asciilifeform: 100k being a very conservative figure.
asciilifeform: titanic is at ~4km, afaik - and, still there...
asciilifeform: in particular, while not knowing where.
asciilifeform: thestringpuller: go fetch treasure from 5km+ down.
asciilifeform: thestringpuller: hardest part, aside from constructing the hypothetical apparatus, is finding the site again without recourse to satellite.
asciilifeform: i suppose that's more than some people can say.
asciilifeform: i've concinced perhaps... 3, 4, 5 people who actually understood what the hell the entire picture is about.
asciilifeform: because it is quite impossible to carry out the proof while day job.
asciilifeform: gabriel_laddel: the crux of the matter, though, is that the only possible proof is experimental. and so what i really should've worked on, instead of essays, would've been some scam like davis has
asciilifeform: decimation: someone did build a clockless 'arm' a few years ago, an israeli firm quietly sells it for military clients, it's a footnote mostly.
asciilifeform: i've been working on 'genuinely' reconfigurable cpu for some years now. but, understand, most of the reading i do - is material as old as i am.
asciilifeform: and when designed - won't be chinese.
asciilifeform: the kind of machine that absolutely requires an fpga - simply has not been designed yet.
asciilifeform: because it doesn't really matter what they cost
asciilifeform: mostly used in miniscule runs, prototypes - where american units generally suffice
asciilifeform: it has very little role to play in the kind of items chicoms make
asciilifeform: decimation: have to understand what fpga is used for
asciilifeform: decimation: bad fit for chinese business model
asciilifeform: shame fpga was invented too late for soviet version
asciilifeform: decimation: problem is lack of fpga with documented layout, so you aren't chained to a vendor's turdware.
asciilifeform: decimation: there is no shortage whatsoever of open 'verilog' code for 10000+ things
asciilifeform: e.g., you can lay out the equivalent of a '486' and get around that performance, but only if you have some idea of the actual physical topology of the circuit.
asciilifeform: thing you must appreciate, if never having worked with fpga, is that they just barely work.
asciilifeform: as in, by the time you're done, the chip is no longer commercially available in quantity.
asciilifeform: but problem is, such efforts are quickly obsoleted.
asciilifeform: and an even gnarlier talmud of combinations that will result in anything like usable performance
asciilifeform: basically, fpga works, but all existing high-density examples are sold by scammers of a kind - who absolutely could not abide the buyer using these items as the schoolbook promised - 'reconfigurable computer'
asciilifeform: (each new series of 'xilinx' chip has considerably altered internals, to make past reverser's efforts less useful)
asciilifeform: hence, some of the gnarliest, moving-target obfuscation in the known world.
asciilifeform: they have literally no way of enforcing their copyright other than obfuscation
asciilifeform: decimation: both 'xilinx' and 'altera' make the bulk of their profit by renting out the right to use 'cores' (as they call them) - libraries, to a normal person
asciilifeform: decimation: understand why the actual internals are secret.
asciilifeform: the serious fpga market (vs. 'lattice' & friends, who largely get used for minor glue logic) is a duopoly.
asciilifeform: and there is literally no alternative (other than 'altera' - who have a considerably fouler turd)
