stabbles

20 ms·

211.
by stabbles 2y ago
Classification per pixel
212.
by stabbles 2y ago
The OCI Distribution Spec is not great, it does not read like a specification that was carefully designed. > According to the specification, a layer push must happen sequentially: even if you upload the layer in chunks, each chunk needs
213.
by stabbles 2y ago
The handbook for utilitarians prescribes that the life of one cow is worth the life of seven chickens
214.
by stabbles 2y ago
Somewhat confusing to see a matrix represented as an array of arrays, it will never be fast.
215.
by stabbles 2y ago
You're too fixated on the example. It's obviously used in cases where you cannot group like that: (-b ± sqrt(b² - 4ac)) / 2a = 2c / (-b ∓ sqrt(b² - 4ac))
216.
by stabbles 2y ago
I blogged about this when I stumbled up on it the first time: https://stoppels.ch/2022/11/30/io-is-no-longer-the-bottlenec... "Trick 2: Creating masks efficiently". It's pretty neat.
217.
by stabbles 2y ago
The original joke was to define an alias alias libtree='curl -LfSs https://raw.githubusercontent.com/haampie/libtree/master/libtree.c | ${CC:-cc} -o /tmp/libtree -x c - -std=c99 -D_FILE_OFF
218.
by stabbles 2y ago
It prints a DAG as a tree: every line is an edge, identical nodes are repeated. With `-vvv` it prints all possible paths as a tree.
219.
by stabbles 2y ago
See `man ld.so` for the search behavior. Note that there are inconsistencies between glibc and musl when it comes to rpath and runpath.
220.
by stabbles 2y ago
What is the relationship between glibc's iconv and libiconv? Both are GNU. They have the same API (but different symbols). Is glibc's iconv more or less a copy of libiconv code? Do they have the same bugs?
221.
by stabbles 2y ago
On very old glibc ldd used to set an env variable and run the executable. The assumption being that PT_INTERP is set to glibcs dynamic linker. The dynamic linker would dump resolved libraries and stop execution if the variable was set. My g
222.
by stabbles 2y ago
`sudo ldd /usr/sbin/sshd` why would you put `sudo ldd` on a slide.
223.
by stabbles 3y ago
More details: https://hudoc.echr.coe.int/app/conversion/pdf/?library=ECHR&...
224.
by stabbles 3y ago
Looks like remote work was getting popular anyhow, just that COVID accelerated it. Does anybody know what the bump in 2017 is?
225.
by stabbles 3y ago
How about wheels in the python ecosystem
226.
by stabbles 3y ago
Sounds similar to folks who can memorize many digits of pi in the sense that they can visualize a landscape of numbers and they just have to play that movie in their head, similar to how ordinary people can visualize commuting from home to
227.
by stabbles 3y ago
I don't get the point of the project. So they show that rounding errors are amplified because a problem is ill-conditioned. Yeah, of course. The higher precision computation they do is probably equally wrong then. Averaging / smoo
228.
by stabbles 3y ago
Computing the color of a pixel is computing the (averaged) integral of some iteration counting function. It looks like there are areas where this function is pretty chaotic -- isn't the problem very ill-conditioned? If so adding more p
229.
by stabbles 3y ago
Regarding "A final optimization", there's another way: `_mm256_cmpeq_epi32` produces `0xFFFFFFFF`, the article suggests shifting to produce a `1` and then add. Instead you can interpret `0xFFFFFFFF` as negative one, and subtr
230.
by stabbles 3y ago
Sure, you can reorder floating point operations and assume no aliasing. Now you can vectorize and avoid redundant memory access. But it's a problem of the language really. Alias analysis in Fortran is much easier than in C / C++.
231.
by stabbles 3y ago
I think they do a lot of good stuff, like LTO and PGO. But in benchmarks you sometimes see like a 4x speedup compared to ubuntu, which is obviously not due to superior compilers. For example: https://github.com/phoronix-test
232.
by stabbles 3y ago
It does not look like correctness is their main concern, they mostly care about performance of the vectorizer ;) Also note that before Intel oneAPI there was the now deprecated Intel Classic Compiler (icc), which also had bugs, since 2003.
233.
by stabbles 3y ago
Good luck with that, cause there are bugs, as the article also points out. Previous release miscompiled Python [1] Current release miscompiles bison [2] It's also the type of compiler that defaults to -ffast-math to look good in benchm
234.
by stabbles 3y ago
"why Schur vectors": stability + they're always computed when you compute eigenvectors. For non-symmetric matrices, the standard approach to compute an eigen decomposition is the QR algorithm, which is an iterative method tha
235.
by stabbles 3y ago
That is an ambitious goal in that issue. What's the reason to get rid of Fortran?
236.
by stabbles 3y ago
All Arnoldi methods basically repeatedly grow and shrink a search space until they contain good approximations to Schur vectors. Shrinking (=restart) is necessary to keep constant memory requirements, as well as performance. Restart is some
237.
by stabbles 3y ago
It's chilling to see the blog archive end at "January 2024"
238.
by stabbles 3y ago
For large and sparse matrices you can use the (restarted) Arnoldi method to compute a partial Schur decomposition AQ=QR where Q is tall and skinny and R has a few dominant eigenvalues on the diagonal (i.e. eigenvalues on the boundary of the
239.
by stabbles 3y ago
Yeah, exactly. Your matrix shows up when you do a 3-point discretization of -u''(x) = λu(x) with u(0) = u(1) = 0, at 5 equidistant (interior) grid points. That's the discrete, 1D version of the problem they're looking at
240.
by stabbles 3y ago
I think that's just how textbooks present it. A fact is stated but you're lacking intuition. If you had played around a bit with Laplacian matrices, like tri-diagonal matrices with stencil [-1, 2, -1], and found that its eigenvalu

More ›