evanb

53 ms·

631.
Audio First: Spotify Acquires Gimlet and Anchor(newsroom.spotify.com)
7 points by evanb 8y ago|1 comments
632.
by evanb 8y ago
There are also sentences like 'z' is the first letter of this sentence. z yields a finite sentence, a fixed point of the procedure of finding all the instances of that letter and enumerating their positions. Some find no
633.
by evanb 8y ago
1) "Not entirely obvious! A great place for a dogwhistle!" I'm obviously not suggesting that whatever graphic designer intentionally slipped a Nazi symbol into Slack's new logo. But, much like the arrow in FedEx, onc
634.
by evanb 8y ago
I'm sure a rational discussion about co-opted symbology is what what Slack wants associated with their new icon/logo, rather than many people's visceral gut reaction that is almost certainly independent of the sense of rotati
635.
by evanb 8y ago
When I was a kid there was a knife sharpening truck that'd come to my block (Queens, NYC) twice a year. Think ice cream man / food truck, but for sharpening knives. They would take your knives, and you come back some time later
636.
by evanb 8y ago
> includes a cutaway of the back to reveal the insides of this astronomical object. Can you really call it an astronomical object if it's no moon? Is the ISS an astronomical object? ;P
637.
by evanb 8y ago
If you're in the Bay Area you should also visit the Marine Mammal Center in Marin. Make sure you time it so you can take the backstage tour!
638.
Brian Kernighan on successful language design [video](youtube.com)
1 points by evanb 8y ago|0 comments
639.
by evanb 8y ago
There's a great old-timey video explaining the differential in a car works: https://www.youtube.com/watch?v=yYAw79386WI
640.
by evanb 8y ago
You can de-PDF anything on the arXiv by chopping off the .pdf suffix and changing the /pdf/ to /abs/. For example, https://arxiv.org/abs/1104.2804
641.
The Robots Have Descended on Trump Country(nytimes.com)
5 points by evanb 8y ago|1 comments
642.
by evanb 8y ago
You may be interested in Leslie Lamport's "Time, Clocks, and the Ordering of Events in a Distributed System" https://amturing.acm.org/p558-lamport.pdf
643.
by evanb 8y ago
You may be interested to learn of arXiv:1510.07655, "Detecting particles with cell phones: the Distributed Electronic Cosmic-ray Observatory" by Vandenbroucke et al. I don't know if any novel results have come out of this kin
644.
by evanb 8y ago
Depending on the step a LQCD calculation might run on 4, 8, 16, 32, or even more nodes (linear solves and tensor contractions, for example). It's coordinated with OpenMPI and MPI (or equivalent, see my other comment on the software st
645.
by evanb 8y ago
As for efficiency: it depends on exactly what code you use and what your problem is. BAGEL, which included hand-coded assembly http://usqcd-software.github.io/bagel_qdp/ was getting something like 25% of peak, sustain
646.
by evanb 8y ago
There's a large stack of community software that was developed with funding from the DOE's SciDAC program. http://usqcd-software.github.io/ People use the stack in a variety of different ways---I'll describe
647.
by evanb 8y ago
I know some Gordon Bell finalists have used these tensor cores for a dramatic speedup. As far as I know, there isn't any QCD code yet. NVIDIA has a small team of developers who write libraries to make QCD screamingly fast https:&#x2
648.
by evanb 8y ago
It could be---I really don't know. SLURM ran the BG/Q very successfully, which was PowerPC.
649.
by evanb 8y ago
It actually doesn't use SLURM, even though SLURM is originally an LLNL development. It uses LSF. I do not know why the shift. https://hpc.llnl.gov/training/tutorials/using-lcs-sierra-sys...
650.
by evanb 8y ago
We're lucky in the sense that the IO for LQCD is small (compared to other scientific applications), in that we're usually only reading or writing gigabytes to terbytes. But also our code uses parallel HDF5 and it's someone e
651.
DOE to Build Next-Generation Supercomputer at Lawrence Berkeley National Lab(energy.gov)
2 points by evanb 8y ago|0 comments
652.
by evanb 8y ago
Not being in industry, I forget that AWS doesn't have 100% utilization 100% of the time. One of the reasons the lab likes us lattice QCD people is that we always have more computing to do, and are happy to wait in the queue if we get
653.
by evanb 8y ago
Ha, I just replied to your other message asking for an AWS comparison. The answer is yes, the communications has to be substantially better than anything you can order up. LLNL also has national security concerns that are unparalleled by m
654.
by evanb 8y ago
It's not remotely a fair comparison, as typical scientific computing requires high-performance communication to run with any modicum of efficiency. It's not just a matter of scale---the architecture of a machine like this differs
655.
by evanb 8y ago
To add: Titan has 1 GPU per node, we were getting about 300 GFlop/sec/GPU sustained. Sierra has 4 GPUs per node, we get about 1.5 TFlop/sec/GPU sustained. (Summit has 6 GPUs per node, also about 1.5 TFlop/sec&#x2
656.
by evanb 8y ago
I'm privileged to be an early-science user of Sierra (and Lassen) to pursue lattice QCD calculations while the machine is still in commissioning and acceptance. It's completely absurd. A calculation that took about a year on Tit
657.
by evanb 8y ago
Michelson did win the Nobel prize, though the citation reads "for his optical precision instruments and the spectroscopic and metrological investigations carried out with their aid", so not explicitly for the null result.
658.
by evanb 8y ago
This has received some serious consideration in Loeb's "The Habitable Epoch of the Early Universe" https://arxiv.org/abs/1312.0613
659.
by evanb 8y ago
If you're ever in The Hague, the Escher museum is a must-see. Seeing Metamorphosis II displayed in the round was just fantastic.
660.
by evanb 8y ago
Physics has some history of PhD work being recognized. Abrikosov won for his very first paper, if I recall correctly. Wilczek also won for a paper he published as a grad student (albeit with his advisor). And de Broglie essentially won f

More ›