_Chris_ has quit [Remote host closed the connection]
Kabouik has quit [Ping timeout: 240 seconds]
ArturShaik has joined #neo900
xmn has quit [Quit: Leaving.]
ctag_ has joined #neo900
ctag has quit [Ping timeout: 256 seconds]
<Joerg-Neo900>
Donnager: help
<Donnager>
Joerg-Neo900: I am a ban management and anti-spam bot, using this plugin: <https://github.com/ncoevoet/ChanTracker>. If you need help, ask the user who invited me to your channel, or ask pinkieval.
ceene has quit [Ping timeout: 256 seconds]
arcean has joined #neo900
arcean has quit [Remote host closed the connection]
arcean has joined #neo900
* enyc
meows
_whitelogger_ has joined #neo900
ceene has joined #neo900
<Joerg-Neo900>
folks, while waiting for metacollin to show up again, let's do something useful and specify a few details for STEP2. To start with - which SoC to use?
<sn0wmonster>
how about lowrisc (RISC-V)?
<sn0wmonster>
is that gonna come out anytime soon?
<sn0wmonster>
its something the Raspi guys founded (linux embedded chip)
<anon->
are you guys going to wait for him?
<sn0wmonster>
their webite says it is expected to finish this year
<sn0wmonster>
it would solve the whole TI chip issue at least
<sn0wmonster>
(althoguh i have no clue how that would complicate the OS integration or anything else)
<sn0wmonster>
the Linux kernel recently added RISC-V support iirc
pkill9 has joined #neo900
<sn0wmonster>
buti guess its a matter of the individual distros incorporating it?
<sn0wmonster>
anon-, i don't think there's any point in waiting any longer
<Joerg-Neo900>
that chip is planned to have a GPU? which one?
<sn0wmonster>
but as there's no one else available, its kind of the default position :)
<sn0wmonster>
i dont know if it has a GPU, where would i find that?
<sn0wmonster>
How is this different to Raspberry Pi?
<sn0wmonster>
not include a GPU.
<sn0wmonster>
Our goals and focus are quite different. The Raspberry Pi exists to improve computer science education worldwide. A fully open-source SoC is not necessary to reach this goal, or even a pragmatic way of achieving it given the timescales involved. While it is an ultimate goal to support all the features of a modern commercial SoC, it will require a number of iterations of the design to achieve this. For example, early versions of our SoC will
<Joerg-Neo900>
I'm very wary about using vaporware chips that have not even proven to work at all. Odds are the chip has SiErr bugs you can't work around, in first few releases
<Joerg-Neo900>
only THEN the chip could get benchmarked if it *really* performs in all relevant aspects
<Joerg-Neo900>
then it might turn out the instruction set of the CPU or the system design (It's SYSTEM on a Chip, remember?) has a flaw in the concept that - while irrelevant for e.g. rPi - makes it not suited for battery powered embedded
<sn0wmonster>
nice
<sn0wmonster>
long haul then rey
<sn0wmonster>
ey*
<sn0wmonster>
Dec 2015: We have untethered the Rocket SoC, which now runs standalone on the KC705 and Nexys4DDR FPGAs. This release also includes support for simulation using Verilator and is accompanied by extensive documentation.
<sn0wmonster>
seems like they are basing this new one off of something that was already around