31 Mar 2021 |
Rob Taylor | it’d be interesting to see if we can use the google cloud ai platform to optimise openlane runs on gke ;) | 21:11:05 |
mehdi | I have access through Tim/Openroad and it is definitely helpful. But I still feel I am not using the gcp to its full potential. Do you have access to gcp? Maybe you could help me enhance my scripts 😉 | 21:35:21 |
| Brad B joined the room. | 21:58:50 |
Brad B | small question: I am starting to get familiar with the OpenLNAE tools - what exactly does the acronym spm stand for? | 21:59:22 |
Brad B | In reply to@_slack_skywater-pdk_U018R1PNQC9:matrix.org small question: I am starting to get familiar with the OpenLNAE tools - what exactly does the acronym spm stand for? I think I found my answer - it is one of the example designs provided in the openlane repository | 22:04:09 |
Anish | In reply to@_slack_skywater-pdk_U018R1PNQC9:matrix.org I think I found my answer - it is one of the example designs provided in the openlane repository serial-parallel multiplier | 22:04:49 |
Anish | In reply toundefined
(edited) serial-parallel multiplier => serial-parallel multiplier is that it stands for, and yep it's a very-small sample design | 22:04:59 |
Brad B | In reply to@_slack_skywater-pdk_U016PTY0C2E:matrix.org serial-parallel multiplier is that it stands for, and yep it's a very-small sample design Thanks! | 22:05:07 |
1 Apr 2021 |
| Sukanta Dey joined the room. | 04:32:16 |
| Sukanta Dey changed their profile picture. | 04:32:18 |
Olof Kindgren | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org mkk ^ Rob Taylor Actually had a talk with Ahmed Ghazy and Mohamed Shalan yesterday about Edalize, OpenLane and cloud use-cases | 06:26:38 |
Rob Taylor | In reply to@_slack_skywater-pdk_U0169AQ41L6:matrix.org I have access through Tim/Openroad and it is definitely helpful. But I still feel I am not using the gcp to its full potential. Do you have access to gcp? Maybe you could help me enhance my scripts 😉 sure, would be happy to see if i can help | 08:42:51 |
Rob Taylor | In reply to@_slack_skywater-pdk_U016HSA1H5Y:matrix.org Rob Taylor Actually had a talk with Ahmed Ghazy and Mohamed Shalan yesterday about Edalize, OpenLane and cloud use-cases nice :) | 08:43:09 |
Nguyen.Dao | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org nice :) Update: the flow works fine with 1x2 and 2x2 CLBs fabrics without timing optimization (OpenPhySyn doesn’t support disable timing loops). For the 4x4 CLBs fabric, it’s stuck at global routing (fastroute)? | 12:54:10 |
| Jamey Hicks joined the room. | 13:00:27 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01L8BR7FFA:matrix.org Update: the flow works fine with 1x2 and 2x2 CLBs fabrics without timing optimization (OpenPhySyn doesn’t support disable timing loops). For the 4x4 CLBs fabric, it’s stuck at global routing (fastroute)? Does the time go up exponentially? | 13:00:55 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Does the time go up exponentially? Could we produce a 2x2 clb gds and then stitch together to make a bigger fabric? | 13:00:56 |
Nguyen.Dao | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Could we produce a 2x2 clb gds and then stitch together to make a bigger fabric? Running time are almost same: 12s (1x2 clbs) and 13s (2x2 clbs) | 13:05:49 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01L8BR7FFA:matrix.org Running time are almost same: 12s (1x2 clbs) and 13s (2x2 clbs) Interesting. Same for 3x3? | 13:06:55 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Interesting. Same for 3x3? I wonder - are you hitting memory limits and it’s going into swap? | 13:07:14 |
Nguyen.Dao | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org I wonder - are you hitting memory limits and it’s going into swap? I haven’t tried 3x3, need to ask Jing to generate codes. I don’t think memory is limited, it used only 1% of memory (64GB) when routing | 13:09:05 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01L8BR7FFA:matrix.org I haven’t tried 3x3, need to ask Jing to generate codes. I don’t think memory is limited, it used only 1% of memory (64GB) when routing Same when going the 4x4? | 13:09:57 |
Nguyen.Dao | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Same when going the 4x4? yes, same memory at 1% | 13:10:19 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01L8BR7FFA:matrix.org yes, same memory at 1% Very odd. Ok let’s get a 3x3 tested, find out where this breaks and then let’s submit a bug | 13:11:16 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Very odd. Ok let’s get a 3x3 tested, find out where this breaks and then let’s submit a bug Or maybe 2x4 if that’s easier | 13:13:29 |
Nguyen.Dao | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Or maybe 2x4 if that’s easier it’s OK with 2x4, global routing took 16s, looks like the running time doesn’t go exponentially | 15:30:51 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01L8BR7FFA:matrix.org it’s OK with 2x4, global routing took 16s, looks like the running time doesn’t go exponentially Weird! | 15:33:59 |
Rob Taylor | In reply to@_slack_skywater-pdk_U01GE4F371D:matrix.org Weird! Ok, it’d be good to prepare a test case that shows the working version and the non working version and submit that as a bug against openlane | 15:34:36 |
_slack_skywater-pdk_U016ELL4X4M | [efabless/openlane] Issue opened by nguyendao-uom | 16:48:29 |
| Nayiri joined the room. | 21:08:28 |