dibanez

7 ms·

1.
by dibanez 9y ago
One of the problems with the current peer review system is the ease with which a reviewer can recommend rejecting a manuscript based on personal conflicts of interest (for example, viewing the work as competitive with their own). Making rev
2.
by dibanez 10y ago
Without belittling the difficulties of rural areas, it seems just as concerning to me that all the lines in this plot are trending upward.
3.
by dibanez 10y ago
Its useful to be able to remove safety checks for speed. I have a C++ code where all data is in array objects. Bounds checking is a compile time option, and it makes the overall code 2X slower. I can do testing with bounds checking on, but
4.
by dibanez 10y ago
In my parent comment are listed compile times from scratch, although there probably exist one or two files that would trigger similarly long rebuilds (included indirectly by most other files). If you change a random line of code, chances ar
5.
by dibanez 10y ago
I did the single file trick for the small library that I control, as a configure option. It definitely does help, 2X speedup in serial compile time typically. however, you lose the ability to compile in parallel, which with my hardware give
6.
by dibanez 10y ago
As others have mentioned, I've resorted to brute force (parallel compile using 16 physical cores, good SSD, 32GB RAM). I have a 20KLOC library that I wrote myself (very few templates), that takes 20 sec to compile in parallel, meaning
7.
by dibanez 10y ago
I've been doing this without thinking about it for a while and after reading it from a beginner's perspective It seems like quite a few steps. It is the "right" thing to do though, as far as I know.
8.
by dibanez 10y ago
Seems reasonable. An interesting bit for me was the CITATION file. The lack of citability for software is a recurring annoyance when writing up work that uses that software.
9.
by dibanez 10y ago
Muratori's compression-oriented programming and Acton's data-oriented design have really helped me in writing HPC code. Carmack's arguments make sense for apps that have a clear main function, although they're less appli
10.
by dibanez 10y ago
I'm 80% "software engineer" and 20% "researcher" and have to play both roles to write supercomputer code (I'm the minority, most peers are more researchers). These issues are important right now, as the govt is
11.
by dibanez 10y ago
Yea, I've seen compilers use lea for some integer math in C/C++, because it is handled by a different part of the CPU and so can happen in parallel with other integer math ops.
12.
by dibanez 10y ago
Both, and it agrees with your prediction. KNL's cores are each much faster than KNC's cores. A KNC core was over 10X slower than a mainstream CPU core, and a KNL core seems to only be about 4.5X slower (on my particular code). I a
13.
by dibanez 10y ago
As someone who has programmed both Phis and conventional x86 CPUs, I can confirm that the Phi is more sensitive to data traversal order and NUMA effects on which core accesses which memory. Also, the latest generation (Knights Landing) has
14.
by dibanez 10y ago
One of the interesting things to keep in mind is that these new Xeon Phi cards can be used as standalone CPUs, not just as PCIe cards like a GPU. This is the "self-hosted mode" the article talks about. So one can now think about c
15.
by dibanez 10y ago
Its essentially a matter of realistic benchmark acceptance. Right now machines are compared by how well they run this benchmark called LINPACK, which has been criticized for being non-representative of real science codes. As mentioned here
16.
by dibanez 10y ago
It may be that mobile graphics gets by fine with FP32 or less, but I worry that if FP64 gets sidelined then none of the effort going into this hardware today will benefit applications that need real precision like science, weather predictio
17.
by dibanez 10y ago
Agreed. I used to debug large "integration tests". Now with unit testing, I'm only debugging a bit of new code atop a foundation that is regularly unit tested.
18.
by dibanez 10y ago
I agree, CUDA took a step in the right direction by allowing annotated C++ source to be compiled for the device. As an HPC developer, I'm using the Kokkos C++ library https://github.com/kokkos/kokkos to switch bet
19.
by dibanez 11y ago
I also work in HPC and have been doing a better job of this but still struggling in some areas. For example, the sin() and cos() implementations are different on some machines, and their only guarantee is being within some ulp of the right
20.
by dibanez 11y ago
Well, Go and Rust can do better threading on this problem than a not-really-optimized C code with OpenMP. Thats a start. Message passing has to be shown next, and then one can decide whether the installation and runtime of Go/Rust&#x2F
21.
by dibanez 11y ago
In order for a language to replace C, it has to do two things: 1) be able to write an OS kernel competitive with Linux in this language. 2) get its compiler(s) ported to new architectures, including the Top500 supercomputers, where I work