04:54 | illwieckz | left the channel | |
04:54 | anuejn | left the channel | |
04:54 | felix_ | left the channel | |
04:54 | BAndiT1983 | left the channel | |
04:54 | vup | left the channel | |
04:54 | felix_ | joined the channel | |
04:55 | vup | joined the channel | |
04:55 | anuejn | joined the channel | |
04:55 | BAndiT1983 | joined the channel | |
04:55 | illwieckz | joined the channel | |
05:02 | dos1 | left the channel | |
05:03 | dos | joined the channel | |
05:14 | tpw_rules | left the channel | |
05:17 | tpw_rules | joined the channel | |
06:39 | Bertl | off to bed now ... have a good one everyone!
| |
06:40 | Bertl | changed nick to: Bertl_zZ
| |
08:37 | aombk2 | left the channel | |
08:39 | aombk | joined the channel | |
08:40 | aombk2 | joined the channel | |
08:44 | aombk | left the channel | |
14:43 | Bertl_zZ | changed nick to: Bertl
| |
14:43 | Bertl | morning folks!
| |
14:57 | yasmin | joined the channel | |
16:00 | se6astian | MEETING TIME, who is here?
| 16:00 | Bertl | is here ...
|
16:01 | yasmin | Yasmin is here
| |
16:01 | se6astian | great, yasmin how are things going on your end?
| |
16:03 | yasmin | Great actually :D
| |
16:03 | yasmin | finally the installation problem was solved
| |
16:03 | yasmin | and i will deliver my first task to betrl today
| |
16:03 | se6astian | very good, wat was your first task?
| |
16:04 | yasmin | i will do the fpga flow for a code and return the results and the schematics for Betrl
| |
16:05 | yasmin | the code it the top module of the peripheral part if remember right
| |
16:05 | Bertl | basically updating the gateware to vivado 2021.x
| |
16:05 | yasmin | i did it before in the 2018 version and i will repeat it for the 2021 version this time
| |
16:05 | yasmin | yes :D
| |
16:05 | yasmin | 2021.1
| |
16:06 | se6astian | ah, I see
| |
16:07 | se6astian | right
| |
16:07 | se6astian | anything else you want to share yasmin?
| |
16:07 | yasmin | no
| |
16:09 | se6astian | right, many thanks !
| |
16:09 | se6astian | quick updates from my side: BAndiT1983 and me worked on a raw12 viewer
| |
16:10 | se6astian | a small gui that shows image content in monochrome or color preview quickly
| |
16:10 | se6astian | today I added switching between next/previous images in same dir
| |
16:10 | se6astian | and BAndiT1983 brought loading and displaying logic to below 1 second on my rather old machine
| |
16:10 | se6astian | so quite useable already
| |
16:11 | se6astian | different zoom/decimation levels are still on the todo list
| |
16:11 | se6astian | gui elements are there already
| |
16:11 | se6astian | if you want to test it
| |
16:11 | se6astian | https://github.com/apertus-open-source-cinema/misc-tools-utilities/blob/hdmi-raw-2021/raw-via-hdmi/scripts/viewer.py
| |
16:11 | se6astian | just provide the path to a raw12 image as parameter
| |
16:12 | se6astian | here is the matching lab task: https://lab.apertus.org/T1276
| |
16:12 | se6astian | holiday today in austria so I was not in our office to complete the hdmi plugin module assembly
| |
16:12 | se6astian | but planned soon
| |
16:12 | se6astian | thats it from me, Bertl do you also have news for us?
| |
16:13 | Bertl | not much, but a few small pieces of information ...
| 16:13 | vup | is here
|
16:13 | Bertl | vup: got something to report?
| |
16:13 | vup | nope
| |
16:14 | Bertl | okay, so I discovered the Aluminum (Aluminium) PCB feature pooling services offer now (some of them for very cheap)
| |
16:15 | Bertl | and thought to me, that might come handy for heat sink prototyping and similar, e.g. for our below sensor heat transport
| |
16:15 | se6astian | oshpark?
