volta83

43 ms·

361.
by volta83 6y ago
Generational indices allow safe unchecked array indexing in Rust. There are many papers that cover how to do this, starting with Alexis Beissegners thesis, up to the latest Rust belt project paper on GhostCell. There are also many libraries
362.
by volta83 6y ago
So what happens if you try to decode a really big PNG, e.g., like any of the > 1Tb large pngs of space photos that are around?
363.
by volta83 6y ago
>That's simply never going to fit on a 8GB, 40GB, or even 80GB high-end GPU. The only way to work with that kind of data is to think about how to split it up, and have the CPU store lots of the data, while the GPU processes pieces o
364.
by volta83 6y ago
> Being able to AVX512 your L1 cached data for a memcpy will always be superior to passing data to the GPU. The two last apps I worked on have been GPU-only. The CPU process starts running and launches GPU work, and that's it, the G
365.
by volta83 6y ago
You can also create races in safe Rust by using the file system, no concurrency primitives necessary (just create a new file, close it, and re-open it assuming it's still there..). What safe Rust doesn't have is _data races_.
366.
by volta83 6y ago
How do you know that the source code they published is the same code that the site uses?
367.
by volta83 6y ago
> Nuclear waste isn't a problem. It never was. If its so easy as you say, why doesn't every country have a small room to keep all the waster? Why do we often "just" throw it in the ocean instead? https://en
368.
by volta83 6y ago
as long as you are not switching compilers during development, those formats are stable..
369.
by volta83 6y ago
Enabling opt-level=1 also does wonders. It enables dead code elimination, and removes a lot of code very early.
370.
by volta83 6y ago
- Knock Knock * Who's there? - GPUs. * Closes door. - Knock knock - Knock knock - Do you really want to ignore ~90% of the compute throughput and ~90% of the memory bandwidth of your machine? - Knock knock - ... ---- An AMD EPYC 7742 (
371.
by volta83 6y ago
Agree, you probably want to just read some cpu time stamp counter before and after, and make sure you ar pinning your threwads, locking the clocks, etc. so that you can get a reliable time from there, or... just use cycles as your unit of m
372.
by volta83 6y ago
> Rust code is just very big. This seems to come with the territory. As someone that uses Rust every day to program STM32s with 128KB of space... i... disagree... Rust is the only high level language I know that allows me to customize th
373.
by volta83 6y ago
Lol no. The only thing this means is that type checking a Rust program might never terminate. Nothing more, nothing less. It might take 10 seconds, it might take 10 days, it might take till the universe freezes. In practice, the compiler -
374.
by volta83 6y ago
The type system is turing complete, and its implementation already supports this, since its required for correctness. People using this feature does not incur any "extra" effort.
375.
by volta83 6y ago
You should open an issue with whoever packages that library. Nobody expects library consumers to be experts in how to build libraries. But this isn't any different that depending on a C library, and discovering that the Makefile they a
376.
by volta83 6y ago
feel you, just pointing out that trying competitive programing and their resources might give you a different point-of-view to attack the issue but like you mentioned, its always nicer if there is no "issue" to be learned in the f
377.
by volta83 6y ago
> The real killer is when I'm trying to prototype. If I don't already know what my interfaces are going to look like, Rust really slows me down. This requires practice, but given the choice, competitive programmer teams picking
378.
by volta83 6y ago
It doesn't do type deduction there. In original C you could declare variables without a type, and these variables with auto have no type. So this "auto" is essentially saying "this variable without a type uses automatic
379.
by volta83 6y ago
> But the thing is that to a zig developer, it's clear that this is explicit static polymorphism getting compiled into a plethora of monomorphic callsites To a Rust programmer it is clear that the toolchain defaults are set up to sa
380.
by volta83 6y ago
And what does it mean there? (automatic storage duration)
381.
by volta83 6y ago
computer graphics is doing a lot of very small (4x4 matrices) linear algebra GPGPU is often about doing few (or one) very big (10^9 x 10^9) matrix op. You can split big matrix op into smaller ones, but with a big matrix you need to combine
382.
by volta83 6y ago
> Any modern discrete GPU will blow away any M1 laptop for training. Counter example: none of the AMD consumer GPUs can be used for training. So no, not any modern discrete GPU blows away the M1. The M1 might not be a great GPGPU system
383.
by volta83 6y ago
+1 I'm in a similar situation as the op, but I didn't buy a new macbook air due to most of the reasons mentioned. Everybody knew that Linux support was going to be broken on day 1 and require multiple months of work to get better.
384.
by volta83 6y ago
How is that better than CUDA streams, CUDA graphs, etc. ?
385.
by volta83 6y ago
its open source, so it has been kept "alive" on and off for years already
386.
by volta83 6y ago
The pipeline is basically: - stream your voice through an encoder: X ms - send encoded packages over the network: 20-100ms latency (fiber vs mobile phone) - potential decoding + encoding (if receiver does not support the senders codec, e.g.
387.
by volta83 6y ago
Thanks for the book, will give it a read!
388.
by volta83 6y ago
and ISPC, and... and...
389.
by volta83 6y ago
> AMD GPUs are better than Nvidia for that Better in what sense? > they had async compute way longer What is that?
390.
by volta83 6y ago
> I'm also not sure why bonds would shield you from liability. You're still giving the company money to do these bad things, why does it matter if you have ownership? Sorry, should have clarified "government bonds" he

More ›