Written by Stephen Nellis
SANTA CLARA, Calif. (Reuters) – Synopsys on Wednesday unveiled a suite of software tools designed to make designing cars, data centers and other large systems that rely on semiconductors easier and faster.
Synopsys is one of the leading companies developing software for designing the chips themselves, helping companies like Nvidia decide how to place hundreds of billions of transistors on tiny squares of silicon. I am.
But with its $35 billion offer to acquire engineering software company Ansys, Synopsys also aims to help customers design the products and systems in which those chips will ultimately be used. At the annual developer conference in Santa Clara, Calif., on Wednesday, Synopsys CEO Sashin Ghazi outlined how some customers are doing it.
For example, Tesla uses virtual simulations of its custom chips to begin writing software for its chips and test how the software will control the car long before a physical product is manufactured. Masu. Other auto customers will likely follow suit, Sasin said in an interview, without naming him.
“Tesla was a pioneer in seeing the car as a software-defined vehicle,” Ghazi told Reuters. “Many European and Japanese OEMs are trying to go down that path, and China is competing to go down that path as well.”
Synopsys also says that operators of the giant data centers that power artificial intelligence systems like co-pilots and chatbots need to know how their software works on tens of thousands of chips, and how much the chips emit during the process. A method for simulating the amount of heat generated is also outlined. Determine the amount of cooling equipment required.
“The speed at which AI models evolve and produce results is limited by how quickly supercomputing architectures and hardware can evolve to meet this new challenge,” said Microsoft Silicon, Cloud Hardware. said Reynold Dasa, corporate vice president of infrastructure. said in a statement.
(Reporting by Stephen Nellis in Santa Clara, California; Editing by Chizu Nomiyama)