| |
16:15 | Bertl | JLCPCB and MultiPCB
| |
16:15 | se6astian | ah
| |
16:16 | Bertl | I ordered some boards just for testing and it seems they are quite nice, so we can use that in the future for prototyping
| |
16:17 | Bertl | (will upload some images later, but nothing unexpected there)
| |
16:17 | cscar | joined the channel | |
16:17 | cscar | Sorry I'm late. But hi!
| |
16:17 | Bertl | on the Magewell USB front, after se6astian got feedback about using sudo/root for the SDK examples
| |
16:18 | se6astian | yes, was about to ask about the sdk based prototype you were creating
| |
16:18 | se6astian | hi oscar!
| |
16:18 | Bertl | I did some more investigations and it turns out that we should be able to address/avoid that by using a proper udev rule for the device
| |
16:19 | Bertl | will prepare and test something in the next few days, so we should have a hash check tool which works with both, PCIe and USB devices
| |
16:19 | Bertl | that's it from my side for this week
| |
16:20 | se6astian | great, many thanks
| |
16:20 | se6astian | any news to report oscar?
| |
16:21 | cscar | Not much. The yearly tax return papers arrived today. All good.
| |
16:22 | cscar | Meaning no taxes ;)
| |
16:23 | se6astian | good :)
| |
16:24 | se6astian | right, anyone else who wants to share/report anything at the meeting?
| |
16:27 | se6astian | right then, many thanks everyone who participated! MEETING CONCLUDED
| |
16:27 | yasmin | Thank you ^^
| |
16:27 | cscar | Thanks!
| |
16:27 | vup | yasmin: are you using the ml enabled version of vivado?
| |
16:28 | cscar | left the channel | |
16:28 | yasmin | yes ML standard
| |
16:28 | vup | interesting, did you have a chance to compare the speed to previous versions?
| |
16:28 | vup | its supposed to be a lot faster
| |
16:29 | vup | (and also the QoR is supposed to be better)
| |
16:29 | yasmin | unfortunately no but this is interesting :D
| |
16:29 | vup | hmm unfortunate
| |
16:30 | vup | seems like I need to free up some space on my ssd and try myself sometime
| |
16:31 | yasmin | :((
| |
16:31 | vup | how much space is it taking for you?
| |
16:32 | yasmin | rounf 65 or 66 GB
| |
16:32 | yasmin | round*
| |
16:32 | vup | holy shit
| |
16:32 | vup | 2019.2 is only taking 19Gb and thats with the SDK
| |
16:32 | yasmin | yes its HUGE :D its installation file was 52 GB, 2018 version was less than 18 GB :D
| |
16:33 | vup | yep
| |
16:33 | vup | seems like they have gained some serious weight
| |
16:33 | Bertl | performance seems to be better, place and route didn't change too much, although it seems to benefit from repeated implementations
| |
16:33 | Bertl | there is a lot of additional stuff in the package
| |
16:33 | vup | you mean they are doing incremental pnr somehow?
| |
16:34 | Bertl | vivado has reused the previous placement results for re-runs for some time now
| |
16:34 | vup | since when?
| |
16:34 | Bertl | so if you do not scratch the data from the previous build, results will be reused
| |
16:35 | Bertl | since 2017.x or so
| |
16:35 | vup | interesting
| |
16:35 | vup | do you have any feeling for how well this works if you flatten the hierarchy?
| |
16:35 | vup | or is this more hierarchy / single module pnr based?
| |
16:36 | Bertl | it seems to work fine with a flattened hierarchy if you have 'rebuild' enabled
| |
16:36 | vup | hmm makes sense
| |
16:36 | Bertl | which helps to get meaningfull debug/schematic/simulation signals anyway
| |
16:37 | vup | well I am doing all debugging / simulation with nmigen, so I don't really care about that
| |
16:38 | vup | but this means this could reasonable work reasonable well with nmigen, as that now actually generates modules (as opposed to migen which always only generated a single huge module)
| |
16:39 | vup | s/reasonable//
| |
16:40 | Bertl | might definitely be worth a try
| |
16:53 | yasmin | left the channel | |
17:39 | Bertl | off for now ... bbl
| |
17:39 | Bertl | changed nick to: Bertl_oO
|