Synopsys Design Software Uses AI to Make Chips More Power Efficient

A man walks through the Synopsys booth during the Black Hat information security conference in Las Vegas, Nevada, US, July 26, 2017. (Reuters)
A man walks through the Synopsys booth during the Black Hat information security conference in Las Vegas, Nevada, US, July 26, 2017. (Reuters)
TT

Synopsys Design Software Uses AI to Make Chips More Power Efficient

A man walks through the Synopsys booth during the Black Hat information security conference in Las Vegas, Nevada, US, July 26, 2017. (Reuters)
A man walks through the Synopsys booth during the Black Hat information security conference in Las Vegas, Nevada, US, July 26, 2017. (Reuters)

Synopsys Inc said on Monday one of its customers used artificial intelligence software to get a 26% gain in the power efficiency of a computer chip, a leap that usually has to wait for a new generation of chip manufacturing technology.

Modern computing chips are made of billions of transistors and wires laid down on a piece of silicon the size of a fingernail. Precisely how all the elements are placed on the chip, along with other design and architecture choices, has a major impact on how well they perform and how much they cost to make.

Major chip firms like Intel Corp or Nvidia Corp can spend two years and hundreds of millions of dollars to perfect their designs. Synopsys is one of the major makers of software used to do that work.

The company has started weaving artificial intelligence called DSO.ai into its flagship chip design suite to help chip designers get better results, faster, while trying to balance trade-offs on speed, power efficiency and cost to meet their business goals. Samsung Electronics Co Ltd and Renesas Electronics Corp have begun using it, with Samsung last year saying it had cut a chip design step that would have taken months down to weeks.

On Monday, Synopsys said the AI system can now take into account what software will eventually run on a chip to squeeze out more gains. A major cloud computing provider that it did not name got a 26% gain in power efficiency versus the best solution found by human designers.

In the past, gains like those came from a new generation of chip manufacturing technology that would come every two years rather than purely from the design. The new software can squeeze much more out of existing chip factories, said Aart de Geus, chief executive of Synopsys.

“It is significant because design is now actually more of the enabler than ever before,” de Geus told Reuters in an interview.



New Nvidia AI Chips Face Issue with Overheating Servers, The Information Reports

The logo of NVIDIA as seen at its corporate headquarters in Santa Clara, California, in May of 2022. Courtesy NVIDIA/Handout via REUTERS/File Photo
The logo of NVIDIA as seen at its corporate headquarters in Santa Clara, California, in May of 2022. Courtesy NVIDIA/Handout via REUTERS/File Photo
TT

New Nvidia AI Chips Face Issue with Overheating Servers, The Information Reports

The logo of NVIDIA as seen at its corporate headquarters in Santa Clara, California, in May of 2022. Courtesy NVIDIA/Handout via REUTERS/File Photo
The logo of NVIDIA as seen at its corporate headquarters in Santa Clara, California, in May of 2022. Courtesy NVIDIA/Handout via REUTERS/File Photo

Nvidia's new Blackwell AI chips, which have already faced delays, have encountered problems with accompanying servers that overheat, causing some customers to worry they will not have enough time to get new data centers up and running, the Information reported on Sunday.
The Blackwell graphics processing units overheat when connected together in the customized server racks the company has designed, the report said, citing sources familiar with the issue, Reuters reported.
The AI chipmaker has asked its suppliers to change the design of the racks several times to resolve overheating problems, according to Nvidia employees who have been working on the issue, as well as customers and suppliers with knowledge of it, the report said without naming the suppliers.
Nvidia did not immediately respond to a request for comment outside regular business hours.
Nvidia unveiled Blackwell chips in March and had earlier said they would ship in the second quarter before encountering delays, potentially affecting customers such as Meta Platforms , Alphabet's Google and Microsoft.
Nvidia's Blackwell chip takes two squares of silicon the size of the company's previous offering and binds them into a single component that is 30 times speedier at tasks like providing responses from chatbots.