azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/azonenberg/scopehal-apps | Logs: https://libera.irclog.whitequark.org/scopehal
Degi has quit [Ping timeout: 256 seconds]
Degi has joined #scopehal
gruetzkopf has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
gruetzkopf has joined #scopehal
<bvernoux> azonenberg, do you confirm UCSC_OpenRAM
<bvernoux> SKY130 is limited to max 40MHz ?
<bvernoux> It is quite disappointing to be so slow do you have any idea (as there is some MCU with SRAM@130nm which reach more than that easily)
<tnt> bvernoux: AFAIK he suspects that the io buffers played a large role in that limit. Which you wouldn't have for an internal user.
<bvernoux> ha ok interesting
<bvernoux> anyway I suspect 130nm will limit max speed to something like 64MHz ?
<tnt> why ?
<tnt> My Athlon 700 MHz was 180nm ...
<bvernoux> yes will be interesting to see what can be reached with the power consumption ;)
<bvernoux> We see for MCU like ARM Cortex they was limited to something like 64MHz with 130nm
<bvernoux> So I think it shall be similar to on RISCV
<tnt> "they" ?
<bvernoux> especially for a good power consumption
<bvernoux> not to be power hungry like Athlon/X86/x64 stuff ;)
<bvernoux> yes RISCV & ARM are similar in term of performance
<tnt> I sure hope to get above 64 MHz, that's slow as hell.
<bvernoux> Yes but it is a tradeof between speed & power consumption too and 130nm also do not help
<tnt> it's not even a tru 130nm process, more like 130/180 hybrid.
<bvernoux> a must have for serious next gen stuff will be to have 45nm but we can always dream ;)
<bvernoux> Anyway even with this hybrid 130/180 hyrbid stuff we could have some cool MCU I think probably limited to <100MHz but depending on peripherals features it could be interesting
<azonenberg> bvernoux: Yeah it can definitely go faster
<azonenberg> 40ish MHz is the fastest i've pushed the current test chip. which is a lower bound on the usable performance of the IP in a proper test harness
<azonenberg> I don't know how much more it can go, they have more test chips on the way that fix some of these issues
bvernoux has quit [Quit: Leaving]
nelgau has joined #scopehal