keldaris

44 ms·

391.
by keldaris 9y ago
If you're going to insult people on HN, the least you could do is substantiate your reasoning so the post has at least some content. And while I'm sure there are many definitions by which I could be called "incompetent"
392.
by keldaris 9y ago
Personally, I run a single Win10 desktop (all of my laptops are Linux-only) that I update roughly annually, by allocating an entire day for going through the updates one by one and getting rid of any Microsoft malware along the way. I am fa
393.
by keldaris 9y ago
By prohibiting a clean-room rewrite you're skirting dangerously close to attempting to prohibit independent reinvention. That is, without hyperbole, one of the most abhorrent ideas I've ever seen. Thankfully, software patents are
394.
by keldaris 10y ago
As a fellow Latvian speaker, I agree completely. Personally, I'm glad translations are terrible and/or absent entirely, this motivated me to learn English as a kid and the benefits of an effectively ubiquitous (among computer-lite
395.
by keldaris 10y ago
Not sure why this got downvoted - it's a perfectly valid personal opinion and, albeit expressed rather strongly, doesn't seem to violate any HN guidelines. Have an upvote.
396.
by keldaris 10y ago
The problem for me as a reasonably educated patient (physics PhD, done some work in biological physics, amateur knowledge of medical literature which I enjoy reading) is that the "top" academically performing doctors leave a publi
397.
by keldaris 10y ago
While I guess it's true that not all that many people understand the restrict annotations in the various compilers, I don't think it's an unreasonable expectation from people writing saxpy-like computational kernels. It'
398.
by keldaris 10y ago
Exactly, the STOKE results shown here are actually inferior to those produced by both gcc and clang when you actually know how to use the compiler and restrict pointer aliasing. Are there any actual examples of stochastic optimization techn
399.
by keldaris 10y ago
(Disclaimer: physics postdoc working in an unrelated area of research, not a domain expert.) I'm assuming you mean GR and the Standard Model (which is a very long way away from QM). If so, there's a number of known issues with the
400.
by keldaris 10y ago
Depending on your workload deliberately avoiding the prefetcher (which isn't particularly easy, mind you) ranges between a mild loss to 2-3 orders of magnitude in performance. As a trivial example, consider the element-wise addition of
401.
by keldaris 10y ago
There's already a few benchmarks that are directly relevant, take a look at the FFTW performance, for instance [1] (sourced from [2]), which consists mostly of single precision AVX instructions. The 1800X is handily beaten even by a de
402.
by keldaris 10y ago
Very true, hence "caveat" not "problem" or "dealbreaker". It's something you have to constantly keep in mind as you look at the results, some very fast array iterations will habitually look terrible, etc.
403.
by keldaris 10y ago
They're referring to AMD Jaguar cores as used in the PS4. https://en.wikipedia.org/wiki/Jaguar_%28microarchitecture%29
404.
by keldaris 10y ago
Guru3D seem to be the only ones to manage memory frequencies over 3 Ghz, most reviewers report a lot of issues with the memory subsystem at the moment. In many reviews (see [1] for a representative example) the 1800X is on par or even sligh
405.
by keldaris 10y ago
Having glanced at the code, it seems fairly easy to extend to modern Intel CPUs, which is great. The biggest caveat is the lack of prefetcher simulation, which literally makes many seemingly horrible cache misses a non-issue, but that'
406.
by keldaris 10y ago
Ultimately, the number of cores doesn't mean anything by itself, it's the aggregate performance that matters. We'd all happily take a 16 Ghz single core CPU over a 4 Ghz quad core any day, if such a thing was possible. It&#x2
407.
by keldaris 10y ago
I don't understand how this is even possible in a company operating on that scale. Granted, I'm a lowly scientific programmer with no clue about running a cloud infrastructure, but I would have imagined that there would be at leas
408.
by keldaris 10y ago
I'm someone who's still running an ancient (2007, IIRC) Nehalem i7-920 based workstation for daily use and was eagerly waiting for this launch to finally decide on a new build. After reviewing all the day one benchmark data, I&#x2
409.
by keldaris 10y ago
I'll have to disagree on a few of the points on that list: > - relative multi-threaded performance per $ Highly depends on the particular workload. Given the few benchmarks we have right now, it seems to go 50-50 once you factor in
410.
by keldaris 10y ago
That's absolutely true, I merely mentioned the most obvious counterexample in terms of deliberate atrocities. Most of the other actors are also guilty of deliberate human rights violations on a fairly massive scale, but I thought my si
411.
by keldaris 10y ago
> There's a false equivalency here: Only one side is barrel bombing people indiscriminately, torturing people by the tens of thousands, and systematically exterminating an entire group of people. That seems blatantly false to me. Th
412.
by keldaris 10y ago
As a recently minted theoretical physics PhD, I feel rather the opposite way on this. The vast majority of Nobel laureates still went through the usual university progression, which isn't at all designed to cater to the smartest studen
413.
by keldaris 10y ago
I've found I'm markedly better at two specific types of programming activities while mildly drunk (the "mildly" part being quite crucial, for me this means two beers or a couple of slowly sipped glasses of whiskey). One
414.
by keldaris 10y ago
That's why it's fine to have ASLR on by default. Most developers don't even seem to be familiar with how ASLR works, and very few actively disable it in practice. Yet for some of us, the ability to do so is quite indispensabl
415.
by keldaris 10y ago
I'm definitely not disabling ASLR over this post, that would be utterly silly. I've been disabling ASLR for many years in the context of specific performance-sensitive applications (mostly computational science stuff) in environme
416.
by keldaris 10y ago
I agree, on both counts. For some applications (web browsers top that list nowadays) ASLR has a lot of demonstrable value as a mitigation measure. I could easily imagine security-critical systems where even proper strong ASLR (with randomiz
417.
by keldaris 10y ago
The paper is extremely well written and their attack is very elegant, so all due credit there, however, I've never understood why ASLR was taken particularly seriously in the first place. It's always been obvious to anyone familia
418.
by keldaris 10y ago
Or, as I would put it, the universe at large is simply vastly richer than our limited intuitions, a natural consequence of the limited range of physical properties that supports human life. It would be really strange if the universe didn&#x
419.
by keldaris 10y ago
Disclaimer: Theoretical physics postdoc, working in an unrelated subject area. Not necessarily an expert, but I do have some comments. > Niels Bohr proposed that electrons don't have enough energy to radiate away, and thus are forc
420.
by keldaris 10y ago
FWIW, I have no personal experience with Adderall or any amphetamines, haven't used any (but am rather curious about trying some day), but I also found your post very confusing due to the phrasing. "One or two orders of magnitude&

More ›