11:50.57 | *** join/#arm-netbook infobot (ibot@rikers.org) |
11:50.57 | *** topic/#arm-netbook is arm-netbook: Don't ask to ask. Just ask! - EOMA68 spec http://bit.ly/ZHqfxA - ML arm-netbook@lists.phcomp.co.uk - Logs http://bit.ly/XELKJq or http://bit.ly/15OCYPD - For Allwinner/sunxi discussions prefer #linux-sunxi - iMX6 Card http://bit.ly/XAUGCD - Flying Squirrel http://bit.ly/WgOYaR - Allwinner A10/A20 Card http://bit.ly/zBLAbY - http://rhombus-tech.net |
12:52.15 | buZz | https://www.avantek.co.uk/store/avantek-32-core-cavium-thunderx-arm-desktop.html |
12:52.18 | buZz | man |
12:52.20 | buZz | it might not be a netbook |
12:52.24 | buZz | but i totally want one |
12:54.35 | *** join/#arm-netbook leviathan (~quassel@pcd566195.netvigator.com) |
12:55.54 | *** join/#arm-netbook oaken-source (~oaken-sou@p5DDB5183.dip0.t-ipconnect.de) |
13:01.42 | jn__ | buZz: btw, the Talos II workstation (with pretty much entirely owner-controllable firmware) has been shipping for a few weeks now. It's POWER9 instead of something that implements arm64, though |
13:02.36 | buZz | yeah powerpc isnt sexy |
13:02.41 | buZz | by a long shot |
13:02.42 | buZz | :) |
13:03.26 | buZz | also, just the motherboard of a talos already costs more than that entire desktop |
13:06.10 | *** join/#arm-netbook OmegaPhil (~kvirc@fsf/member/OmegaPhil) |
13:08.52 | jn__ | avantek certainly has a price advantage here |
13:10.04 | buZz | and they are just a reseller of that board even |
13:10.15 | buZz | so you can get it for less, probably |
13:11.53 | jn__ | but yeah, pretty cool to see a non-x86 desktop machine at all |
13:12.29 | buZz | uhuh |
13:12.49 | buZz | i feel kinda forced to now get one, to make sure those companies making them will stay around |
13:12.56 | buZz | sucks :) |
13:14.08 | jn__ | if i got one, i'd probably want to port OpenBMC to it, and that's a looong project, that realistically depends on having schematics |
13:14.49 | buZz | that old facebook project? |
13:15.01 | jn__ | it has outgrown facebook |
13:15.21 | buZz | i hope it outgrows yocto soon aswell :P |
13:15.23 | jn__ | at least IBM and google are also onboard |
13:15.23 | buZz | for your sake |
13:16.34 | jn__ | yeah⦠i never really used yocto myself, but it looks complicated from afar |
13:17.40 | buZz | i dont think you can find a single openembedded dev that now says moving to yocto was a good idea :D |
13:31.41 | jn__ | oh wait⦠there is actually a difference between "desktop" and "workstation" |
13:31.45 | jn__ | https://www.avantek.co.uk/store/avantek-32-core-cavium-thunderx-arm-desktop.html |
13:31.49 | jn__ | https://www.avantek.co.uk/store/avantek-thunderx2-arm-workstation-thunderx2station.html |
13:32.31 | jn__ | the workstation is a lot more expensive and has a BMC. the desktop doesn't have one |
13:39.59 | buZz | i highly doubt general workstation usage requires a BMC :) |
13:40.05 | buZz | or even desktop usage |
13:41.10 | jn__ | i guess a "workstation" is at the boudary of desktop and server, so a BMC and remote management kind of make sense |
13:44.28 | buZz | i wouldnt get any work done on a system knowing there could be a remote admin looking over my shoulder at any time :P |
13:52.27 | jn__ | that's why i wouldn't want a closed-source bmc firmware |
13:53.15 | buZz | i dno how a open-source bmc would save me from a spying administrator |
13:53.23 | buZz | cant he still just do that |
13:53.37 | buZz | isnt that the whole idea of a BMC's existance |
13:53.52 | jn__ | it doesn't help when you distrust your admin |
13:54.07 | jn__ | but it helps when you are or when you trust the admin |
13:57.38 | buZz | i dno why i would want to remote admin the machine i'm sitting at |
13:58.40 | jn__ | it makes more sense for server use |
13:58.47 | buZz | :) |
13:59.02 | buZz | i gather, i like using ILO stuff on server |
13:59.10 | jn__ | i.e. you're travelling but your workstation at home is powered off, so you power it on to run some big computations |
