daniel-cussen

It doesn't matter that much they're not super accurate exact, they do in fact count ticks n tocks j dandy. I wonder if they are counted equally, interesting question if i do say so myself about my own question, i know rising edge is steady w the next rising edge, n falling edge w falling edge, but yeah...no i think that's a specification figure of merit, that they be balanced w respect to each other. N Intel® despite it's many failures, long ago forecast n long overdue due to the sheer difficulty of their business model n how long they kept it going--they were saying it was going to end soon in the late seventies already--n you know what, i'm fine w that. They don't make old chips rot like other software companies i could mention, bits rot but glass is timeless. Which brings me back to the point, in my analysis the problem is not the clocks being inaccurate but rather the jitter, which means a single run will not suffice in describing a repeatable exact clock time taken for eg an inner loop, which is what is worth bothering with. The minimum jitter attainable currently is 1 cycle, n then i guess you run the same code repeatedly n take the minimum with more repeatability as a consequence of that low jitter.

In the early nineties it was not so, you'd get the same number of clock cycles again n again n again.

N then it gets tricky because cycle counts are thresholds set within which, if voltages n frequency are proper, the operation will complete deterministically w a soft error rate no greater than the system as a whole, about one per 30 years of hammering on the chip at sea level. Which is not enough for my tastes, n the jitter is a fucking mess.

I much prefer the GA144, least jitter of any platform bar none, sensible because it is fully asynchronous logic, no clock anywhere in sight until you connect it to an oscillator, n even then the vibration doesn't rock the system like that of a grandfather clock synchs w another grandfather clock with whose pendulum's swing the former clock's pendulum swing is aligned. GA144 it's pretty easy to tell average case complexity of a function down to the tens of picoseconds, at which point you have to check that there's no open window letting a draft in. In fact the time trial will tell you such a draft is coming into the house, it happened to me, while not from a parallel universe in spacial dimensions it is by all means from another universe in the time dimension.

“will unnecessarily cost taxpayers billions of dollars.”

This is undisputably true. H&R Block n Intuit are taxpayers, n free-file will cost them billions of dollars.

Being real, online taxes through the SII work great in Chile.

Passive voice, better title is Kevin Mitnick died, or passed away, sugar coat it like he never did if you want, we all know what it means. I suddenly feel bad for him, though. He deserved the black ribbon.

Erdosian means having the genetic trait Paul Erdós had, which was the uncommonly good reaction to similar stimulants to those he took, without which he could not do any math research at all. He was steroidal, i'm steroidal too in my research, it's like bodybuilding in that steroids are part of the game. As long as they do not damage your health, it's fine. I literally get time dilation from the stimulants i take, which i do under prescription of course, but at the same time i know it's not fair to play sports for instance, because they're performance enhancing drugs (n generally outright forbidden). But sports are zero-sum, research is positive-sum, so that's what i do instead.

In particular the critical trait is amplified creativity to the point i share Erdos view that "Before, when I looked at a piece of blank paper my mind was filled with ideas. Now all I see is a blank piece of paper.". Same exact symptoms.

In addition, my being steroidal is justified by the fact that i got brain surgery, like the radio transcranial magnetic crap, against my will n without my knowledge, electroshock too, the works. So perhaps it is understandable in my case.

I wager in some part of your analysis you're not dead on the money but in broad strokes this all rings true. The political acumen necessary to survive as a programmer is intense, it is one advantage black programmers i've met had a better handle on than white programmers who are more often autism-spectrum n therefore terrible at office politics, cept when being oblivious is advantageous which it can be. It was ironic in the scheme of racial stereotypes present in American culture, that both black coders were employed while the white coders were not, this was in a hacker house, but it was simply because they were better coders. No issues there. We got along great, but in terms of office politics, n without being aggressors either, j very good instincts. It isn't talked about in plain terms enough even on HN, well i guess it is discussed a lot, but never like "alright, here's the strat if you're facing X" like you talk about in the face of other commentors's naysaying.

Tracking helped me in particular by making my homeroom (the British school equivalent, j called a class) more civilized, n fair to students w less academic talent than me. Given the tracking in English n at one point, briefly, Mathematics, it meant the other boys had a shot at getn the best score in the class or at all enjoying leniency in grading, which my presence as a future Erdosian mathematician made impossible, n their parents would deny them privileges n beat them sometimes if they weren't excellent students, so they'd bully me for monopolizing the academics completely. The girls did not, because i believe there was not an expectation they be the best student n plus i was very good-looking to them, though not to myself. Future model too. But the bullying was the main challenge, no matter what they threw at me the academics were nothing in difficulty compared to the bullying. So for that reason alone tracking helped me out, n the homeroom out too.

