Remnant44

6 ms·

1.
by Remnant44 23d ago
I'm one of the weirdos that still holds on to mercurial as I genuinely prefer it's methods and workflow to git most of the time. How the author describes handling work trees is exactly how they work in hg, and I find it enormously
2.
by Remnant44 1mo ago
This was an unexpected find on HN but I found it very moving. Good writing lets you do something that is almost impossible otherwise; to get a glimpse inside someone else's mind and experience, the thing that is forever hidden from us
3.
by Remnant44 1mo ago
I haven't had a chance to dig into the repo at all, so that's excellent, thank you!
4.
by Remnant44 1mo ago
Looks very promising - I've been looking for a good delaunay library that supports constrained delaunay. It looks like their performance benchmark is including multithreading, which although a useful feature, makes performance comparis
5.
by Remnant44 3mo ago
Do you have any sources for this? I have also heard of rabbit starvation many times over the years, and it has always been in the context of too taking in too little fat -- essential fatty acids -- as well, due to the extreme leanness of th
6.
by Remnant44 3mo ago
I agree with you, but also: outside of anything else, amdahls law means that as the parallel performance grows, we become _more_ limited by the inherently serial code, and thus single core performance, not less. Given that single core perfo
7.
by Remnant44 4mo ago
It's a little dramatic to say avx512 is dead versus 10 - rather, I would say that avx10 finalizes a universally available set of avx512 extensions. For AVX 10.1, there's essentially, no difference after Intel backed out of reducin
8.
by Remnant44 7mo ago
There's a whole lot of us out there. I don't know if there's still a future in the thing that I love, which is where all the malaise comes from.
9.
by Remnant44 8mo ago
I love me some isa extension, I'd love to know what these are intended and useful for for though. 1 bit inference? I hear they could be useful in crypto as well, but that's out of my field.
10.
by Remnant44 8mo ago
While you may be correct in the sense that, in a public acquisition statement, people should be inferring enormous context and not taking anything said at face value. It's simultaneously true that this is the farthest thing from effect
11.
by Remnant44 8mo ago
Fort what it's worth, I had the exact same experience you did when I started writing SIMD code explicitly with intrinsics. I avoided it for a long time because, well, it was so damn ugly and verbose to do simple things. However, in act
12.
by Remnant44 8mo ago
AVX doesn't require alignment of any memory operands, with the exception of the specific load aligned instruction. So you/the compiler are free to use the reg,mem form interchangibly with unaligned data. The penalty on modern mach
13.
by Remnant44 8mo ago
There are many situations where your data is essentially _majority_ unaligned. Considerable effort by the hardware guys has gone into making that situation work well. A great example would be a convolution-kernel style code - with AVX512 yo
14.
by Remnant44 8mo ago
which honestly, shouldn't be neccessary today with avx512. There's essentially no reason to prefer the aligned load/store commands over the unaligned ones - if the actual pointer is unaligned it will function correctly at hal
15.
by Remnant44 11mo ago
In practical use for simd, various min/max operations. On Intel at least, they propagate nan or not based on operand order
16.
by Remnant44 11mo ago
I've run into this as well. Problem is that linear RGB is most definitely not a perceptually uniform space, so blending in it frequently does something different than you want. Use linear for physically based light and mixing, but if y
17.
by Remnant44 11mo ago
Essentially ever other use case for a computer. Whether you're playing games, or editing videos, or doing 3D work, or trying to digest the latest bloated react mess on some website.. ;)
18.
by Remnant44 11mo ago
I've had just the smallest touch of this caring for my elderly parents, and you have my deep empathy. It's exhausting and really really hard.
19.
by Remnant44 11mo ago
totally - especially given how bandwidth constrained CPUs still are, going wider than 512 doesn't make much sense. 512 itself was a stretch for quite a long time (and all the negative press on the original implementations was a consequ
20.
by Remnant44 11mo ago
Sure.. in detail and abstracted slightly, the byte table problem: Maybe you're remapping RGB values [0..255] with a tone curve in graphics, or doing a mapping lookup of IDs to indexes in a set, or a permutation table, or .. well, there
21.
by Remnant44 11mo ago
Yes and no. I think neon is undersized for today at 128bit registers -- if you're working with doubles for example, that's only two values per register, which is pretty anemic. Things like shuffles and other tricky bitops benefit
22.
by Remnant44 11mo ago
I'm just happy that finally, with the popularity of zen4 and 5 chips, AVX512 is around ~20% of the running hardware in the steam hardware survey. It's going to be a long while before it gets to a majority - Intel still isn't
23.
by Remnant44 1y ago
I think the most interesting thing here is the near-lack of NUMA effects on memory access, giving fairly easy to achieve high bandwidth memory. Combined with the double-width fabric links, it's an interesting part because it gives a gl
24.
by Remnant44 1y ago
For sure.. for what it's worth though, I have run across several references to arm also implementing uop caches as a power optimization versus just running the decoders, so I'm inclined to say that whatever it's cost it pays
25.
by Remnant44 1y ago
I was going to mention this as well. Source: chipsandcheese.com memory latency graphs
26.
by Remnant44 1y ago
ARM instructions are fixed size, while x86 are variable. This makes a wide decoder fairly trivial for ARM, while it is complex and difficult for x86. However, this doesn't really hold up as the cause for the difference. The Zen4/5
27.
by Remnant44 1y ago
Man, I miss Joel's blog. So much developer wisdom that is still relevant even if aged now.
28.
by Remnant44 1y ago
Link time optimization is definitely not new, but it is incredibly powerful - I have personally had situations where the failure to be able to inline functions from a static library without lto cut performance in half. It's easy to dis
29.
by Remnant44 1y ago
Just like talking to Grandpa!
30.
by Remnant44 1y ago
Great timing on this paper. I actually just started tackling a problem that is essentially exactly what is under discussion here (creating a coherent UV set for implicit geometry), so I'm very looking forward to reading it in depth. At

More ›