NVIDIA acquired PGI to accumulate energy for high-performance computing

[It168 news] NVIDIA has recently made some moves in the field of high-performance computing, and has bought a product from everywhere in terms of high-performance computing compiler technology reform, the Portland group, an enterprise with a long history, is called PGI ).

NVIDIA announced that the purchase and sale of goods had been completed in July 30, but did not reveal more financial details.

Founded in 1989, PGI is mainly used in the high-performance computing industry to develop compilers, regulators, and R & D tools, such as Fortran, C, and C ++. Previously, it compiled Fortran, C/C ++, and other languages for NVIDIA Cuda parallel computing platform. It also provides support for ARM, IBM, and other vendors.

According to Ian Buck, NVIDIA's GPU computing software manager, NVIDIA has been working with PGI over the past five years to promote its graphics technology in general computing. Although PGI will be purchased from NVIDIA, it can still work for other vendors, including AMD and Intel.

Buck wrote in a NVIDIA blog: "PGI and its talented staff will work under PGI to develop openacc, Cuda Fortran, and multi-core x86 Cuda x86, of course, there is also gpgpu. they will then serve a broad range of customers, including chip makers, research laboratories, and HPC computing cores."

Openacc is a revolution. It is a standard created by NVIDIA, PGI, and super computing startup cray for parallel computing. It has to discuss the roles of people, scientists, and companies that have the same parallel computing standards in heterogeneous CPU/GPU systems. Developers of parallel computing can briefly describe the entire instruction compiler, so that each application can be optimized by GPU acceleration.

With the development of science, technology, and economy in various major companies, the demand for supercomputing is also growing, which also improves the rapid development of high-performance computing, earl Joseph PHS, vice president of computing technology at IDC, said: "high-performance computing servers, especially supercomputers, are not limited to scientific research offices. Industrial innovation and competition have been combined with high-performance computing, and they are urgently integrated ."

With the rapid development of the high-performance computing market, competition in this market is becoming increasingly tense. As a veteran of this market, NVIDIA is indispensable to fully exert its own advantages. With intel's top500 in the year of speaking, it has been putting pressure on nvidia. This purchase from various locations is also related to the enhancement of its GPU accelerator technology. At the same time, NVIDIA's Denver plan, combined with arm, and trust will also bring a great impact to intel.

NVIDIA acquired PGI to accumulate energy for high-performance computing