-
Wanted to record two papers that are reporting the significant benefits provided by alternative number systems for DNN training accuracy and computational efficiency:
https://arxiv.org/pdf/1603.010…
-
Go supports complex numbers, but ~nobody uses them.
The only thing complex numbers add to Go is their complexity to Go's runtime, compiler, and standard library.
I propose we remove them from Go…
-
l2_fused_mv.cpp ought to include cfloat for DBL_EPSILON. Compilation fails on gcc 4.8.5.
-
For any software to accurately interoperate with a geospatial dataset it must be given or make an assumption about the datum and projection used for the geospatial content. It is unacceptable to omit …
-
- Setup CMake build system (#1)
- Setup unit test system (#2)
- Setup CI configuration (#3)
- Setup coding style format (#4)
- Add Dockerfile (#13)
-
So, this is just a general curiosity question from someone who at best may fool with this library on his Arduino's at some point.
I found this repo through [this article on Bonsai](https://www.cnx-…
-
Dear developers of universal,
we are developing a fully templated simulation package for the numerical solution of PDE problems that would allow to test the arithmetic performance of unums/posits f…
-
The standard posit has a maximum fraction at 1.0 of nbits - sign_bit - 2_region_bits - 3_exponents_bits = 64 - 1 - 2 - 3 = 58 bits. This is more bits than captured by a regular double, and thus double…
-
Likely related to the failures of 64bit_posit, but unclear where these failures come from as the fraction size of a posit is covered by a regular double.
-
We don't have Apple gear to build on, so I am looking for folks that can keep a Mac build clean.
We don't have Intel compiler gear, so I am looking for folks that can keep an Intel build clean.
We d…