lukefleed

29 ms·

1.
Who Owns the Memory? Part 2: Who Calls Free?(lukefleed.xyz)
10 points by lukefleed 9mo ago|0 comments
2.
Who Owns the Memory? Part 1: What Is an Object?(lukefleed.xyz)
12 points by lukefleed 9mo ago|1 comments
3.
by lukefleed 10mo ago
That's an interesting question. I don't have too much experience, but here's my two cents. For matrix function approximations, loss of orthogonality matters less than for eigenvalue computations. The three-term recurrence mai
4.
by lukefleed 10mo ago
Thanks! I used perf to look at cache miss rates and memory bandwidth during runs. The measurements showed the pattern I expected, but I didn't do a rigorous profiling study (different cache sizes, controlled benchmarks across architec
5.
by lukefleed 10mo ago
Thanks!! I'm currently working on expanding that work. I will post something for sure when it's done.
6.
by lukefleed 10mo ago
Sorry for the late answer. The blog post is a simplification of the actual work; you can check out the full report here [1], where I also reference the literature about this algorithm. On the cache effects: I haven't seen this "en
7.
by lukefleed 10mo ago
Hi there, thanks! I started doing this for a university exam and got carried away a bit. Regarding Rust for numerical linear algebra, I kinda agree with you. I think that theoretically, its a great language for writing low-level "high-
8.
Cache-friendly, low-memory Lanczos algorithm in Rust(lukefleed.xyz)
141 points by lukefleed 10mo ago|22 comments
9.
by lukefleed 1y ago
For many applications, casting to u16 and wasting 6 bits is perfectly fine. The "trouble" is only worth it when you're operating at a scale where those wasted bits add up to gigabytes. This is common in fields like bioinforma
10.
by lukefleed 1y ago
It's not about poorer instructions; a get_unchecked on a Vec<u8> is just a single memory access, which is as good as it gets. The difference is likely down to cache locality effects created by the benchmark loop itself. The bench
11.
by lukefleed 1y ago
The other replies nailed it, but I'll add my two cents. Vec<u8> may be the right call most of the time for most use cases. This library, however, is for when even 8 bits compared to 4 is too much. Another example, if all your val
12.
by lukefleed 1y ago
That's a good catch!! Thank you, you are right. I (incorrectly) assumed that a single u64 could capture the entire bit_width-value read starting from byte_pos. However, as you said, this assumption breaks for some large bit widths. I a
13.
by lukefleed 1y ago
BEXTR basically does the same thing, yes. I'm sticking with the portable shift-and-mask, though. My bet is that LLVM is smart enough to see that pattern and emit a BEXTR on its own when the target supports BMI1. Using the intrinsic dir
14.
Engineering a fixed-width bit-packed integer vector in Rust(lukefleed.xyz)
89 points by lukefleed 1y ago|21 comments