: Oh sure | 00:18 |
| daniellimws | mithro: For the v2x repo you asked me to also add the "coding: utf-8" thing to the top of python files. Should we require this for the python files in other repos too? | 00:22 |
| mithro | Yeap | 00:22 |
| daniellimws | mithro: Do you wanna add it to https://github.com/SymbiFlow/ideas/issues/47 | 00:22 |
| tpb | Title: Updated the repositories to compile with the current copyright best practices · Issue #47 · SymbiFlow/ideas · GitHub (at github.com) | 00:23 |
| mithro | @daniellimws Done! | 00:24 |
| *** az0re has joined #symbiflow | 00:33 | |
| daniellimws | mithro: By third_party code does this mean all the git submodules? For example inside https://github.com/SymbiFlow/symbiflow-xc7z-automatic-tester there's devmemX, linux-xlnx, u-boot-xlnx, zynq_bootloader, should they be put into a third_party directory? | 00:36 |
| tpb | Title: GitHub - SymbiFlow/symbiflow-xc7z-automatic-tester: Tool for automatically testing FPGA designs using a Zynq Series 7 board. (at github.com) | 00:36 |
| mithro | daniellimws: Yeap! | 00:37 |
| sf-slack | <timo.callahan> Hi @borisnotes, I'm trying to find an IRC client that works well, until then please use *symbiflow* to ask me questions (I can see them since the posts are echoed to Slack). | 01:48 |
| *** futarisIRCcloud has joined #symbiflow | 01:48 | |
| *** citypw has joined #symbiflow | 01:52 | |
| *** circ-user-VOF5C has joined #symbiflow | 01:56 | |
| *** circ-user-VOF5C is now known as tcal | 01:58 | |
| *** Bertl is now known as Bertl_zZ | 02:55 | |
| -_whitenotifier-9- [symbiflow-arch-defs] rakeshm75 opened issue #1436: Branch : Quicklogic : Primitives for the post-layout simulations - https://git.io/Jfk0Q | 03:22 | |
| mithro | https://antmicro.com/blog/2020/04/systemverilog-linter-and-formatter-in-fusesoc/ | 04:19 |
| tpb | Title: Antmicro · SystemVerilog linting and formatting with FuseSoC - Verible integration (at antmicro.com) | 04:19 |
| tcal | whois tpb | 04:33 |
| *** az0re has quit IRC | 05:30 | |
| *** craigo has joined #symbiflow | 05:55 | |
| *** _whitelogger has quit IRC | 06:03 | |
| *** _whitelogger has joined #symbiflow | 06:05 | |
| *** OmniMancer1 has joined #symbiflow | 06:23 | |
| *** OmniMancer has quit IRC | 06:26 | |
| sf-slack | <pgielda> A bot | 07:06 |
| *** ZipCPU has quit IRC | 08:05 | |
| *** ZipCPU has joined #symbiflow | 08:06 | |
| *** borisnotes has joined #symbiflow | 08:14 | |
| *** nonlinear has quit IRC | 08:25 | |
| *** nonlinear has joined #symbiflow | 08:30 | |
| *** ssb has quit IRC | 08:36 | |
| ZirconiumX | tpb: version | 08:44 |
| tpb | ZirconiumX: The current (running) version of this Supybot is 0.83.4.1. The newest version available online is 0.83.4.1. | 08:45 |
| *** borisnotes has quit IRC | 08:45 | |
| ZirconiumX | tcal: ^ | 08:45 |
| *** ssb has joined #symbiflow | 09:18 | |
| *** Bertl_zZ is now known as Bertl | 09:33 | |
| *** futarisIRCcloud has quit IRC | 09:56 | |
| *** Bertl is now known as Bertl_oO | 11:43 | |
| *** gsmecher has joined #symbiflow | 14:14 | |
| * hackerfoo uploaded an image: Screen Shot 2020-04-22 at 7.09.46 AM.png (844KB) < http://sandbox.hackerfoo.com:8008/_matrix/media/v1/download/sandbox.hackerfoo.com/FLKvfGKtFMkIRytedQYsJvAO > | 14:18 | |
| hackerfoo | What would I do if I had 5 n1-highmem-96's? I'd do 400 VTR tests at the same time, man. | 14:19 |
| hackerfoo | Running titan_quick_qor on 144 different flags: https://github.com/HackerFoo/vtr-nix-test/blob/master/tests.nix#L69-L82 | 14:25 |
| tpb | Title: vtr-nix-test/tests.nix at master · HackerFoo/vtr-nix-test · GitHub (at github.com) | 14:25 |
| ZirconiumX | It might be interesting to put a Cyclone V model up | 14:25 |
| ZirconiumX | Based on what I know of it | 14:26 |
| *** citypw has quit IRC | 14:46 | |
