jrk

9 ms·

61.
by jrk 4y ago
Lighting and composition can do a lot to compensate for a weak camera, but read the post: this isn’t saying “Zoom video looks bad [potentially because of bad lighting and composition],” it’s saying “even high-end webcams are enormously wors
62.
by jrk 4y ago
It's not exactly the same, but a similar left-to-right tree outliner, specifically optimized for writing, is https://gingkoapp.com
63.
by jrk 5y ago
Patents are generally not written by engineers. They are written by patent attorneys / outside consultants. The engineers have a few conversations with the attorneys, and share technical material (in this case, something specific about
64.
by jrk 5y ago
Yep!
65.
by jrk 5y ago
The use of "conservative" here doesn't seem to refer to political alignment. It is common to talk about "conservative business culture" and "conservative dress" to refer to cultural factors — exactly aroun
66.
by jrk 5y ago
They did not use the exact same mRNA sequence: https://github.com/NAalytics/Assemblies-of-putative-SARS-CoV... The different doses were guesses made to balance effectiveness with severity of side effects. (Earlier Mo
67.
by jrk 5y ago
The rumor has long been that the future Mac Pro will use 2-4 of these “M1X” dies in a single package. It remains to be seen how the inter-die interconnect will work / where those IOs are on the M1 Pro/Max die.
68.
by jrk 5y ago
I think the question is raised because Firecracker was released years ago and has been posted and discussed many times before.
69.
by jrk 5y ago
Video and VFX generally don't need to keep whole sequences in RAM persistently these days because: 1. The high-end SSDs in all Macs can keep up with that data rate (3GB/sec) 2. Real-time video work is virtually always performed on
70.
by jrk 5y ago
Yes, but only at the very extreme. It's normal that a high core count part at low clocks has higher efficiency (perf/power) at a given performance level than a low core count part at high clocks, since power grows super-linearly w
71.
by jrk 5y ago
Note the point in the essay that statistical significance is meaningless if the model does not correspond to reality — which, in this case as in many, they very much do not.
72.
by jrk 6y ago
Coauthor of the paper here: this tangent does not apply to this post. The key thing about the actual work linked in this post is that it is pointing out and addressing an important case where discontinuous computations (polygon coverage at
73.
by jrk 6y ago
> The period of growth was a period of massive decline in the price of energy. This is referring to the decades before 1970, not after the oil crisis. > We've since had 50 years of stagnation in energy prices. This is the perio
74.
by jrk 6y ago
Lots of the comments are talking about this as an MIT/Harvard issue. To be clear, MIT and Harvard are the first to file suit, but this radical policy by ICE dramatically affects every college and university, and every international stu
75.
MacOS Catalina: Slow by Design?(sigpipe.macromates.com)
2031 points by jrk 6y ago|998 comments
76.
by jrk 7y ago
Aside from the fact that you may spend a week or more spreading virus all over the place and to potentially dozens or hundreds of other people with this approach, there is at least a 1 in 10 chance that, if infected, you will need hospitali
77.
by jrk 7y ago
A key idea of this language is that you _don't_ give up control over your data structures. You program your algorithm against abstract data structures, as in most high-performance DSLs, but then you _explicitly control_ the implementat
78.
by jrk 7y ago
Nice work! @epenson: unfortunately, it only took a few hours for someone to create https://fractal.parts/swastika and get it on the gallery :(
79.
by jrk 7y ago
This is referring to Electron for the (PC-based) editor UI, not the deployed game UI. (Plenty of criticisms of Electron still hold, of course.)
80.
by jrk 8y ago
Good to see the TVM troll back. As ever, it's unclear what "substantially more advanced" means, and to cherry-pick just some of the biggest users, there's a large and growing amount of Halide code in Android, Photoshop,
81.
by jrk 8y ago
Sorry, poor choice in the default Bootstrap style we adopted years ago. Under normal conditions the styles kick in in a second or less, but I've fixed it nonetheless.
82.
by jrk 8y ago
This paper is not about the core of Halide, it's about extending it with automatic differentiation, and about a range of applications that makes possible. But TVM indeed includes more components than just its language and compiler—in p
83.
by jrk 8y ago
The RISC V V-extension follows a Cray-style model, and builds on the work with Hwacha to do mixed-precision vectors much more cleanly: https://pdfs.semanticscholar.org/b9e8/fcf11b662a31cecd5a08d5... https://
84.
by jrk 8y ago
Cool, but a bit grandiose and historically arrogant: "[T]he industry's first CPU performance virtualization software" "A New Frontier For JIT Compilers… JIT compilers use as input a virtual machine code format… Dynimizer
85.
by jrk 9y ago
Is there a security story for this interface? Internal NVMe devices are generally trusted and not passed around, but plugging untrusted things directly into the PCIe bus, with coherent access to memory, is a potentially huge security hole.
86.
by jrk 9y ago
To be clear, I had read the whole post, I was just being terse since the emphasis seemed to be so heavily on an apples to bananas comparison (I believe 100% of the results cited in the prose, many in bold, are with mismatched precision and
87.
by jrk 9y ago
[Edited] The top line results focus on comparing four TPUs in a rack node (which marketing cleverly named “one cloud TPU”), running ~16 bit mixed precision, to one GPU (out of 8 in a rack node), also capable of 16 bit or mixed precision, bu
88.
Can High Performance Be Portable?(cacm.acm.org)
2 points by jrk 9y ago|0 comments
89.
Halide: Decoupling Algorithms from Schedules for High-Performance Img Processing(cacm.acm.org)
3 points by jrk 9y ago|0 comments
90.
by jrk 10y ago
Related: my immediate thought was that this totally ignores basic aesthetics and readability. This is partly a matter of taste, but to say that the only virtue of Markdown is popularity ignores why Markdown became so popular, when ReST, Tex

More ›