charleshn

11 ms·

31.
by charleshn 1y ago
> Spawning processes generally takes much less than 1 ms on Unix It depends on whether one uses clone, fork, posix_spawn etc. Fork can take a while depending on the size of the address space, number of VMAs etc.
32.
by charleshn 1y ago
Indeed, see https://semianalysis.com/2025/01/31/deepseek-debates/ for a well researched article about the actual cost.
33.
by charleshn 1y ago
See also Rendezvous hashing [0], which is simpler and more general. [0] https://en.m.wikipedia.org/wiki/Rendezvous_hashing
34.
by charleshn 1y ago
Because it's actually fairly slow. Among other things, the OSD was not designed with NVMe drives in mind - which is fair, given how old it is - so it's nowhere close to being able to handle modern NVMe IO throughput and IOPS. For
35.
by charleshn 2y ago
As someone with a phenotype more subject to skin cancer, I have been using an app called SkinVision for years. From my anecdotal experience it tends to flag the same moles as dermatologists do, and they have actual dermatologists review ima
36.
by charleshn 2y ago
Definitely, I'm playing fast and loose with "lightweight formal method here", thanks for making it clear. I was mentioning it in the same context as e.g. the Amazon paper on lightweight formal methods [0] where they use prope
37.
by charleshn 2y ago
On lightweight formal methods, a favorite of mine that's not widely known is trace verification using Linear Temporal Logic [0]. Basically you just need to log events - which you can even have for free in event-driven architectures etc
38.
Recent Linux Improvements That Impact TCP Throughput [pdf](conferences.computer.org)
2 points by charleshn 2y ago|0 comments
39.
by charleshn 2y ago
Is the Quantum X800 generally available now? Looks like it was announced in March. The SN5600 was released last year on the other end.
40.
by charleshn 2y ago
I don't believe the pricing is in the same ballpark. There's also other differences, such as port counts. AFAICT Spectrum switches at 400Gbps have up to 128 ports whereas equivalent Infiniband NDR Quantum only have 64 [0]. When b
41.
by charleshn 2y ago
Indeed, it seems that 80usec is just given as an upper bound based on the 1MB buffer at 100Gbps. It is definitely possible to go much lower than 80usec on Ethernet. But obviously it depends on the scale, utilisation etc. At the sizes of GPU
42.
by charleshn 2y ago
See also the classic "Are You Sure You Want to Use MMAP in Your Database Management System?" which mentions this common pitfall of mmap, and others, in the context of DBMS. [0] https://db.cs.cmu.edu/papers/202
43.
Nvidia's Blackwell Reworked – Shipment Delays and GB200A Reworked Platforms(semianalysis.com)
67 points by charleshn 2y ago|6 comments
44.
by charleshn 2y ago
This answer by Peter Cordes is much more complete: https://stackoverflow.com/a/38549736
45.
Facebook's Tectonic Filesystem: Efficiency from Exascale (2021) [pdf](usenix.org)
2 points by charleshn 3y ago|0 comments
46.
by charleshn 3y ago
It is the case, see the vm.dirty_expire_centisecs sysctl [0] at the VM layer, and the commit interval for e.g. ext4 [1] at the filesystem layer, and [2] for their interaction. [0] https://docs.kernel.org/admin-guide/sys
47.
by charleshn 3y ago
I've seen around 1-3% on non micro benchmarks, real applications. Aee also this benchmark from Phoronix [0]: Of the 100 tests carried out for this article, when taking the geometric mean of all these benchmarks it equated to about
48.
by charleshn 3y ago
It's not just the loss of an architectural register, it's also the added cost to the prologue/epilogue. Even on x86_64, it can make a difference, in particular for small functions, which might not be inlined for a variety of
49.
by charleshn 3y ago
> Why are BLAS written in assembly for each CPU architecture? On that specific subject of SIMD portability, embedded DSL like Highway developed by Google [0] can help. Hopefully it'll gain traction and usage. [0] https://
50.
Overcoming the Iotlb Wall for Multi-100-Gbps Linux-Based Networking(peerj.com)
4 points by charleshn 3y ago|0 comments
51.
by charleshn 3y ago
I really like magic-trace [0]. https://github.com/janestreet/magic-trace Not that the exact tracing relies on Intel PT - support for AMD was added recently but uses perf so suffers from the same sampling/skew issu
52.
by charleshn 3y ago
On that subject, it's worth knowing that with the default BIOS settings the whole socket shows up as a single NUMA node, which means that most workloads don't benefit from automatic NUMA balancing or manual NUMA-aware placement wi
53.
by charleshn 3y ago
strace's -Z flag is quite useful to filter only syscalls that failed. In a similar way, -k is useful to get tracebacks.
54.
by charleshn 3y ago
The original report [0] shows the corruption due to NUL bytes at the end of the file (see the hexdump). This comment [1] from Ted Ts'o details the exact chain of events leading to it. [0] https://issuetracker.google.com&#x2F
55.
by charleshn 3y ago
It's a shame more people aren't familiar with arithmetic intensity and the roofline model [1], which give a simple mental model for performance profiles on both CPU and GPU. I always find this graph [2] covering decades of hardwar
56.
by charleshn 3y ago
It's likely because of delayed allocations (delalloc): https://issuetracker.google.com/issues/172227346#comment6 because the only guarantee which data=ordered provides is the security guarantee that stale data wo