| -_whitenotifier-9- [symbiflow-arch-defs] kgugala opened issue #1437: quicklogic: investigate negative cost warnings - https://git.io/JfkFJ | 15:08 | |
| *** gsmecher has quit IRC | 15:08 | |
| *** gsmecher has joined #symbiflow | 15:22 | |
| mithro | acomodi: The results from your pull request at https://github.com/SymbiFlow/fpga-tool-perf/pull/73#issuecomment-617846483 look great! | 15:27 |
| tpb | Title: tools: add symbiflow support and re-enable CI on all supported tools by acomodi · Pull Request #73 · SymbiFlow/fpga-tool-perf · GitHub (at github.com) | 15:28 |
| mithro | https://www.irccloud.com/pastebin/vSjcMm6K/ | 15:28 |
| tpb | Title: Snippet | IRCCloud (at www.irccloud.com) | 15:28 |
| sf-slack | <acomodi> @mithro Yep :) what is needed now is a nice collection of results, so we can analyze the statistics for each run | 15:30 |
| -_whitenotifier-9- [fpga-tool-perf] mithro opened issue #74: Figure out why kokoro is using wrong GitHub token on this repo - https://git.io/JfkbQ | 15:52 | |
| *** citypw has joined #symbiflow | 15:53 | |
| -_whitenotifier-9- [fpga-tool-perf] acomodi opened issue #75: Make symbiflow toolchain installation dynamic - https://git.io/JfkbF | 15:54 | |
| *** citypw has quit IRC | 16:00 | |
| *** tcal has quit IRC | 16:10 | |
| *** tcal has joined #symbiflow | 16:27 | |
| *** OmniMancer1 has quit IRC | 16:38 | |
| *** futarisIRCcloud has joined #symbiflow | 18:10 | |
| *** adjtm has quit IRC | 18:23 | |
| *** adjtm has joined #symbiflow | 18:25 | |
| shapr | hackerfoo: do you have a blog or other place to publish your fpga layout investigations? | 19:13 |
| hackerfoo | Yeah, I haven't updated hackerfoo.com in a while. | 19:15 |
| hackerfoo | I've been working on accelerating QoR measurements with Nix: https://github.com/HackerFoo/vtr-nix-test | 19:16 |
| tpb | Title: GitHub - HackerFoo/vtr-nix-test: Using Nix to run VTR tests (at github.com) | 19:16 |
| hackerfoo | So now I can test algorithms as fast as I can make them. | 19:17 |
| shapr | exciting! | 19:20 |
| shapr | Seems like everyone I know is switching to nix to reduce compile times. | 19:20 |
| hackerfoo | I think the biggest improvement is moving data around automatically and keeping it organized. When running hundreds of configurations and flag settings, it's hard to not mess something up and draw the wrong conclusions. | 19:22 |
| hackerfoo | Or to do something dumb like making a code change or switching branches but forgetting to recompile. | 19:23 |
| shapr | btw, do you also have a patreon | 19:24 |
| shapr | ? | 19:24 |
| ZirconiumX | hackerfoo: what's measurement noise like? | 19:25 |
| shapr | My company is switching to nix to reduce compile times. | 19:25 |
| hackerfoo | No, I don't. | 19:25 |
| shapr | is there some other way outsiders could support your efforts? | 19:26 |
| hackerfoo | ZirconiumX: I'd estimate ~2% from previous runs, but now I can manage load much better because I broke every run into it's own derivation, where before a derivation might run a few tens in parallel. | 19:27 |
| ZirconiumX | It's funny, because I seem to have a fairly different methodology for "improvements" | 19:28 |
| hackerfoo | I'm keeping limiting task to 80 on a 96 core machine to avoid saturating it, but I might bring that up to 90. | 19:28 |
| ZirconiumX | I'd rather take the "latency" hit of multiple samples than limit "throughput" | 19:29 |
| ZirconiumX | So I use SPRT. | 19:29 |
| hackerfoo | shapr (IRC): Google is funding the work on Symbiflow (I'm an employee), but there are others such as Dave Shah (https://www.patreon.com/fpga_dave) that are funded by the community. | 19:30 |
| tpb | Title: David Shah is creating open source FPGA tools | Patreon (at www.patreon.com) | 19:30 |
