addaon

6 ms·

91.
by addaon 4mo ago
Since when are we doing 32-layer planar transistor logic on a single chip? Even ignore the use of FETs for eDRAM… I didn’t realize we had decent logic density possible on BEOL.
92.
by addaon 4mo ago
> Top floor existed at all because it was Boeing's entry for a heavy cargo plane competition Yes, but it turns out the hump is great for area ruling (aerodynamic drag reduction at transonic speeds), as observed by the 747-300's
93.
by addaon 4mo ago
Woulda given the AI credit if it had gone with "The merchants of Assur called it Nergal."
94.
by addaon 4mo ago
Meanwell is the standard answer for this sort of thing; something in the SD-500 family or sized/optioned as you need. You'll have to do the connectors yourself; you may be able to find junction-post-to-barrel-plug leads of the rig
95.
by addaon 4mo ago
Why use the original power bricks, with the space claim and awful routing, instead of just going to a single dc/dc... either directly if no individual power control is needed, or to a relay block or switch block if automated / man
96.
by addaon 4mo ago
Me too!
97.
by addaon 4mo ago
> it's unlikely that you'll get people that are in overlapping networks Networks all overlap if you search deep enough, ask Kevin Bacon. For good-enough talent and starting with random nodes, yeah, you'll probably not over
98.
by addaon 4mo ago
Yes, making people an offer they can't refuse isn't scalable. It generally requires both deep pockets and the time investment to understand what would motivate them to leave a good job for one that they think can be great.
99.
by addaon 4mo ago
I am aware of several startups that started this way, and have been involved in some. The quality of the results depends on who you seed the graph with, of course; but I’ve seen it work well. As requested by the original poster, it doesn’t
100.
by addaon 4mo ago
Ask the best people you’ve ever worked with who the best people they’ve worked with are. Recurse. When names start repeating through different graph paths, make those people an offer they can’t refuse. Once they join, ask them to do the sam
101.
by addaon 4mo ago
> costs are only likely to decrease over time >> There was no argument that the price will forever keep going down It's very hard for me to read the first quote as anything other than a continuing decrease in expected value of
102.
by addaon 4mo ago
In order of safest to least safe: commercial jets, automobiles, general aviation aircraft, motorcycles, messing around with Jim.
103.
by addaon 4mo ago
Everyone is pointing out examples around image reproduction, which are valid and interesting… but the case that comes up in nature is violet (beyond blue in spectrum) vs purple (mix of red and blue) pigments.
104.
by addaon 4mo ago
But that's my point. If you're basing your society shape around adopting a technology based on it continually decreasing in price, but you only get a few decades of that behavior before saturation and then you're at the mercy
105.
by addaon 4mo ago
> it’s costs are only likely to decrease over time What is this based on? We're well past a 50-ish year deflationary period in the cost of major appliances (refrigerators, washing machines, etc). We're pretty clearly at or near
106.
by addaon 4mo ago
Okay, that video is great. Product questions that I couldn't find an answer to. From https://www.crowdsupply.com/modos-tech/modos-flow , I see "On the go, you can power Flow at up to 40 Hz with a single USB Ty
107.
by addaon 4mo ago
They’re not dead until they’re warm and dead.
108.
by addaon 4mo ago
Fixed-point CORDIC is not hard… but a trivial implementation will be slower than a hardware float implementation, true. Several of the micros I work on have a CORDIC accelerator on a (usually only one) low-latency core designed for BLDC con
109.
by addaon 4mo ago
I've almost never seen a real complexity difference between floating point and fixed point math in a language that supports fixed point either natively or through operator overloading. Yeah, scaling sometimes changes across an operatio
110.
by addaon 4mo ago
Depends how you measure. The choice of floating point lat/long is very favorable for people in the area of Null Island, and while the population there nearly zero, the population density is either NaN or infinite, depending on how yo
111.
by addaon 4mo ago
> Those are the specific inventions. In what way is electricity an invention? Electricity is a physical phenomenon. Various machines for doing work with electrical energy, storing electrical energy, converting other types of energy to el
112.
by addaon 4mo ago
> This comes up all of the time with stuff like lat/lng values. People are convinced you have to use doubles because of the inaccuracy of floats. Using any floating point representation for lat/long makes no sense at all to me.
113.
by addaon 4mo ago
This is insane in so many different ways. 1) Why would you use floating point instead of fixed point or other fractional representation (uh, percent?) for this sort of thing? 2) Why would you design an interface that's specifically set
114.
by addaon 4mo ago
> due to a dominant recessive gene for pigment I assume this should be "a double recessive gene"? Although looks like a lot of blueberry varieties are not diploid, so...
115.
by addaon 4mo ago
Today, I write the code. It’s trivial and takes a lot less time than writing the spec, and since I’m using conventional tooling for WCET and stack sizing it’s nice to get those right up front. The LLMs sometimes tweak the code slightly for
116.
by addaon 4mo ago
With the proof checker.
117.
by addaon 4mo ago
That's up to the harness. Right now, my harness doesn't have any tools that would allow the agent to change contracts. Also, agents working on a Rocq proof don't have write access to the C+ACSL code; they can review it, and p
118.
by addaon 4mo ago
I've been playing with related ideas recently, and can talk about them at length... but one thing that's surprised me is just how effective frontier models (ChatGPT-5.5 in particular) are at completing certain manual proofs in the
119.
by addaon 4mo ago
"If something requires attention, it’s by definition relevant" Not really. Consider an assembly language for a processor with a very orthogonal register set. The number of registers used by a block of code is relevant, but the ide
120.
by addaon 4mo ago
> an app that helps you detect if a video (tt/reels) is lying about those "do this and you'll make 10k a month" There’s a Unix CLI tool that implements an accurate version of this… check out /bin/yes.

More ›