Scientific IT and artificial intelligence were formerly separated worlds, using different types of calculations on distinctly different equipment. But the two fields merge regularly, as shown by a new massive machine in Berkeley, California.
On Thursday, the laboratory of the Ministry of Energy near the University of California in Berkeley, said that it had selected Dell Technologies to deliver its next flagship supercomputer in 2026. The system will use Nvidia chips Adapted to AI and simulations common to energy research and other scientific fields.
The National Laboratory of Lawrence Berkeley expects the new machine – to be named Jennifer DoudnaA Berkeley biochemist who shared the Nobel Prize for Chemistry 2020 – to offer more than one speed increase ten times compared to the most powerful current system in the laboratory. If it is fully equipped, the machine could be the largest resource in the energy department for tasks such as the training of AI models, said Jonathan Carter, associated laboratory director for computer sciences at the Berkeley Center.
The supercomputer is distinguished by its technological choices, which indicate the growing desire for government laboratories to adopt more technologies from commercial AI systems. The NVIDIA fleas, although widely used by large cloud companies as well as in Superordinators, were transmitted by the Department of Energy for three previous machines which were assembled by Hewlett Packard Enterprise. Dell was hardly a player in the highest end of the Superordinators market, but he succeeded in large IA commercial facilities.
“HPE has swept the Doe space,” said Addison Snell, Managing Director of Intersect360 Research, who follows the supercomputer market. “It’s a big victory for Dell.”
Supercalculators – which are computer systems that occupy entire parts used first for jobs such as the design of weapons and cracking codes – have long been symbols for national prowess in technology.