| hackerfoo | shapr (IRC): I can help you get started if you are interested in contributing work to the project. | 19:30 |
| hackerfoo | ZirconiumX: https://en.wikipedia.org/wiki/Sequential_probability_ratio_test ? | 19:31 |
| tpb | Title: Sequential probability ratio test - Wikipedia (at en.wikipedia.org) | 19:31 |
| ZirconiumX | Yes | 19:32 |
| ZirconiumX | I actually have source code if you're curious | 19:32 |
| hackerfoo | From my experience with sensors, it's much better to gather clean data than to clean it up afterwards. | 19:33 |
| hackerfoo | ZirconiumX: I am curious. | 19:33 |
| ZirconiumX | https://github.com/ZirconiumX/yosys-testrunner | 19:33 |
| tpb | Title: GitHub - ZirconiumX/yosys-testrunner: Program for robustly comparing two Yosys versions (at github.com) | 19:33 |
| hackerfoo | Maybe we could combine both approaches for even better data. | 19:33 |
| hackerfoo | ZirconiumX: Thanks. | 19:33 |
| ZirconiumX | The point of SPRT is to have bounded error while maintaining optimality | 19:33 |
| ZirconiumX | I can go to incredibly small error for an old versus new test because it's generally pretty decisive | 19:34 |
| ZirconiumX | And I change the nextpnr seed each run to explore the design space as much as possible | 19:34 |
| ZirconiumX | As much as anything, I'm not asking "how do these compare", I'm asking "is this stronger" | 19:35 |
| ZirconiumX | My current measurement for "stronger" is "higher Fmax", but I specifically picked a multinomial model to support multiple outcome conditions | 19:35 |
| hackerfoo | I have the advantage of unlimited cores, so there isn't a need to saturate them. Many tasks are limited by available RAM, so it might be best to say at 80 cores, for 8GB per task. | 19:36 |
| ZirconiumX | Okay, sure | 19:41 |
| hackerfoo | Right now my aim is to minimize runtime without sacrificing (too much) quality. | 19:41 |
| hackerfoo | The goal is to improve VPR's runtime by 10x. | 19:41 |
| sf-slack | <pgielda> @hackerfoo - this is what we already did on (original) Distant approach (not using RBE) -- doing everything on one machine to a certain point (configure, common parts) then just cloning the hard drive a lot of times and forking a lot of small machines in GCP and running everything in parallel. Internally we were running it on Ryzens which was even faster for single core things (GCP did not have Epycs at that point and | 19:44 |
| sf-slack | even now I cannot spawn more than ~24 cores of n2s in total on GCP) | 19:44 |
| sf-slack | <pgielda> (I mean the "What would I do if I had 5 n1-highmem-96" thing) | 19:44 |
| hackerfoo | I had to fix a bug in Nix to handle waiting on more than 1k fds, but now it should be able to scale to thousands of cores now. | 19:48 |
| *** zkms has quit IRC | 19:54 | |
| *** zkms has joined #symbiflow | 19:56 | |
| sf-slack | <pgielda> Cool | 20:07 |
| sf-slack | <pgielda> I have to play with Nix once I have time | 20:07 |
| sf-slack | <pgielda> Oh, the main reason why it was better to spawn smaller machines was that a lot of things was single core anyway and single core performance drops once you are using more cores | 20:08 |
| sf-slack | <pgielda> In fact its a pity nobody has something like GCP-style on demand machines that have less cores, but the cores are super fast, specially cooled, can be unreasonably expensive even -- there are still workloads where you cannot really benefit from a lot of cores but rather from a smaller amount of really fast cores. | 20:10 |
| sf-slack | <pgielda> Ryzens/Threadrippers seem to be closest to that ;) | 20:14 |
| daveshah | For place and route, memory bandwidth/latency is another big constraint | 20:15 |
| sf-slack | <pgielda> (while still having a lot of cores ; )) | 20:15 |
| sf-slack | <pgielda> yes, true. | 20:15 |