13:59.28 | buZz | so i dont need to move to the ice cold server room (who ever thought cpus should use LESS power was a idiot, we're all freezing in DCs now) |
13:59.44 | buZz | WakeOnLan isnt rocketscience |
14:00.51 | jn__ | another usecase: run your irc bouncer on the BMC ;) |
14:02.32 | buZz | rofl |
14:02.48 | buZz | does it cost more than a raspberry pi zero? |
14:02.59 | buZz | if so, i dont see it as a win to use it to replace that ;) |
14:03.53 | jn__ | just a few orders of magnitude more ;) |
14:05.09 | jn__ | i wish avantek would provide more information about the ThunderX desktop⦠at least they list the SoC's model number, which they don't list for the workstation |
14:05.23 | jn__ | that photo is nice, but doesn't say anything about the machine :/ |
14:05.25 | buZz | :P |
14:05.46 | buZz | yeah but you do realize avantek doesnt produce these? |
14:05.55 | jn__ | THEY DON"T TELL ME |
14:06.16 | buZz | gigabyte and asus both make Cavium ATX boards |
14:06.48 | buZz | bit sad that cavium doesnt make some themselves |
14:06.58 | buZz | maybe just overexpensive reference implementation or something |
14:07.41 | buZz | https://cavium.com/news/gigabyte-announces-thunderxstation-industry-s-first-armv8-workstation-based-on-cavium-s-thunderx2-processor |
14:07.49 | buZz | some more info? maybe enough for you |
14:07.55 | buZz | AMI MegaRAC BMC that says |
14:08.21 | jn__ | that's sounds like the workstation sold by avantek |
14:11.00 | jn__ | the thing is: even if i find some information about a mainboard, i can't be sure that it's the one that avantek uses, so avantek's information policy(?) is still crap |
14:12.33 | buZz | well, they probably have email addresses |
14:12.36 | buZz | you could just ask ;) |
14:12.53 | buZz | or just buy a board directly |
14:14.35 | jn__ | yes, you're right :/ |
14:30.06 | lkcl | cavium networks are pretty cool |
14:36.25 | jn__ | cavium also has a BSD-licensed BDK (BIOS dev kit?) for at least CN81xx, but i'm not sure how to obtain it |
14:39.40 | buZz | cute :) |
14:39.50 | buZz | jn__: walk in with a assaultchair |
14:40.04 | buZz | they wont know how to react so have no choice but to hand it over |
14:40.35 | jn__ | heh :> |
14:45.39 | jn__ | (https://github.com/coreboot/coreboot/tree/master/src/vendorcode/cavium here's a copy, but i don't know how to get it *from cavium*. it probably requires signing an NDA) |
14:48.55 | buZz | licenses on those files says they can be redistributed freely |
14:49.02 | buZz | so doubt it requires a NDA |
14:49.15 | buZz | if it does, good luck getting me to sign it cavium! |
15:10.48 | *** join/#arm-netbook isacdaavid (~isaac@189.215.145.143.cable.dyn.cableonline.com.mx) |
15:30.49 | lkcl | jn__, there's nothing to say. it's a chip, with bare PCIe, bare RAM socket, and SATA connections. |
15:31.28 | lkcl | that's it. oh and 10GE. |
15:31.41 | lkcl | anything you want you plug in as PCIe cards. |
15:31.48 | lkcl | there's really not a lot to say. |
15:58.11 | buZz | i think jn__'s specifically talking about the BMC |
16:00.56 | jn__ | i talked about a lot, so what was the context? |
16:02.11 | jn__ | cavium doesn't make BMCs. Aspeed and Nuvoton do. and the "AMI MegaRAC⢠BMC" firmware is from AMI, i guess |
17:50.13 | *** join/#arm-netbook oh2ftg (~oh2ftg@dd7-3pyft4kh94zjhbdlt-3.rev.dnainternet.fi) |
18:08.06 | *** join/#arm-netbook opbolt (~starcard@host-92-15-35-104.as43234.net) |
18:46.10 | *** join/#arm-netbook focus (~jm@fpc83061-enfi22-2-0-cust5.20-2.static.cable.virginm.net) |
20:31.39 | *** join/#arm-netbook focus (~jm@81.141.72.246) |
20:34.08 | *** join/#arm-netbook agusanasuga (~nasuga@unaffiliated/nasuga) |
20:49.47 | *** join/#arm-netbook focus (~jm@81.141.72.246) |
21:24.24 | *** join/#arm-netbook Gamayun (~Gamayun@163.47.115.18) |
22:12.49 | *** join/#arm-netbook agusanasuga (~nasuga@unaffiliated/nasuga) |