adgjlsfhk1

3 ms·

31.
by adgjlsfhk1 1mo ago
Julia has pretty good design heritage for how to deal with this sort of thing. you build the right abstractions and everything works (the main key is making sure the compiler elides bounds checks)
32.
by adgjlsfhk1 1mo ago
The really big problem is that $$ don't buy freedom, they buy freedom for your lawyers to talk with ARM's lawyers about what features you are allowed to build.
33.
by adgjlsfhk1 1mo ago
it's more efficient than paying 1000x for fhe
34.
by adgjlsfhk1 2mo ago
Julia
35.
Quasipolynomial Cryptanalysis of the McEliece Cryptosystem(eprint.iacr.org)
1 points by adgjlsfhk1 2mo ago|0 comments
36.
by adgjlsfhk1 2mo ago
This is one of the nice things about JIT languages. you defer the compiler time decisions to runtime and this get to choose based on what the user has
37.
by adgjlsfhk1 2mo ago
right, but that's a reticle size chip. to put something in a phone it has to be ~10-30x smaller
38.
by adgjlsfhk1 2mo ago
The size of model we're talking about running doesn't need much if any dram.
39.
by adgjlsfhk1 2mo ago
I don't think this works out from a cost/silicon perspective. Small models already run pretty well in software (since the weights fit in cache) and big models require silicon area proportional to the size of weights. On a mobile d
40.
by adgjlsfhk1 2mo ago
Heat is really useful if the energy form you want is heat. If you want electricity or motion it's not the best.
41.
by adgjlsfhk1 2mo ago
the GPU competition is using 16 gpus, so the actual comparison is that the CPU has <1/10th the bandwidth
42.
by adgjlsfhk1 2mo ago
> Federal prosecutor success rate is over > 90% Prosecutors pick their cases. Defense doesn't. The cases that aren't 90%+ sure aren't charged.
43.
by adgjlsfhk1 2mo ago
I think HDR wasn't what changed. It's resolution. At 720P and below, you don't notice dither because pixelation is the more obvious problem. Once you're at 1080p (and it's more obvious at 1440p/4k), you start h
44.
by adgjlsfhk1 2mo ago
The counterpoint is that you should just keep your images at 10 bit minimum or 12+ bit until the OS/monitor can dither for you. Lossy compression will do much better with a 10/12 bit image than a dithered 8 bit one.
45.
by adgjlsfhk1 2mo ago
red/green/blue
46.
by adgjlsfhk1 2mo ago
Max effort is the way to give the highest perf, but not highest perf/$. Having claude (or other models) use a lower effort can often be 80% as smart but get to the results 10x faster for the problems where it works.
47.
by adgjlsfhk1 2mo ago
Testing at max effort likely doesn't produce optimal results.
48.
by adgjlsfhk1 2mo ago
right, the thread you responded to is asking why there hasn't been a parallel criminal suit
49.
by adgjlsfhk1 2mo ago
that definition only converges for Re(z)>1. for Re(z)<=1, you need alternate formulas
50.
by adgjlsfhk1 2mo ago
Julia has roughly 2 different IRs that live pre LLVM (one for type inference, one for Julia level optimization). This is a proposed replacement to both of these. The reason not to go MLIR is that writing Julia is a lot nicer than writing C+
51.
by adgjlsfhk1 3mo ago
median of median does give you guarenteed n*log(n) but it doubles your memory reads per pass making it pretty poor. single random is almost guarenteed to take fewer passes (and median of 3-7 random values can make the number of extra passes
52.
by adgjlsfhk1 3mo ago
it's arguably more of a cpu bug than a computer bug. The problem is that predictable data determined whether a cmov or a branch is faster. cmov is only faster than if when the branch is unpredictable. Summer the compiler doesn't k
53.
by adgjlsfhk1 3mo ago
if you have decent (randomized) pivoting, you never hit the worst case or anything like it
54.
by adgjlsfhk1 3mo ago
the problem with oral societies isn't that they unable to transmit information (in the Shannon sense) it's that they're really bad at transmitting information through time while separating true facts about history from fables
55.
by adgjlsfhk1 3mo ago
Binface's speculation to the media is that the Monster Raving Loony party may split the vote with Farage instead as that is closer to where his loyalties lie.
56.
by adgjlsfhk1 3mo ago
> SpaceX started moving into their own constellations because lower prices open up new markets, and this lets them access those markets without having to lower launch prices the alternate take here is that even at current prices, falcon
57.
by adgjlsfhk1 3mo ago
it's not just that they have the same asymptomatic. once you do the radix conversion (e.g. base 10 has log(10)/log(2) times more symbols), any 2 definitions are only off from each other by a constant
58.
by adgjlsfhk1 3mo ago
none of that explains why the dense states (e.g. ma, ny, ca, tx) have this solved
59.
by adgjlsfhk1 3mo ago
if the proposal was ml-kem+ecc hybrid that removed sha2 and replaced it with sha3, no one would be objecting
60.
by adgjlsfhk1 3mo ago
if no one should implement it, why standardize it?

More ›