| daveshah | which means more small machines almost certainly win there too | 20:15 |
| sf-slack | <pgielda> those machines had a lot of RAM | 20:15 |
| sf-slack | <pgielda> I have a lot of RAM everywhere just because of software like that, otherwise mostly eaten by the browser or not used ;) | 20:16 |
| hackerfoo | pgielda: I think they're all large machines anyway, so the smaller machines are just pieces of a bigger machine. So my theory is that 96 single-core machines will have more overhead than a single 96 core machine, and be far less flexible, and require 96 disks instead of one large shared disk. | 20:17 |
| sf-slack | <pgielda> you can actually attach same read only disk to multiple machines in GCP | 20:17 |
| hackerfoo | And it would be worse if they were actually individual machines. | 20:18 |
| sf-slack | <pgielda> it just gets slower but to a certain point it scales | 20:18 |
| sf-slack | <pgielda> but I agree | 20:18 |
| hackerfoo | I thought of that. I think a ZFS disk can be shared read/write. I haven't tried it yet. | 20:19 |
| sf-slack | <pgielda> also bigger drives are faster in GCP I think so to a certain point it makes sense to make this drive bigger than needed | 20:19 |
| sf-slack | <pgielda> but its all slightly too complicated when you can run one big machine, I agree | 20:20 |
| hackerfoo | daveshah: I have a Hades Canyon NUC (i7, 32GB RAM, nvme) that outruns my Xeon workstation by 1.5x-2x single threaded, and probably beats the cloud machines by 2-4x. I think this is partly clock speed, but that doesn't entirely explain the difference. It might be better at handling the heat (ironically), because it's a 100C part with massive blowers on it. | 20:34 |
| daveshah | For short term single threaded workloads the i9 in my laptop beats my Threadripper desktop | 20:36 |
| daveshah | Definitely not the case for more than a minute or two though | 20:36 |
| hackerfoo | It would be interesting to collect data to determine the best machine. We could make an FPGA benchmark suite. | 20:37 |
| *** OmniMancer has joined #symbiflow | 21:16 | |
| -_whitenotifier-9- [vtr-xml-utils] acomodi opened issue #11: Add license automated check in Travis - https://git.io/JfITh | 21:22 | |
| -_whitenotifier-9- [symbiflow-xc7z-automatic-tester] acomodi opened issue #3: Add license automated check in Travis - https://git.io/JfITj | 21:22 | |
| -_whitenotifier-9- [python-sdf-timing] acomodi opened issue #40: Add license automated check in Travis - https://git.io/JfIke | 21:23 | |
| -_whitenotifier-9- [symbiflow-bitstream-viewer] acomodi opened issue #6: Add license automated check in Travis - https://git.io/JfIkv | 21:23 | |
| -_whitenotifier-9- [fasm] acomodi opened issue #22: Add license automated check in Travis - https://git.io/JfIkf | 21:23 | |
| -_whitenotifier-9- [symbiflow-arch-defs] acomodi opened issue #1438: Add license automated check in Travis - https://git.io/JfIkJ | 21:23 | |
| -_whitenotifier-9- [prjxray] acomodi opened issue #1306: Add license automated check in Travis - https://git.io/JfIkU | 21:23 | |
| -_whitenotifier-9- [sv-tests] acomodi opened issue #773: Add license automated check in Travis - https://git.io/JfIkI | 21:25 | |
| *** OmniMancer has quit IRC | 21:26 | |
| *** OmniMancer has joined #symbiflow | 21:27 | |
| *** OmniMancer1 has joined #symbiflow | 21:29 | |
| *** OmniMancer has quit IRC | 21:31 | |
| *** tcal has quit IRC | 21:47 | |
| *** somlo has quit IRC | 21:47 | |
| *** tcal has joined #symbiflow | 21:51 | |
| *** somlo has joined #symbiflow | 21:51 | |
| *** craigo has quit IRC | 21:58 | |
| *** tim_ has joined #symbiflow | 22:20 | |
| *** tim_ is now known as Guest16974 | 22:20 | |
| *** Guest16974 has quit IRC | 22:27 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!