Nvidia starts laying groundwork for future open and parallel code

Graphics processors are becoming a must-have in computing, so Nvidia is stepping up its work with standards and open-source communities to downstream technologies once largely exclusive to the company's development tools. A lot of work is being done specifically around programming languages like C++ and Fortran, which are …

[Anonymous Coward]

"...past work done by Nvidia..."

Now done for free by you.

"Of course, Nvidia's own compiler will extract best performance and value..."

Of course.

"The memory model in C++ 11 ... C++ 17 ... C++ 20 ... C++ 23"

What do any of those have to do with the fact that Nvidia should be contributing this code to Vulcan and not trying to trick people into using their proprietary offerings?

[Anonymous Coward]

"What do any of those have to do with the fact that Nvidia should be contributing this code to Vulcan and not trying to trick people into using their proprietary offerings?"

Exactly. Even back in the 90s, gcc would perform with the same level of perf. on T2D/T3E like native Cray tools.

The future for HPC is still (as it was, funnily, things don't seem to change in HPC) in MPI, heir of PVM.

Why not, NVidia, investing in MPI, gcc and some fancy CUDA libs linkable to it ?

[Pascal Monett]

"piping has been laid"

So, the road will be covered up, then three weeks later it will be re-opened for another round of piping, and so on and so forth for the rest of the year ?

Because that has been going on for the past six months in front of my office . . .

[teknopaul]

As a Linux user

NVidia talking up open source always sounds weak to my ears.

[EduQuint]

What about Intel's OneAPI?

Intel has a OneAPI initiative that is for unifying the development for CPU, GPU, FPGA, among others.