It'll get USA screwed on the World stage too. For me getn tracked in math classes pretty much the whole time, n to the point we had a class w four people in it, smallest class-size in the school already w the smallest class sizes in the country, all four of us, on that table getn the closest thing to tutoring the Chilean education system could give us in its most expensive school International School Nido de Aguilas, n then it's like...without tracking i would have hated math, got bullied much more for always cracking the curve, all that shit.

Elon Musk already complained about the basic difference between the CCP politicians n the American senators n congressmen is the former have a distantly n vertically superior knowledge of math n science than their American counterparts, being engineers w a v rigorous sink-or-swim training, than the American politicians who are generally lawyers n only one of which has a PhD. There's as many PhDs as comedians in Congress for fuck's sake, like not hating the comedian i appreciate there being one, that's cool, but for there only to be one PhD, what the fuck?! Some people need to have a PhD, it matters sometimes it is generally a good way to get technical expertise spesh w subjects w high stakes that require planning for executing correctly, which is why i went into algorithms instead, my laptop had no choice but to obey n compile what i wrote, n run it, automaton after all, that computer was my only audience but an audience nonetheless. There it's fine, n that's why coders don't need degrees for the most part, low cost of failure, not like a civil engineer which i would not allow building a bridge j like that.

Let's look at your profile, it says,

[insert interesting falsehoods here]

which is a confession. You insert interesting falsehoods here, on Hacker News.

You yourself confessed, in your own words, you're full of shit.

Yeah because 1% going to trial sounded a little too high, most liars go for 2%, more believable. Stanford tried 2% of the accused, 1% is pushing it. 2% annual inflation. 2% of rape accusations are false. Say 1% or 0.75% and people call bullshit.

N again, you confessed that you "[insert interesting falsehoods here]" [sic brackets]. You wrote that w brackets in the original, i don't escape brackets as this isn't coding, normally brackets when i quote are when not absolutely sure word-for-word quote, to protect the integrity of my testimony for my eventual trial. Which is getn closer n closer, any day now, the longer you wait the less time left until your day in court, n it's been more than 14 years waiting for Stanford to try me.

Dude Common Lisp 100%!

Don't fret if you think you're mediocre. I myself have been trying to get through On Lisp since 2008. Then, after that, Let over Lambda. Not better, but v strong book, i can tell the guy while not the best at humblebragging has so much cool stuff there.

ANSI Common Lisp n that's plenty to not be mediocre! I did get through the whole thing in 2009, what a great book!

Find yourself.

I started out imprentation in 2008 as a business co-founder, now i'm the solo founder of fgemm SpA, j published the paper about it n have the provisional patent for the algorithm's implementation on a machine.

You can do it.

Finance is numbers. Coding is numbers.

You can do it.

So i'm in favor of police going back to the basics, good old-fashioned detective work. The forensics are not what they're cracked up to be, there's very little verifiability n verification, too easy for the forensics lab to lie through their teeth or contaminate the scene of the crime. Like when the Duke Lacross players had to hire a very expensive lawyer to do the messy work of proving evidence actually traced back to three other men that same day, that was noted as very difficult forensic work because OF COURSE the DA didn't want verification. Local DA Dan Nifong wanted to tamper w it, anything to pull off the lynching, mostly out of sadism n participating in the culture. Three equals three, easy to see it was three men's evidence left behind in Candice Mangum's evidence place, figured it would be a slam dunk requiring only a little bit of finesse. This lawyer got commended in front of all America by the legal profession, n proved Candice Mangum was pretty much correct it would be a slam dunk. It was only when these young men scraped together huge money for the hardcore IDEALISTIC civil rights lawyer (unicorn basically) to counter the momentum of the shakedown w perfect lawyering against Nifong's zero shame of screwing up on purpose w all the evidence--an uphill battle all the way...in 2006 iirc, like now in 2023 again.

Wow.

I was going to say the same as donkeybeer, agree it is surely an APL variant.

True. Too late to edit. Yeah that is confusing, guess I gotta write differently when discussing that. I just shave characters for character counts, which are often a problem particularly on Twitter. It's not because I don't like typing out the whole word, I generally refrain from that sort of abbreviation. It is also stylistically unique--I use a unique style for the same reason as, and vindicating, Auguste Rodin who faced problems due to his statues being too literal.

Yeah n how much do you get paid, n for what? You get paid to take on professorial duties, TA'ing, lab assistant, that sort of thing usually. Pretty rough in many ways.

Thanks, help finding typos is appreciated! It'll make things easier for when this gets published in a peer-reviewed journal, which is in the works.

Fgemm SpA, coming soon.

No comment.

