smaddox

47 ms·

481.
by smaddox 8y ago
That's already a thing. Intel has its most advanced fabs in the US for exactly this reason.
482.
by smaddox 8y ago
I'm glad you've found happiness. Have you found meaning? Personally, meaning is more important to me than happiness, though meaning often leads to happiness.
483.
by smaddox 8y ago
I think I remember reading at some point that another hypothesis is that it acts as a repository for bad bacteria, so that your immune system doesn't forget how to make antibodies for those bacteria. I think I remember the sinuses bein
484.
by smaddox 8y ago
Having previously worked in materials science (though not specialised in battery tech), this does sound promising, but without the academic paper (is there a prepub available?) it's impossible to know with any confidence.
485.
by smaddox 8y ago
That's hairy? It looks pretty clean to me. Sure, there are a lot of match cases, but they follow a fairly clean and clear pattern. Pulling them out into a hundred single-use functions certainly wouldn't make it easier to understan
486.
by smaddox 8y ago
Nice! This should be amazing for writing toy compiled languages! This looks way easier to get up and running than an LLVM backend.
487.
by smaddox 8y ago
It's interesting that booleans are stored as 0 or -1 [1]. I wonder what drove that decision. I just checked, and Rust's LLVM stores booleans as 0 or 1. [1] https://cranelift.readthedocs.io/en/latest/ir.ht
488.
by smaddox 8y ago
That article is completely devoid of any technical information supporting the claim that Intel nerfed the 10nm scaling.
489.
by smaddox 8y ago
I'm currently using my MacBook pro for development, so vulkan isn't an option right now. I'll definitely keep that in mind, though.
490.
by smaddox 8y ago
Not sure where you heard that Intel renamed 12nm to 10nm, but you're probably thinking of another fab. Intel's 10nm cell density is comparable to Global Foundry's 7nm [1]. And yes, Intel's 10nm process is very late ; th
491.
by smaddox 8y ago
AMD has taken the lead in performance per watt??? Do you have a source? Everything I've seen shows Intel is either matched or better. If you also take capitol cost into account, maybe AMD is in the lead; I don't really know. Eithe
492.
by smaddox 8y ago
OpenGL is pretty bad, but it's not clear to me that Vulkan is really any better. Importantly, though, it gives more control and so makes it possible to build a good API on top.
493.
by smaddox 8y ago
Great work on the gfx portability layer! I hope to be able to use it in the future, but for the time being I've abandoned gfx in favor of raw OpenGL calls in order to avoid the type complexity in gfx (I spent multiple hours trying to c
494.
by smaddox 8y ago
No... You could theoretically store information in the spin state of a few valance atoms, but all of the tightly bound atoms are already filling all of the spin states.
495.
by smaddox 8y ago
Not my area of expertise, but based on what I've read, you general try to get as close as possible through pure mappings, and then use rejection sampling to throw out the samples that don't fall under the desired distribution. Thi
496.
by smaddox 8y ago
> I still haven't figured out what the universe would like but I am at its service. If you're trying to figure out "what the universe would like", you're going to be searching until the heat death of the universe
497.
by smaddox 8y ago
In the field of condensed matter physics.
498.
by smaddox 8y ago
Thanks. This completely allays my concerns about the PCG's I would actually use. Edit: And I like that MCG with 64-bit multiplicand that she shows. I might switch to that for Monte Carlo applications.
499.
by smaddox 8y ago
Interesting read, thanks! The deflection about xoshiro is not particularly convincing, though. It's much more likely that you'll want to multiply your random stream by a multiple of 57 than you'll want to xor it with a 43-bit
500.
by smaddox 8y ago
xoshiro has flaws: http://www.pcg-random.org/posts/a-quick-look-at-xoshiro256.h...
501.
by smaddox 8y ago
Not my area of expertise, but based on what I've read, Linux maps part of kernel space into userspace memory in order to reduce the overhead of context switches. Spectre puts all of that memory at risk of being read, even though it is
502.
by smaddox 8y ago
Operator calculus is much more general than multiplication. It can involve exponentiation, differentiation, integration, etc.
503.
by smaddox 8y ago
Not to mention operator calculus. Operators do not, in general, associate. Feynman developed his own notation for operator calculus.
504.
by smaddox 8y ago
Interesting read so far. The fish farming story is a nice example. For further analysis around the same topic, https://equilibriabook.com is a great read.
505.
by smaddox 8y ago
What did you find incredibly bad about it? I might agree if you said Aladdin 2 was bad, but I thought 3 was quite good.
506.
by smaddox 8y ago
I'm not the original commenter, but I think it's fair to say that throwing compute at the problem is absurd from an academic point of view (where the goal is understanding), but not from a commercial point of view. Although there&
507.
by smaddox 8y ago
And since "fast enough" is a part of "correct", the order should really be "correct, fast enough, simple".
508.
by smaddox 8y ago
Rust is a much smaller language and stdlib than C++ (though still quite large). Rust makes an effort to make many things explicit that are implicit in C++, such as numeric type conversions. Rust's traits are conceptually simpler than c
509.
by smaddox 8y ago
I can usually get away with visualizing indices as being just before a memory cell (i.e. on the boundary between cells), whereas lengths are spans of cells. For example: | H | e | l | l | o | Index 0 1 2 3 4 5
510.
by smaddox 8y ago
Reminds me of this NPR article from 2014 with pictures showing how small "the biggest catch" is today compared to 60 years ago: https://www.npr.org/sections/krulwich/2014/02/05/257046530&#x

More ›