thecompilr

8 ms·

31.
by thecompilr 4y ago
That OP just pointed how the provided example would actually work in an OOP project, not their fault to have to work with an example that is indeed an inventory management problem.
32.
by thecompilr 4y ago
large die area
33.
by thecompilr 4y ago
To those like me wondering what ISA it is, the official documentation says: Runs binaries for x86, Arm, and RISC-V in addition to native ISA I am quite a bit skeptical. OTOH, if it were aarch64 compatible with 1024 bit SVE support, it could
34.
by thecompilr 4y ago
Well, at least it was slower a while ago, for the networking code I was working on. Could be outperforming for other use cases. As for perf, when you run it you can see that a lot of cpu is sent on work stealing and other bookkeeping. Which
35.
by thecompilr 4y ago
For me async is about ergonomics first of all. When you perform parallel tasks on multiple threads it is hard and ugly (in cross platform Rust at least) to implement any sort of intricate cross communication, as communication between thread
36.
by thecompilr 4y ago
I'm the co-author of one of the papers referenced in the blogpost, (Fast Quicksort Implementation Using AVX Instructions), we did write the AVX512 code back in 2015, just had nowhere to run it, at least publicly. The paper also very ex
37.
by thecompilr 4y ago
Thanks, you should use AES-128-GCM, it is about 40% faster and is much more common on the internet.
38.
by thecompilr 4y ago
Maybe, maybe not, since it is a custom configuration can be anything. A coffee lake CPU should be capable of 25Gbps of AES-GCM for a single connection. Not so much for ChaCha20-Poly1305. There is a significant difference.
39.
by thecompilr 4y ago
But it doesn’t say what TLS cipher suite it is using, making it a pointless benchmarks
40.
by thecompilr 4y ago
Yes, Rust with const generics could probably do that, but I am not aware of an existing implementation.
41.
by thecompilr 4y ago
I don't recall exactly, but at least a day each I would say? They were all performance critical, so had to put some effort. Two of them were for internal nginx patches, but were storing data a bit differently so code couldn't be q
42.
by thecompilr 4y ago
That reminds me that I had to reimplement basic hashmaps in C at least three times in my career. What a waste of time.
43.
by thecompilr 5y ago
Hi, ReadySet engineer here. It does replicate using the replication log.
44.
by thecompilr 5y ago
Ukraine want to punish Russia in whatever way it can, which is totally understandable, but it will leave Russians with only internal propaganda, which will make them even easier to control for the regime. Not sure how is it strange?
45.
by thecompilr 5y ago
This is what Putin wants, don't do this! Please!
46.
by thecompilr 5y ago
Probably red green blue = amd nvidia intel
47.
by thecompilr 5y ago
Yeah, I had heart palpitations, it made me to reduce my coffee intake to just three per day and one red bull a week.
48.
by thecompilr 5y ago
When I was working at Intel, Boris Babayan, one of the original creators of Elbrus and an Intel fellow back then, came to visit the team I was working on. He was working very hard to sell the VLIW to us. That was post Sandy Bridge, which tu
49.
by thecompilr 5y ago
There is now doubt the M1 is very much cooler, but they also made the laptop much thicker to allow sufficient airflow without hitting the fans. I am pretty confident that this is the main reason to make it so much thicker than the Intel one
50.
by thecompilr 5y ago
That is not how you compare operating systems. They used the same system. It is not a CPU benchmark.
51.
by thecompilr 5y ago
The post I replied to said M1 only gonna help user applications, not servers, I claim that it helps both.
52.
by thecompilr 5y ago
Huh? I am developing natively in macOS for a Linux server using Rust. Everything just runs fine on both. Unless I need to go really low into syscalls I have no reason to be writing and testing my code on Linux.
53.
by thecompilr 5y ago
Awesome, now where do I sign for the free fastly account? Oh right, they only do enterprise, so who the hell cares? Cloudflare OTOH caters for the whole spectrum.
54.
by thecompilr 5y ago
It doesn’t matter. That wasn’t a bash against Graviton 2, I am just really looking toward the next generation, because I think it will be so much better, but sadly no announcements yet.
55.
by thecompilr 5y ago
I mean it was pretty competitive almost two years ago, but AMD made a huge leap since then. Not sure why saying a two year old CPU is outdated is a controversial take. https://www.anandtech.com/show/15578/cloud-cla
56.
by thecompilr 5y ago
That is nice, but where is Graviton3? Graviton2 is two years old by now, it can hardly compete on performance and power any longer.
57.
by thecompilr 6y ago
They describe how they got the info in the original article, and then deduced the rest. They called to confirm the suspcicions, which were already enough by themselves. They got lucky that the details were sufficient to make the killer th
58.
by thecompilr 6y ago
Much easier would be to break the 512b operation into 4 128b uops on the small cores.
59.
by thecompilr 6y ago
I’m in this picture and I don’t like it
60.
by thecompilr 6y ago
Clang is definitely not better for ARMv8 yet, in fact clang doesn’t even have an optimization profile for the M1 yet.

More ›