_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
Micko has joined #prjcombine
Micko has quit [Client Quit]
Micko has joined #prjcombine
<Wanda[cis]> azonenberg: I'm curious, do you know if anyone ever took die shots of the System ACE CF chip?
<Wanda[cis]> I've been wondering for a long time wtf that thing really is
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #prjcombine
<azonenberg> Wanda[cis]: i'm about 80% sure it was disclosed in some doc or other that it was some generation of coolrunner with a special firmware on it
<azonenberg> but i cant point to anything off the top of my head
<azonenberg> i did decap the Platform Flash but cracked the die. it was made by stmicro on an old-looking process
<azonenberg> i actually found a single XCF02 the other day in my parts bin that had been lurking there for ages and should decap it and grab a nicer photo
<Wanda[cis]> coolrunner huh.
<azonenberg> Looking for the source, if i'm actually remembering correctly. https://www.edn.com/xilinx-to-use-siliconsystems-tech-in-system-ace/ related but not too useful
<Wanda[cis]> it doesn't match xpla3 nor xc2c tq144 pinout
<azonenberg> yeah ok i also see in DS080 that it runs core at 3.3
<Wanda[cis]> ... I guess it could be a custom package option
<azonenberg> maybe it was an xc9500?
<azonenberg> or it could be an asic
<Wanda[cis]> xc9500xl pinout also doesn't match
<Wanda[cis]> cannot be xc9500 or xc9500xl because it has two VCC banks
<azonenberg> welp. guess i remembered wrong
<Wanda[cis]> cannot be xc9500xv either because that'd require 2.5V core supply
<azonenberg> yeah its probably an asic then
<azonenberg> some bog-standard mcu of the era perhaps
<azonenberg> if you get your hands on one i can decap it
<Wanda[cis]> I do have a few on various devboards, though I'd prefer not to destroy them
<azonenberg> yeah thats the tricky bit
<azonenberg> getting one you can sacrifice
<Wanda[cis]> also I think we're like on different continents
<azonenberg> yeah but how much does it cost to ship a tqfp?
<azonenberg> a few postage stamps?
<Wanda[cis]> ... true
<azonenberg> of all the potential blockers, that's not the big one
<Wanda[cis]> also in similar questions
<Wanda[cis]> seen any versal die shots yet?
<azonenberg> win-source claims to have 557 in stock for $54 each in single units. and i see a bunch on aliexpress for low tens of usd each. anybody's guess if they're real
<azonenberg> some are obvious fakes that dont even have a xilinx logo
<azonenberg> a few at least look legit at a glanec
<azonenberg> if you're that interested i can drop $30 and decap whatever shows up
<azonenberg> as scott manley says "excitement guaranteed, we're just not sure what kind"
<azonenberg> i have not seen any ultrascale, ultrascale+, or versal decaps to my knowledge
<azonenberg> this seller claims to have 54K in stock which i doubt (or its a sign that they have 54K blank tqfp's they can relabel depending on how many you order)
<azonenberg> the package is ink printed not laser engraved in that pic. i see photos of both looking online, but i dont know which marking scheme(s) were used on real parts
<azonenberg> so that doesnt narrow it down much
<Wanda[cis]> I can give you some photos
<Wanda[cis]> ML505
<azonenberg> So it is ink printed
<azonenberg> that matches what the aliexpress seller has
<azonenberg> i'll bite. $30 for microscope food
<azonenberg> it'll make an interesting blog post or something even if it's fake
<Wanda[cis]> ML510 (holy shit this board is dusty)
<azonenberg> lol yeah
<Wanda[cis]> SP605
<azonenberg> ok so based on those
<azonenberg> i'm most confident in the ink printed parts
<azonenberg> the ones with laser engraving are more sus
<azonenberg> the ones with only the part number, date code, and "JAPAN" almost 100% fakes
<azonenberg> i.e. there's no way xilinx would make one without a logo
<Wanda[cis]> and the oldest I have, a XUP Virtex-II Pro board
<Wanda[cis]> before RoHS apparently
<azonenberg> ah yeah a non-G part
melnary has quit [Remote host closed the connection]
melnary has joined #prjcombine
Micko has quit [Ping timeout: 256 seconds]
_whitelogger_ has quit [Remote host closed the connection]
Micko has joined #prjcombine
Micko has quit [Quit: Client closed]
<Wanda[cis]> okay I am DONE with this cursed refactor
<Wanda[cis]> 163 files changed, 21440 insertions(+), 17810 deletions(-)
<Wanda[cis]> the core interconnect model has been disentangled from ISE tile/wire/pip/bel naming model, so I can now add support for toolchains that aren't ISE (or Vivado, which has basically the same naming model)