Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
celrod
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
91.
▲
by
celrod
3y ago
Oops, I meant downgrade from 7950X to 3950X. It told me to buy a 3950X.
92.
▲
by
celrod
3y ago
I agree. "Workstation" is too general.
93.
▲
by
celrod
3y ago
Yes, it is. That is why this is a terrible recommendation.
94.
▲
by
celrod
3y ago
It recommended I pay the $150 extra to downgrade from the 3950X to the 7950X. Also, that I pay $1400 for a 3080 TI, but a substantially less powerful GPU would do just fine. I asked for a workstation, not a gaming rig.
95.
▲
by
celrod
3y ago
Every time I try it, it doesn't seem to work right. Because I don't feel like debugging my debugger, I just go back to gdb. Yet I only hear other people speaking of it positively, so it seems they don't have that experience.
96.
▲
by
celrod
3y ago
You could use C++ lambdas instead. They're basically structs with a call operator. See also the comparisons from the functional header https://en.cppreference.com/w/cpp/header/functional
97.
▲
by
celrod
3y ago
As a grad student, one of my more cynical professors said that to write a paper, you take a simple idea and then obfuscate it until it sounds complicated, and therefore seems impressive. He also said the hardest part of reading papers is un
98.
▲
by
celrod
3y ago
Taking the example: cmpb $115, %cl sete %dl addl %edx, %eax vs cmpb $115, %cl jne _run_switches_jmptgt1 mov $1, %dl _run_switches_jmptgt1: addl %edx, %eax The argument about why `
99.
▲
by
celrod
3y ago
I'd say the ways to increase performance are 1. decrease the number of instructions needed (has to be done in software, but also dependent on ISA, e.g. using AVX512 can help a lot here, so long as you don't end up executing more s
100.
▲
by
celrod
3y ago
FWIW, IIRC my Skylake-X CPU normally has around 2 instructions per clock when I run perf. It has a pipeline width of 4 uops internally. So it's falling far short of a typical basic block/clock cycle. I would also not expect the X4
101.
▲
by
celrod
3y ago
Given that the M1 and M2 perform similarly to AMD and Intel CPUs with far higher clock speeds, it seems most software can use wider dispatch. Note that dispatch doesn't mean vector width, which is harder for software to take advantage
102.
▲
by
celrod
3y ago
The ARM cortex x4 will have 10-wide dispatch: https://www.anandtech.com/show/18871/arm-unveils-armv92-mobi... It'll probably be roughly the same chip as one of their upcoming V chips. The specs there sound se
103.
▲
by
celrod
3y ago
By chordal editing, I just meant pressing combinations of keys at the same time. Lots of editors support this to some degree, but most (AFAIK) are either modal, or don't support keyboard-driven workflows. One of the other classic emacs
104.
▲
by
celrod
3y ago
Chordal editing isn't as popular as modal editing, but I do like it and think it's something uncommon in other editors, at least to the same extant as emacs. The thing I like about emacs is the integration and consistent experienc
105.
▲
by
celrod
3y ago
Given that gastic stapling/restrictive surgeries exist, perhaps this is related? I have a severely obese friend who got such a surgery. He is substantially healthier now, shedding over half his original body weight. EDIT: But see other
106.
▲
by
celrod
3y ago
The vqsort README says it is a non-issue on that generation CPU (Skylake-X) based on their benchmarks: https://github.com/google/highway/tree/master/hwy/contrib/so... At least on their low cloc
107.
▲
by
celrod
3y ago
Is this GPT-generated crap? It is wrong and useless, like GPT always is, but sounds intelligent to fool the ignorant. The compiler can inline functions, but it is essential for it to do so. It does not adjust heuristics based on being in a
108.
▲
by
celrod
3y ago
Yes, it is required for all functions to be inlined in Julia. Note that Julia 1.8 and newer supports callsite inlining, so you can, for example, do `@inline exp(x[i])` inside your `for` loop, and this will cause `exp` to get inlined there,
109.
▲
by
celrod
3y ago
One approach you could try now is creating StartUp packages: https://julialang.github.io/PrecompileTools.jl/stable/#Tutor...
110.
▲
by
celrod
3y ago
Probably that eigen uses expression templates to avoid the needles creation of temporaries.
111.
▲
by
celrod
3y ago
Stan is a DSL for statistical modeling named after Stanislaw Ulam, the inventor of Monte Carlo methods. Stan's primary method of fitting models is via (Markov Chain) Monte Carlo. https://en.wikipedia.org/wiki/Stani
112.
▲
by
celrod
3y ago
Clang does figure it out. Note that clang has a branch, not a loop. It jumps to returning 0, else it continues to do the same thing gcc does. I'd say it figured it out, even if the code isn't as pretty. I tried using `final` to ge
113.
▲
by
celrod
4y ago
I don't think it is likely to help C++, as templates and `auto` arguments (in C++20) were already specializing. These probably cover most of the cases you'd encounter there. I'm less familiar with Rust, but I'd have assu
114.
▲
by
celrod
4y ago
There was a fun example in the Julia compiler around a year ago. Part of the compiler was O(N^2) in `let` block nesting depth. That is let x = foo(), y = y, z = 2y ... end would be a depth of 3. It didn't seem like that sh
115.
▲
by
celrod
4y ago
It checks in debug mode, but these checks are turned off in release mode by default. They can be turned on in release mode, but C/C++ could add `-ftrapv` too[0]... You also have access to `checked_mul` and friends, so manually writing
116.
▲
by
celrod
4y ago
User time is the amount of CPU time spent in user mode. It is aggregated across threads. If you have 8 threads running at 100% in user mode for 1 second, that gives you 8s of user time. Total CPU time in user mode will normally increase whe
117.
▲
by
celrod
4y ago
I haven't found it to be that big an issue. Three cases that annoy me: [1] Slow commands like `straight-pull-all`. I may change how I manage packages in the future, since this is the command that really comes to mind. Because it'l
118.
▲
by
celrod
4y ago
Thanks, I'll give eshell-vterm a try.
119.
▲
by
celrod
4y ago
I second the suggestion of getting ergonomic keyboards with thumb clusters. I also have the Kinesis advantage 2, and am also a fan of the convex keys. I also ordered a Kinesis 360 several months ago, but have not received it yet.
120.
▲
by
celrod
4y ago
I haven't tried eat (Emulate A Terminal) yet, but I've been meaning to: https://codeberg.org/akib/emacs-eat Eat is an terminal emulator supporting eshell, written in elisp. It also supports full screen progra
More ›