aseipp

103 ms·

871.
by aseipp 6y ago
Modern optimizing compilers are already effective black boxes with massive "surface area complexity" that fail in strange ways and have bugs. That they have heuristics written by humans doesn't necessarily mean they're e
872.
by aseipp 6y ago
> If that software then exfiltrated valuable info – session keys, or a digest of hotwords seen, or every string typed into targeted HTML form fields – how would you know? How does this magical exploit work, exactly? How would it extract
873.
by aseipp 6y ago
> but imagine a scheme whereby the IV is chosen to be the first few bytes of the private key xor the port tuple. Again, the NIC doesn't choose the IV. It is given an IV by the host system, which is derived from key exchange in softw
874.
by aseipp 6y ago
Okay, but I'm asking what does it look like? What kind of vulnerability? They presumably aren't using SuperDuper Secret Wifi to mirror data wirelessly to the moon, right? There are a limited number of outcomes at some point. They
875.
by aseipp 6y ago
How do you imagine this wiretapping works? This NIC is installed in your servers, and is fed AEAD keys that are derived from key exchange with a client by your TLS stack (and so private keys exist on the host, not on the NIC). This allows t
876.
by aseipp 6y ago
No it does not, Firefox still uses PDF.js. This is extremely easy to verify by just right clicking and hitting "Inspect Element" on any PDF opened in Firefox (and also you can then just load PDF.js into Chrome on a page, load a pd
877.
by aseipp 6y ago
GPU databases are limited primarily by memory constraints on the card (e.g. ~32GB maximum per card for GV100 or whatever) and interconnect latency/bandwidth, not by raw parallel scan speed. If scan speed was all that mattered, we'
878.
by aseipp 6y ago
There's a long tail of various issues to sort out and "language runtimes that have all the kinks ironed out" is definitely one of them. But I think that's just always going to be the case for any alternative trying to ge
879.
by aseipp 6y ago
It's not really an issue on "desktop class" devices you want to run Linux on and similar to x86 in practice. When people say "modern ARM devices" they are always talking about 64-bit ARMv8. Of those available chips,
880.
by aseipp 6y ago
Most of those don't matter and didn't matter anyway: Chromebooks and Apple desktops are a tiny portion of a small market; Apple's volume chips are iPhones. Intel never really had a grip on any kind of mobile handset market an
881.
by aseipp 6y ago
They absolutely are rolling their own JIT (doing otherwise would reduce its value substantially or simply be infeasible.) The sentence above is just badly written in the original article. The PHP 8 JIT uses the "DynASM" library to
882.
by aseipp 6y ago
You've been able to buy Ice Lake laptops since late last year. So I'm not sure if this is supposed to be a "gotcha", because it's not a good one.
883.
by aseipp 6y ago
You're actually asking for a thing that has already been available for quite a while until now (click through errors). So if you're just now thinking "maybe they should do that", they already have, and it's a strong
884.
by aseipp 6y ago
FreePDK is not a "real" cell library in the sense that the core logic cells it defines are not real ones that can be manufactured/fabricated by any entity. It is a fake kit that can be used to both "estimate" chip a
885.
by aseipp 6y ago
5.4 was released in November 2019, while Navi with the 5x00 series was released in July 2019, so that's actually 2 release cycles or so of the kernel before hitting stable support. For comparison, Intel already has Gen12 graphics suppo
886.
by aseipp 6y ago
You have to be careful what you're talking about when you talk about AVX-512 performance - it's more complicated than you might think at first (like I did!) First, the instruction set and the datapath width are different; you can
887.
by aseipp 6y ago
Jedi Knight: Fallen Order uses UE4 (I can't tell from your phrasing if you meant otherwise, but I figure I'd clarify.)
888.
by aseipp 6y ago
An ECP5 will sit on the order of ~100mW and you can clock those up to dozens of MHz. They can have multiple cores running in parallel (an ECP5 85k will fit dozens, probably well over a hundred RISC-V cores if you do your homework.) Even a l
889.
by aseipp 6y ago
Meh, it's just more proprietary middleware, which is nothing unusual for the Machine Learning crowd. They don't care. Everyone will still use PyTorch or TensorFlow (or just CUDA) and whether or not it directly talks with nvidia&#x
890.
by aseipp 6y ago
I'm not making a claim about the necessity of experimentation. I spent years (and working a paid job) doing programming language work, and also design hardware these days in my spare time, so I'm not against that. I'm specifi
891.
by aseipp 6y ago
GPU based databases haven't reached broad adoption because sending things over the PCIe link is a huge waste of time if you can avoid it. Working around this with custom design like NVLink/NVSwitch do is ridiculously expensive (
892.
by aseipp 6y ago
You have a very limited understanding of what "works" and what is "impossible", it seems. Actual security engineering occurs in hostile environments all the time, in extremely non-perfect environments, with non-perfect t
893.
by aseipp 6y ago
Your motherboard UEFI blob and chip both have to support it. The vast majority of systems are limited by the fact their UEFI implementation does not enable (or allow you to enable) SGX at all, and at least on my Ice Lake laptop, SGX was dis
894.
by aseipp 7y ago
Yes, it does matter, because LLVM is an incredibly complex piece of software. And when you work on a compiler, it turns out you'll have to work on the backend. When I worked on a compiler day-in-and-out, there were single files in LL
895.
by aseipp 7y ago
This is the mandatory rule for Chez Scheme which was only broken once when their entire backend was rewritten, and also (from what I have heard) a large guiding principle for the C# compiler at Microsoft. It's extreme but it's a g
896.
by aseipp 7y ago
It's not the same thing. Chisel will help you do things like write more reusable RTL, because the language itself is more expressive -- so the potential level of abstraction is much higher than what you can achieve in Verilog. You can
897.
by aseipp 7y ago
Nice set of articles, thank you! io_uring is actually pretty easy to use with liburing, and seeing more people adopt it is exciting. The problem with io_uring, really, is that every kernel release has now become much, much more exciting tha
898.
by aseipp 7y ago
Why would you ignore it because of that? LLVM's release cycle is short-ish, but its version bumps are precisely the kind you should pay attention to -- it's a project that changes rapidly with a large amount of features and many s
899.
by aseipp 7y ago
SystemVerilog improves things a lot actually, but the problem is that there is (currently) no freely available, robust synthesis frontend that supports the majority of the useful SV features (e.g. interfaces). In fact I don't think the
900.
by aseipp 7y ago
The extent to which computer programmers literally don't read anything about computing history or alternative systems is quite funny. I can't wait until, sometime in the next few years, someone announces they're "moderni

More ›