--- Log opened Sun May 15 00:00:23 2022
00:04 < fenn> https://libre-soc.org/ "We're building a chip. A fast chip. A safe chip. A trusted chip. A chip with lots of peripherals. And it's a VPU. And it's a 3D GPU... Oh and here, have the Libre-SOC source code..." https://git.libre-soc.org/
00:36 < fenn> .title https://f4pga.org/
00:36 < saxo> F4PGA - the GCC of FPGAs
01:59 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
02:41 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
03:28 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has joined #hplusroadmap
04:41 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::93] has joined #hplusroadmap
05:00 -!- xaete_ [~user12345@2607:9880:1a40:73:ec39:34b:2f4f:b0ba] has quit [Ping timeout: 240 seconds]
05:01 -!- xaete [~user12345@cpe688f2ec6e533-cm688f2ec6e530.cpe.net.fido.ca] has joined #hplusroadmap
06:43 < kanzure> that seems to be lkcl
06:46 < kanzure> .wik Libre-SOC
06:47 < saxo> "Libre-SOC is a libre soft processor core originally written by Luke Leighton and other contributors, announced at the OpenPOWER Summit NA 2020." - https://en.wikipedia.org/wiki/Libre-SOC
06:54 < kanzure> .tw https://twitter.com/lkcl/status/1525836924413435906
06:54 < saxo> @kanzure real-time tamper-detection through deliberate-unstable PLLs routed throughout the ASIC: the detection method is the change in phase. total opposite of making a PLL stable. any change in the program being run will create a detectable EM signature https://libre-soc.org/SEP-210803722-Libre-SOC-8-core/ (@lkcl, in reply to tw:1525058975389171712)
07:07 < kanzure> 07:07 < lkcl> we're adding Galois Field operations as first-level primary opcodes
08:43 < juri_> can confirm, definately lkcl.
09:05 < lsneff> Galois field ops? Why?
10:32 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
12:36 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds]
12:50 < fenn> "Open Pre-trained Transformer Language Models by Meta AI" supposedly a GPT-3 equivalent, open source (apache license)  https://github.com/huggingface/transformers/releases/tag/v4.19.0
13:00 < fenn> "Train and Deploy Transformer models with Amazon SageMaker and Hugging Face DLCs"
13:01 < fenn> .wik galoid field
13:01 < fenn> derp
13:01 < saxo> Article not found: https://en.wikipedia.org/wiki/Galoid_field gave 404 | Searched en for 'galoid field' | https://en.wikipedia.org/wiki/No_result_found gave 404 | Searched en for 'No result found'
13:01 < fenn> .wik galois field
13:01 < saxo> "In mathematics, a finite field or Galois field (so-named in honor of Évariste Galois) is a field that contains a finite number of elements.  As with any field, a finite field is a set on which the operations of multiplication, addition, subtraction and division are defined [...]" - https://en.wikipedia.org/wiki/Galois_field
13:02 < fenn> "The most common examples of finite fields are given by the integers mod p when p is a prime number."
13:03 < fenn> presumably for simplifying crypto code
13:04 < fenn> maybe it reduces wasted memory in the gaps between array elements
13:06 < fenn> it's useful for bitwise operations, but i don't understand why or how
13:10 < fenn> it might be the case that when a computer engineer says "galois fields" they mean "galois fields of integers 2^N"
13:21 < kanzure> https://fosdem.org/2022/schedule/event/libresilicon/
13:23 < kanzure> https://nlnet.nl/pres/20210518/OHW_chips/DavidLanzendoerfer-LibreSilicon.pdf
13:23 < kanzure> https://libresilicon.com/
13:30 < maaku> fenn: that is usually the case (galois fields of integers 2^N)
13:30 < maaku> but in this case it does sound like a crypto accelerator. maybe fast modular arithmetic?
