Researchers at Technion have developed a software package that allows computers to perform calculations directly in memory, bypassing the CPU.
This approach, known as in-memory computing, aims to improve computing speed and energy efficiency by eliminating data transfers between the CPU and memory.
Professor Shahar Kvatinsky and his team at Technion have been working on addressing the 'memory wall' problem for years, which involves finding ways to perform computations directly in memory.
The team has developed a platform called PyPIM, which enables developers to write software for in-memory computing systems in Python and includes libraries to convert Python commands into machine-level instructions in memory.