leaseable mainframes, sega genesis, cough cough
Does anyone has any idea what web UI software they are using for hosting this cloud-like interface? From their GitHub page it seems that Minicloud is powered by OpenStack.
Libre-SOC Releases First Non-IBM OpenPOWER Chip in Decade
If only it was on a tinier Manufacturing Process Node.
Seriously, 180nm?
Can’t we at least get GloFlo 12 nm?
I think they need to prototype on the cheap and those older nodes are perfect for prototyping, before going down to 12nm. Keep in mind Intel and AMD are taking some time just to port the current gen chips to a lower manufacturing node (about a year maybe). And they’re behemoths! It will probably take a few years before we see maybe 45nm chips from Libre-SOC.
Libre-SOC is primarily a community project, from what I understand, so I do not think they could afford to get a run on GF 12 nm; especially for a chip that, as far as I know, does not have any customers waiting in the wings to buy enough chips to pay off the likely million dollar/euro cost of designing a chip for a high-end (even if no longer bleeding edge) node.
This particular run on TSMC 180 nm is (I think) a one-time opportunity to let a few (not just Libre-SOC) chip design community projects see real silicon; I think it was paid for by Google; I somewhat doubt that the Libre-SOC project would have been able to afford this out of pocket, given the limited rate at which far simpler community hardware projects like PowerPC Notebook are able to fundraise for far smaller amounts.
I could be wrong though, maybe Libre-SOC does already have sufficient funding to pay for (or get sponsored for) a run on real silicon just for themselves, but if they do it is not enough for a high-end node like GF 12 nm.
Easy? Arguable, perhaps, though inexpensive though it is not.
From what I understand of the project…
They have a fairly sizeable grant from NLNEt (EU) under security umbrella, but with several projects/milestones. So it’s not all available at once.
The current plan is the 180nm is proof of the basic ISA, tool process and core design. A 45nm is planned as microcontroller/BMC/EC type of chip. And then a 12/14 process aimed at being a mobile grade SOC. And they have a couple different customers lined up for both these chips.
POWER10 delay
From this NextPlatform article about Intel potentially purchasing Global Foundries, there was this mention of POWER10 running behind schedule:
While we presume that the high end Power10 chips are set to come out sometime this fall, we have no idea how tough it might be and if Samsung is any good at making complex server chips, which are different from making memory, flash, and low-end client chips. What we can tell you is that the Power10 and System z processors are running later than expected, and this may have been a contributing factor in IBM not being able to win the two big exascale supercomputer deals at Oak Ridge National Laboratory and Lawrence Livermore National Laboratory, which together account for $1.2 billion in spending.
If you are curious, there is a June article about why IBM is suing GlobalFoundries that covers a bit of the history of the switch to Samsung for POWER10.
Welp I guess I better decide if I want a talos fucking SOON
Hi all, I’m founder of 3mdeb and as some of you may know we working on hostboot (POWER firmware developed by IBM) port to coreboot on behalf of Insurgo. First platforms we working on are Raptor Talos II and Talos II Lite.
We looking for Talos II, Talos II lite and POWER9 QEMU testers, if you are intrested feel free to follow this. Please note initial boot time performance show 4x better results. Asciinema from boot process you can find here.
We would appreciate Level1Tech community feedback about hostboot to coreboot port for POWER.
I don’t know that Wendell still has a Talos II machine, IIRC, it was on loan, but I’ll make sure he sees this post.
I will more than likely be replacing most of my stuff with a talos 2 of some sort. Pm me if you have anything new in the future. I’ll be composing ppc g4/5 notes until then.
I’d love to test. I no longer have the system – I sent it back to them.
I had on my list to pick up a blackbird system but never got 'round to it.
Project looks awesome though
and would be glad to tweet about it and drive some traffic to it!
POWER10
This Phoronix article is mainly rephrasing a recent tweet by Raptor and linking to the relevant parts of the OpenPOWER github:
https://www.phoronix.com/scan.php?page=news_item&px=IBM-POWER10-Not-All-Open-FW
The OMI bridge firmware is binary only and the PPE I/O processor is also distributed as just compiled firmware.
Ian Cutress at Anandtech has a general overview about POWER10 and the newly-released IBM E1080 than uses it.
Firefox JIT
Some good non-POWER10 news, there seems to be good progress on the Firefox JIT engine; I am not familiar with Firefox’s JIT architecture, but ClassicHasClass (who has previously worked on TenFourFox) describes this as the first of three steps in the JIT porting process.
TenFourFox
This is in fact the last year of active product-style development of TenFourFox, a port of Firefox to PowerPC Mac OSX; after 2021-10-05 it will be in what ClassicHasClass labels as “hobby mode”, wherein will be no more builds or defined versions, just a rolling release source repository that he will maintain for his own usage. This was announced back in March of this year:
Did you end up starting a group for this? Thanks; still curious 
Idk
Want one
I have not finished reading it, but this appears to be/include a very thorough analysis of the Cell processor.
A few tidbits to help place Cell (specifically its general purpose Power Processing Unit (PPU)) in context of other PowerPC cores:
… IBM constructed a new CPU that follows version 2.02 of the PowerPC specification (which happens to be the last PowerPC spec before being rebranded to ‘Power ISA’).
…
The PPU shares some history with the PowerPC 970 (called G5 by Apple), both are descendants of POWER4
… the PPU implements the PowerPC ISA version 2.02 , including optional opcodes for floating-point square root[12]. It’s also been extended with a group of SIMD instructions called Vector/SIMD Multimedia Extension (VMX). On the other side, some elements are missing from the original specification, such little-endian mode (in fact, Cell only operates in big-endian) and just a handful of opcodes.
I thought that I knew all that there is to know about the PS3’s Cell BE and Cell BE in general, but it seems that this covers some things not even covered by Sony, IBM, PS3 Scene, and some other Linux kernel code.