Uh, you clearly didn't get the proof. N-bit multiplications performed with a single addition and a single move. Proof involves showing superiority over the number of additions in Russian Peasants.

Not that tiny, n not such little energy, if Intel AVX-512 requires throttling the chip down to 60% of full speed (n similar slowdowns are necessary on Xeon Phi, 1.4 GHz to 1.2 GHz for full) SIMD when using "heavy" operations like multiplication (and also count-leading-zeroes aka CLZ) so it's a lot of energy for sure.

Unfortunately i cannot answer questions as unlike i chip i must now sleep. It's 2:42 AM here.

First thing in the morning i'll be on it.

At the time, i needed academic affiliation, meaning be in college or more likely have a professor vouch for me. What i ended up doing was return to Stanford undergrad n take classes related to algorithms, show my algorithm portfolio in office hours, then get referred to other profs, one of them being Jeffrey Ullman, in 2019. N then after emails back n forth we met in person in the Gates building, it went from there.

Very happy to have met Professor Jeffrey Ullman.

Yeh, it bears many resemblances to Babbage's difference engine.

Sorting is very fast since i designed a sorting algorithm tailor-made for this problem, which is about as fast as your approach. Difference being the numbers are 32-bit, so iterating through the entire array of all possible 24-bit mantissas (i know they're 23 bit, but the implicit initial 1 of normal (vs denormal) values is explicit here) would be way too slow. Otherwise you'd be right, 8-bit values you can just use counting-sort, no problem. Or 14 bits, same deal. Now also notice there's a difference between the matrix size and the mantissa size, we care about the mantissa size, so it's 24 bits.

One addition and one move replacing one multiplication? Absolutely makes things much cheaper.

Are you saying that GPUs invest area on floating-point multipliers because FLOPS are an important marketing metric?

Yes. That is precisely what i'm saying. If i'm mistaken in saying that that's one thing, but as far as it being what i'm saying, it very much is. It's been an important guiding principle for some time now in the project, that recent chips--including FPGA's--tend to have hard IP for floating point multiplication.

Now spending a lot of chip area on getting more FLOPS is not necessarily a bad decision if there is no alternative for achieving fast matrix multiplication. Almost any method is sensible if there was no better alternative available when the decision to use that method was made. In addition, fgemm only really makes sense when matrices contain over 1000 elements per row or column, not sure how much more than 1000 per vector but more than that. Small and in particular small and dense matrices are still best multiplied exactly the way GPUs multiply them, with many floating-point multiplier circuits in parallel. It's not stupid in the least.

Yeah so NVidia n Google have the same business model i'm going for, Google having TPUs in its datacenters that do work that cannot be reverse engineered. Google does not sell TPUs. You can use them by sending Google the work, and you'll benefit from much lower cost and faster speed. NVidia has a similar offering, just not as well-known. That's the correct business model in my analysis, and what fgemm will sell. Sell the work.

It's brutally hard. I had an easier time buying skylinesort.com n posting the skylinesort algorithm there, than publishing through professors n academia. Typically not feasible for undergraduates, least of all anybody not paying tuition. Same way professors are expected to have an undergraduate degree at the very least (4 profs at Stanford have just an undergraduate degree), a Master's degree (a handful have that and no more), but typically a PhD is required (literally all the other professors have PhD's). Is required. Who requires it? Who says, "I require a PhD."? Is expected? Who expects it? Who says, "I expect a PhD."? Passive voice is typical in academia. Very rare to get around the gatekeeping, frankly. I couldn't publish on arxiv for years because of lack of academic affiliation alone.

Took years to get to this point in terms of the effort I dedicate to getting recognition for my work.

Yeah i address that, that's a gotcha and if the numbers are exponentially distributed the algorithm does not work. It is not universal. It depends in part on the exponent for which the numbers are exponentially distributed, n the other optimizations you use. This is the purpose of ongoing experiments. The Fibonacci series is an interesting case, since you get rid of the two largest numbers in each pass.

Yeah hey the paper will not be that painful to read if you can already perform the reduction steps. I'll answer further questions.

Yeah so for floating point an exponential distribution is bounded in how many elements it can contain for a given exponent, so it works out quite nicely. It does not work on bignums.

Nice to see someone use lower-case i like i do!

I think you're right in this case. Alright I'll edit if I still can. I don't use any variables anyway in my post.

EDIT: alright I fixed all the single letter abbreviations.

I don't know why trade secret applies, but i remember reading it does. Perhaps in the rationale, or the preimage. It doesn't make all that much sense, come to think. I think Intel tried it? Intel for sure used copyright to protect chips. Hey thanks, i did not know about 17 U.S.C. §§ 901-914.