| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 05:53 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.57.63> has joined #libre-soc | 05:54 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.57.63> has quit IRC | 07:34 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 07:34 | |
| *** Ritish <Ritish!~Ritish@2401:4900:627d:50a2:8af6:53d2:9841:6ccf> has joined #libre-soc | 08:09 | |
| *** Ritish <Ritish!~Ritish@2401:4900:627d:50a2:8af6:53d2:9841:6ccf> has quit IRC | 08:11 | |
| *** Ritish <Ritish!~Ritish@2401:4900:627d:50a2:8af6:53d2:9841:6ccf> has joined #libre-soc | 08:13 | |
| *** Ritish <Ritish!~Ritish@2401:4900:627d:50a2:8af6:53d2:9841:6ccf> has quit IRC | 08:23 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 08:40 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.56.195> has joined #libre-soc | 08:41 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.56.195> has quit IRC | 10:51 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 10:52 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 11:23 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 11:27 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 11:31 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 12:16 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 12:47 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 12:51 | |
| *** lxo <lxo!~lxo@linux-libre.fsfla.org> has quit IRC | 12:58 | |
| openpowerbot | [mattermost] <lkcl> https://libre-soc.org/openpower/sv/rfc/ls005/ | 14:12 |
|---|---|---|
| openpowerbot | [mattermost] <lkcl> i started RFC ls005 - it's real basic (1 page). | 14:12 |
| openpowerbot | [mattermost] <lkcl> it's intended to be submitted as a... a... "radio-button-2", here https://openpower.foundation/isarfc/ | 14:13 |
| *** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has joined #libre-soc | 15:12 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has quit IRC | 15:43 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.56.92> has joined #libre-soc | 15:43 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.56.92> has quit IRC | 17:00 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.42.16> has joined #libre-soc | 17:03 | |
| *** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has quit IRC | 17:03 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@176.59.42.16> has quit IRC | 17:14 | |
| *** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-109-173-83-100.ip.moscow.rt.ru> has joined #libre-soc | 17:14 | |
| lkcl | https://science.slashdot.org/story/22/12/22/1314242/john-cleeses-classic-silly-walk-burns-more-calories-than-a-normal-gait-study-finds | 17:31 |
| *** bagiyal8989 <bagiyal8989!~bagiyal89@2409:4053:2d0c:1e85:4d3e:bd61:a492:a815> has joined #libre-soc | 17:41 | |
| *** bagiyal <bagiyal!~bagiyal@2409:4053:2d0c:1e85:4d3e:bd61:a492:a815> has joined #libre-soc | 17:48 | |
| bagiyal | everyone | 17:48 |
| *** bagiyal8989 <bagiyal8989!~bagiyal89@2409:4053:2d0c:1e85:4d3e:bd61:a492:a815> has quit IRC | 17:49 | |
| *** bagiyal <bagiyal!~bagiyal@2409:4053:2d0c:1e85:4d3e:bd61:a492:a815> has quit IRC | 18:01 | |
| *** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has joined #libre-soc | 18:49 | |
| programmerjake | lkcl: i expanded on the differences of fp16 vs bf16: https://git.libre-soc.org/?p=libreriscv.git;a=commitdiff;h=674ea72e5fb3aae98d97bd19c883e183b7d0f677 | 20:01 |
| programmerjake | bf8 is e5m2 in https://arxiv.org/abs/2209.05433 -- they call it fp8 along with e4m3. that paper is by nvidia, intel, and arm: https://wccftech.com/nvidia-intel-arm-bet-their-ai-future-on-fp8-whitepaper-for-8-bit-fp-published/ | 20:05 |
| lkcl | programmerjake, i'm attempting to keep the RFC to 1 page | 22:21 |
| lkcl | things like that are really good for going into the "discussion" page | 22:23 |
| programmerjake | ah, ok. i don't have my desktop unpacked so idk how many pages the pdf is, the footnote shouldn't be on page 2, that's just ikiwiki not understanding latex | 22:23 |
| lkcl | it's on page 2 because the last line is migrated to page 2 | 22:24 |
| lkcl | https://libre-soc.org/openpower/sv/rfc/ls005/discussion/ | 22:25 |
| programmerjake | k, note the order shown by ikiwiki is not the order in the original markdown | 22:25 |
| lkcl | correct. doesn't matter. the PDF as a submission to the OPF ISA WG is what matters. | 22:26 |
| lkcl | i've recorded the links (and also discussion here) | 22:26 |
| lkcl | keeping this one *real* brief is a priority, sparking discussion, and linking to the things that toshywoshy mentioned | 22:27 |
| lkcl | which he can't talk about :) | 22:27 |
| lkcl | but are definitely relevant | 22:27 |
| lkcl | good links to wikipedia. makes things clear. | 22:28 |
| lkcl | mmmm.... don't IBM use "bfloat16"? | 22:28 |
| lkcl | we discussed this before for another RFC | 22:28 |
| * lkcl ponders... | 22:28 | |
| programmerjake | k, just trying to make sure you didn't mistake the incorrect ikiwiki rendering as meaning guaranteed-2-pages | 22:28 |
| lkcl | not a priority. | 22:29 |
| lkcl | ah yes, ls002. | 22:29 |
| lkcl | https://libre-soc.org/openpower/sv/rfc/ls002/ | 22:29 |
| lkcl | nggh have to retain consistency sigh | 22:29 |
| * lkcl will update FP32 to BFP32 | 22:30 | |
| programmerjake | feel free to replace bf16 with bfloat16, and fp8/16/32/64 with bfp8/16/32/64 | 22:30 |
| programmerjake | imho having both bf16 and bfp16 is too confusing so should use bfloat16 if possible | 22:31 |
| programmerjake | actually, leave fp8 as fp8, since it matches the intel, arm, nvidia paper | 22:32 |
| programmerjake | and since it isn't ieee | 22:32 |
| * lkcl brainmelting | 22:32 | |
| programmerjake | sorry... | 22:32 |
| lkcl | :) | 22:33 |
| lkcl | i'm on semi-vague-holiday-shutdown :) | 22:33 |
| programmerjake | i can do the search/replace | 22:34 |
| lkcl | but figured it'd be sensible to (slowly, no pressure) do this before things restart next year | 22:34 |
| lkcl | appreciated | 22:34 |
| lkcl | remember the conversation with toshaan, he recommended getting this done... soon. | 22:35 |
| programmerjake | so, did you push the last of your changes? | 22:35 |
| programmerjake | rest* | 22:35 |
| programmerjake | i'm assuming so | 22:36 |
| lkcl | yes. | 22:36 |
| lkcl | i'm looking to reduce lines further, but it's a different location | 22:37 |
| lkcl | ok i'm stopping | 22:37 |
| programmerjake | done: https://git.libre-soc.org/?p=libreriscv.git;a=commitdiff;h=5997033c3e258bc2c5c48f147abb1c4a10c2a21f | 22:49 |
| programmerjake | sorry it's all 1 commit, too annoying to use the web interface for multiple commits when copying to/from my text editor | 22:50 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!