13:41 < fenn> `josuah> F4PGA did not really do anything on their own, more like a structure to receive/distrubte funds, documenting other project's effort, and round the corners for putting a first foot in the fpga world
13:45 < fenn> https://joelw.id.au/FPGA/CheapFPGADevelopmentBoards
15:04 -!- juri_ [~juri@178.63.35.222] has quit [Ping timeout: 260 seconds]
15:20 -!- Malvolio is now known as Guest4289
15:20 -!- Guest4289 [~Malvolio@user/malvolio] has quit [Killed (lithium.libera.chat (Nickname regained by services))]
15:33 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has quit [Remote host closed the connection]
15:40 -!- Malvolio [~Malvolio@user/malvolio] has joined #hplusroadmap
16:33 < kanzure> https://github.com/internet2-WG/rust-aluvm
16:40 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
16:47 < lsneff> That page has a pretty poor comparison with wasm
16:48 < lsneff> Might write an HDL implementation of that
17:09 -!- mirage33523 [~mirage335@2607:fb91:143e:22c:2cf0:a4ca:27c0:3214] has joined #hplusroadmap
17:09 -!- mirage33523 [~mirage335@2607:fb91:143e:22c:2cf0:a4ca:27c0:3214] has quit [Client Quit]
17:10 -!- mirage33523 [~mirage335@2607:fb91:143e:22c:2cf0:a4ca:27c0:3214] has joined #hplusroadmap
17:11 < lsneff> the spec (https://github.com/Internet2-WG/aluvm-spec) has barely any information about the ISA
17:11 < lsneff> definitely wouldn't call that a spec
17:12 -!- mirage335 [~mirage335@64.79.52.86] has quit [Ping timeout: 252 seconds]
17:14 -!- mirage3352397 [~mirage335@2607:fb91:1407:c422:55f0:bb3b:e86:c5af] has joined #hplusroadmap
17:15 -!- mirage335239732 [~mirage335@172.58.191.160] has joined #hplusroadmap
17:17 -!- mirage3352397329 [~mirage335@172.58.191.18] has joined #hplusroadmap
17:18 -!- mirage33523 [~mirage335@2607:fb91:143e:22c:2cf0:a4ca:27c0:3214] has quit [Ping timeout: 252 seconds]
17:18 -!- mirage3352397 [~mirage335@2607:fb91:1407:c422:55f0:bb3b:e86:c5af] has quit [Ping timeout: 252 seconds]
17:21 -!- mirage335239732 [~mirage335@172.58.191.160] has quit [Ping timeout: 252 seconds]
17:21 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 260 seconds]
17:24 -!- mirage3352397329 [~mirage335@172.58.191.18] has quit [Quit: Client closed]
17:26 < muurkha> fenn: there are different Galois fields that can be used for computer engineering applications, and it is fairly unusual to use the multiplication operation of GF(2ⁿ), though for example CRC calculation and LFSR desing do use it
17:27 < muurkha> the reason for using Galois fields rather than some other kind of field in crypto is to make all the field elements and operations on them computable, and in particular computable in bounded time
17:28 < muurkha> real numbers are a non-Galois field but almost all of them are uncomputable
17:28 < muurkha> curve25519 uses GF(2²⁵⁵ - 19), which is not an example of GF(2ⁿ)
17:29 < muurkha> as I understand it
17:29 < muurkha> s/desing/design/
17:49 < lsneff> Anyone know how I could get my hands on a cheap, small X-ray tube?
17:49 < lsneff> Thinking an open-source xrf gun could be cool
17:50 < muurkha> for that you need a high-voltage X-ray tube
17:50 < superkuh> Or you can get an Amptek Cool-X pyroeletric x-ray generator.
17:50 < superkuh> Very small, minimal power needed (just heat cycle the crystal under dilute vacuum)
17:52 < muurkha> neat
17:52 < muurkha> thus no high-voltage power supply needed
17:53 < lsneff> The cycling sounds like it’s unpredictable
18:48 < fenn> lsneff: how about an alpha particle source from a smoke detector
18:49 < fenn> much slower but no high voltage electronics
18:49 < fenn> or vacuum
18:49 < fenn> and then a PIN diode for the detector
18:50 < fenn> or rather, a crapload of them, to get enough signal in a reasonable time
18:52 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::93] has quit [Quit: Leaving]
19:12 -!- xaete [~user12345@cpe688f2ec6e533-cm688f2ec6e530.cpe.net.fido.ca] has quit [Ping timeout: 240 seconds]
19:49 -!- xaete [~user12345@72.138.138.18] has joined #hplusroadmap
20:45 < muurkha> fenn: is the idea that the alpha particles produce secondary X-rays by impacting the sample under test?
20:45 < fenn> i think they cause x-ray fluorescence directly
20:45 < muurkha> what do you mean by "cause x-ray fluorescence directly"?
20:45 < fenn> it's been 10 years since i looked into it
20:46 < fenn> i mean they stimulate the atom by banging into it, and then the excited electrons(?) give off x-rays?
20:46 < muurkha> ah, that's what I meant, ok
20:47 < muurkha> I was confused because the structure of your reply seems to suggest that you meant it as a mutually exclusive alternative
20:47 < fenn> well it wouldn't be "secondary x-rays" if there is only one x-ray photon involved
20:48 < muurkha> oh, I see
20:48 < muurkha> good point!
20:48 < muurkha> the eclipse is currently spectacular btw
20:55 < fenn> eclipsed by clouds
20:59 < muurkha> there's another hour or so of totality
21:00 < muurkha> maximum is in 11 minutes
21:02 < superkuh> Yeah, really pretty here.
22:08 -!- xaete [~user12345@72.138.138.18] has quit [Ping timeout: 248 seconds]
22:15 -!- xaete [~user12345@72.138.138.18] has joined #hplusroadmap
22:41 -!- Netsplit *.net <-> *.split quits: EnabrinTain, s0ph1a, muurkha
22:42 -!- Netsplit over, joins: muurkha
22:42 -!- Netsplit over, joins: EnabrinTain
22:42 -!- Netsplit over, joins: s0ph1a
22:42 -!- xaete [~user12345@72.138.138.18] has quit [Ping timeout: 240 seconds]
22:48 -!- Netsplit *.net <-> *.split quits: RubenSomsen, potatope
22:48 -!- Netsplit over, joins: potatope
22:48 -!- Netsplit over, joins: RubenSomsen
22:50 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 246 seconds]
22:55 -!- xaete [~user12345@2607:9880:1a40:73:ad59:be3d:97ec:1a59] has joined #hplusroadmap
--- Log closed Mon May 16 00:00:24 2022