[J-core] We had a summit thing!

Rob Landley rob at landley.net
Fri Jun 17 20:18:03 EDT 2016

Most of the j-core engineers were in Austin earlier this week, and we
got together and talked about stuff! (Normally we've got people in
Tokyo, Toronto, Virginia, Texas, California, and Jeff's location is "in
a plane".)

Alas, it was pretty time constrained (we were mostly in town for other
reasons, which may have run over ever so slightly) so we didn't get to
cover nearly everything we wanted to, but maybe we can do it again
sometime. (We _were_ going to go down to 6th street in the evening and
tweet/email out to everybody about it so at least local Austinites could
meet up if they wanted to, and then everybody looked at their airplane
schedules and realized that more than one person's plane left at 6:30
am, and the rest a couple hours later. Put a damper on the evening's
activities, that did.)

Lots of small news items, but they're things like "Niishi-san gave Rich
Felker enough information about how to program the DMA that he can
probably get the sdcard and ethernet using DMA in a release or two",
which would be really fun if the non-DMA version of the Ethernet and
sdcard drivers had gone upstream yet. (Although I think they're both in
Rich's git tree if you can guess the right branch to look at. I keep
having to ask him. Since the numato board hasn't got ethernet and we
STILL haven't got the turtle kickstarter up, yes we talked about that
but it was one of the agenda items we didn't get to because of time

Another "not sure how interesting this is to anyone else" is that we may
wind up doing a test asic of 2-way SMP later this year, which means we
can post actual speed and power consumption numbers on the 150 nanometer
process. (Alas, no kickstarter to let you get your own chips this time,
it's a very small production run. One of those "multi project wafer"
things. The hardware guys had a nickname for it but I keep thinking
"sidesaddle" which isn't it. Total unit volume in the tens, not
thousands, and a half-dozen places we need to send them to. Not a lot of
spare chips.)

As usual, we all came out with a longer todo list than we started with.


More information about the J-core mailing list