Viewing profile — etep

The edit distance form work to zork is 1. For this, and because of tab completions (perhaps tab completions post-date zork, but nonetheless) I find the work:zork story compelling. …

Surfer deserves to hit the front page also. Much better than gtk wave. Nice work Spade & Surfer!

PAX Markets (YC W25) | Founding SW or HW Engineer | Menlo Park, CA | Full-time https://pax.markets We're building a crypto exchange, on a chip. We sell ultra-low latency market acc…

I'm curious about the results under the section titled "Demonstration." The author claims that ground truth is that each goroutine utilizes 10% of the CPU time (so stipulated that …

Wasn't aware of it, will try it - thank you!

I have a theory that news occurs with a far lower frequency than people think, far lower than even when people try to account for their knowledge of the 24 hr. news cycle. In my th…

Thank you for the insight! Just for clarity, I'm using "autopilot" as a catch-all term that refers the automated systems that plausibly seem to have caused these crashes.

It's not nice that so many writers parrot that the solution is "more pilot training." Under the hypothesis that there needs to be a solution (one that I think is correct), then the…

https://xkcd.com/1138/

For chips, power scales with voltage squared. Is also true that P=IV (since both are true, these observations cannot be in contradiction). Apparently, for chips, the current must b…

I think it is linear for frequency and non-linear for voltage, i.e. P~f C V^2. But in many current CPUS, the feature that adjusts frequency also adjusts voltage. That's why I stipu…

It should be the case that if the same amount of work is done, then the energy used will be the same. If it takes less work to compile the web assembly, then less energy (holding a…

If Bezos gave all his money to everyone in the world we would all get maybe $13 or so. If all the billionaires did the same we would all get $1000 or so.

Hi stjepang, If the following criteria is met, then perhaps the branch mis-predict penalty is less of a problem: 1. you are sorting a large amount of data, much bigger than the CPU…

I think maybe we were talking past each other. Yes there is more than one reason. It's far easier to add capacity by adding sets, as opposed to ways. But they can't add sets in the…

Hi, it is the main reason L1 hasn't grown. By your reasoning, no cache should be able to grow, because then their latency would increase too much. But instead, all other CPU caches…

At a high level it's true that smaller is faster, but it's also true that those L1s could have grown by adding sets (not ways) and achieved the same latency. L2 has grown, but stay…

If it was easy to build high performance caches with high associativity, we would certainly see higher associativity. Ideally, you want a fully associative cache, but it's too expe…

I had wondered, why are the L1 caches not growing, while L2 and L3 capacities continue to grow: a significant limitation on L1 cache size is actually the fixed tradeoff between ass…

My point exactly, not the most lucrative. So what remains is the implicit assertion they are fleecing the big time guys. And somehow the big time guys remain in business... so what…

People like to say the market is a zero sum game, but I have always been suspicious of this truism. Its fair to say that participants can leave the market, but in practice that isn…

If the market is a zero sum game, then for every winning trade, there must be a losing trade. Obviously some teams must be winning consistently, which implies that some are losing …

I really like the "parallel" by default paradigm, or perhaps rather, I think it has great potential. When you consider any hardware design (Verilog, Chisel, etc.) the default is to…

In my hazy recollection turbotax once upon a time made it easy to see their own DAG, essentially an outline style index into their various interview questions. Now such navigation …