Anvidia, which is known to be competing with Intel through the development of ARM-based processors established in Israel, recently recruited two senior employees who headed the development teams of the previous companies in which they worked: Rafi Marom of Intel and Sagi Lahav of Apple. The two have recently updated their LinkedIn pages and now as mentioned, Anvidia has officially confirmed the appointments.
Anvidia today officially confirmed that Rafi Marom, Intel’s Planning Director at the Israel Development Center and the man responsible for the success of Tiger Lake processor architecture and Sagi Lahav, who was responsible for developing Apple processors that are known to be arm-based, will jointly manage the central CPUs at Anvidia’s expanding development center.
Until a few years ago, the company focused on developing graphics processors (GPU) To the worlds of gaming and design. After it was discovered that these processors have significant advantages for the field of artificial intelligence, the company focused on developing hardware and software to accelerate artificial intelligence applications. Anvidia’s GPU is designed for parallel data processing so that it can perform many tasks in a short time – a significant advantage for algorithm training or inference. But this is not clear enough for all artificial intelligence applications, which are carried out in “conventional” data centers. In addition, Intel has also begun to develop graphics processors and penetrate the Anvidia market.
Anvidia’s answer was to compete with Intel in its home ground – data centers that to date have used Intel’s main processors and Anvidia’s graphics cards and AI accelerators.
Last year NVIDIA announced the GRACE processor (CPU) The first in its development for complex artificial intelligence calculations and for the benefit of high performance computing, where it is intended for demanding tasks such as natural language processing, recommendation systems, etc. It will provide a 10-fold improvement in performance in the tasks of training giant artificial intelligence (AI) models. The processor is based on an array of low-energy cores from ARM, and is named after the American computer pioneer, Grace Hopper. Now as stated it turns out that this announcement is a step on the way to the development of cpu processors
As mentioned, it is now taking from its competitors the best of all worlds – Rafi Marom who is responsible for the demanding development of an important Intel chip, together with Sagi Lahav who did the same at Apple and specialized in arm systems.