*** tpb has joined #yosys | 00:00 | |
*** cr1901_modern has joined #yosys | 00:08 | |
azonenberg | ZirconiumX: down the road i want to build a 2-way superscalar, hyperthreaded rv32i core to see how fast i can push it | 00:13 |
---|---|---|
azonenberg | in-order though, i'm not trying to be BOOM | 00:13 |
azonenberg | i also want to try to build something tiny and size optimized a la picorv32, maybe even start with the picorv32 design, but hand tune it with RPMs to try and squeeze the max clock rate and least area i can out of it | 00:14 |
ZirconiumX | azonenberg: I think there's potential in a very wide in-order core, akin to what Nvidia are trying with Carmel. | 00:14 |
azonenberg | So, i have another design that isnt riscv based that i want to get back to | 00:15 |
ZirconiumX | I presume you've heard of SERV | 00:15 |
azonenberg | it's 3 IPC VLIW but all instructions are ternary and the pipeline is Y-shaped | 00:15 |
azonenberg | specifically meant for cryptographic code | 00:15 |
azonenberg | no integers, possibly not even arbitrary branching | 00:15 |
azonenberg | it's meant to be a black box sitting between a tcp/ip stack and application layer, or similar | 00:16 |
azonenberg | but trying to be agile enough that you can load new algorithms in the future rather than being tied to e.g. AES | 00:16 |
azonenberg | and with guaranteed single cycle execution of all instruction, no caches, no timing side channels | 00:16 |
azonenberg | dedicated rom/ram for instructions, memory mapped fifo for tx/rx data, constant storage/sbox rom, and scratchpad for temporary state | 00:17 |
azonenberg | Basically there's three ALUs that each do r32 op r32 op imm32 | 00:17 |
azonenberg | except the third one operates on the output of the first two, instead of registers | 00:17 |
azonenberg | so you only need one regfile write port despite doing 3 or 5 IPC (depending on if you count a ternary operation as one or two "instructions") | 00:18 |
*** azonenberg_mobil has joined #yosys | 00:20 | |
ZirconiumX | CPU architecture is a very interesting thing | 00:20 |
*** emeb_mac has joined #yosys | 00:23 | |
*** gsi__ has joined #yosys | 00:59 | |
*** vonnieda has joined #yosys | 01:01 | |
*** gsi_ has quit IRC | 01:03 | |
azonenberg_mobil | interesting indeed | 01:16 |
*** pie_ has quit IRC | 01:55 | |
*** emeb_mac has quit IRC | 01:59 | |
*** emeb has left #yosys | 02:39 | |
*** PyroPeter has quit IRC | 02:49 | |
*** PyroPeter has joined #yosys | 03:02 | |
*** azonenberg_mobil has quit IRC | 03:10 | |
*** pie_ has joined #yosys | 03:19 | |
*** emeb_mac has joined #yosys | 03:45 | |
*** _whitelogger has quit IRC | 04:29 | |
*** _whitelogger has joined #yosys | 04:31 | |
*** Max-P has quit IRC | 04:33 | |
*** Max-P has joined #yosys | 04:33 | |
*** citypw has quit IRC | 06:00 | |
*** ravenexp has joined #yosys | 06:30 | |
*** emeb_mac has quit IRC | 06:37 | |
*** citypw has joined #yosys | 07:16 | |
*** X-Scale has quit IRC | 07:18 | |
*** dys has quit IRC | 07:25 | |
*** gsi__ is now known as gsi_ | 07:34 | |
*** dys has joined #yosys | 08:02 | |
*** dys has quit IRC | 08:04 | |
*** trmm has joined #yosys | 09:38 | |
trmm | back on the LPF versus parameters in the verilog; which one wins? if the LPF has a drive strength or slew rate, can the verilog override it? | 09:41 |
*** citypw has quit IRC | 09:47 | |
*** ravenexp has quit IRC | 10:23 | |
*** Jybz has joined #yosys | 10:54 | |
*** Jybz has quit IRC | 10:56 | |
*** mirage335 has quit IRC | 11:18 | |
*** mirage335 has joined #yosys | 11:29 | |
*** pie__ has joined #yosys | 11:36 | |
*** pie_ has quit IRC | 11:37 | |
*** pie__ has joined #yosys | 11:37 | |
*** MoeIcenowy has quit IRC | 11:38 | |
*** MoeIcenowy has joined #yosys | 11:38 | |
*** pie__ has quit IRC | 11:39 | |
*** pie__ has joined #yosys | 11:39 | |
*** pie__ has quit IRC | 11:40 | |
*** pie__ has joined #yosys | 11:41 | |
ZirconiumX | Right, I *believe* I have a live value table implementation | 11:49 |
ZirconiumX | Or at least part of one | 11:53 |
ZirconiumX | Is it preferable to use symbiyosys or yosys-smtbmc? | 12:07 |
ZirconiumX | (maybe a question for ZipCPU) | 12:07 |
*** rrika has quit IRC | 12:31 | |
ZipCPU | SymbiYosys | 12:32 |
ZipCPU | It's much simpler than yosys-smtbmc, gives you access to more solvers, and handles formats and format conversions for you | 12:32 |
*** rrika has joined #yosys | 12:33 | |
*** pie__ has quit IRC | 12:51 | |
*** pie__ has joined #yosys | 12:52 | |
*** pie__ has quit IRC | 12:57 | |
*** X-Scale has joined #yosys | 13:46 | |
*** Stary has joined #yosys | 13:52 | |
*** emeb_mac has joined #yosys | 13:58 | |
*** pie_ has joined #yosys | 14:27 | |
*** pie_ has joined #yosys | 14:28 | |
*** emeb_mac has quit IRC | 14:37 | |
*** rohitksingh has joined #yosys | 15:07 | |
*** Xark has joined #yosys | 15:56 | |
*** emeb_mac has joined #yosys | 16:04 | |
*** rohitksingh has quit IRC | 17:23 | |
*** pie_ has quit IRC | 17:24 | |
*** emeb_mac has quit IRC | 17:25 | |
*** pie_ has joined #yosys | 17:27 | |
*** pie__ has joined #yosys | 18:05 | |
*** pie_ has quit IRC | 18:06 | |
*** emeb has joined #yosys | 19:22 | |
*** maikmerten has joined #yosys | 19:29 | |
maikmerten | my HX8K dev board keeps growing wings: https://pasteboard.co/IkMO5UO.jpg | 19:30 |
tpb | Title: Pasteboard - Uploaded Image (at pasteboard.co) | 19:30 |
*** pie___ has joined #yosys | 19:47 | |
*** pie__ has quit IRC | 19:47 | |
*** pie___ has quit IRC | 20:15 | |
*** dys has joined #yosys | 20:19 | |
*** maikmerten has quit IRC | 20:38 | |
* ZirconiumX sighs | 20:39 | |
ZirconiumX | I really don't understand formal methods | 20:40 |
ZirconiumX | I've read a lot of articles on it | 20:40 |
ZirconiumX | But I just don't 'get' it | 20:40 |
ZirconiumX | I can't even prove that after reset my program counter points to the reset address | 20:41 |
*** lutsabound has joined #yosys | 20:42 | |
*** cr1901_modern has quit IRC | 20:51 | |
*** emeb has quit IRC | 20:52 | |
*** emeb_mac has joined #yosys | 20:58 | |
ZirconiumX | pepijndevos: you ever looked at bit-serial architecture? | 21:03 |
*** cr1901_modern has joined #yosys | 21:20 | |
*** pie_ has joined #yosys | 21:50 | |
*** trmm has quit IRC | 23:38 |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!