Why do all supercomputers in the world use linux?
In: Technology
In the old days, each computer was built with a purpose made operating system. Software was written to run on specific hardware. Then IBM made OS/360 and showed that the OS could be abstracted enough from the hardware that software written and compiled could run on newer hardware. Customers flocked to IBM.
Cray used to have a proprietary Cray Operating System (COS). That meant you could write software that would run on a Cray, but only on a Cray. Customers challenged them to adopt Unix, so Cray created Unicos, which was their Unix variant. C Compilers (cc) for Unix systems started around $3000 and went up from there. Even Microsoft Visual Studio cost thousands of dollars into the early 2000s.
GNU came along (GNU Not Unix) to create an open source and free standard C & C++ version of those expensive compilers called gcc. Of course, this proved the value of open source software and unchained developers from paying OS makers for their proprietary and expensive compilers.
GCC also made it easier to develop software to run on the open source and free Linux kernal. Soon, developers were preferring Linux over (often) proprietary and expensive Unix variants. University developers in particular could create new solutions without needing such a big budget for developer suites. This drove innovation on Linux, including on MPP clusters using Beowulf.
The supercomputer companies, universities, and governments soon found that an entire ecosystem had grown around Linux and their customers were demanding it.
Latest Answers