Synopsys deal is ‘culmination of everything I showed you’
In this year’s flurry of massive AI deals (with several billion dollars in the pocket), Nvidia’s announcement Monday that it will invest $2 billion to expand its longstanding partnership with Synopsys may seem merely incremental. In an interview with Jim Cramer shortly after the news broke, Nvidia CEO Jensen Huang suggested that it wasn’t. “This is a huge deal,” Jensen said. Here’s why: Synopsys provides software and tools that enable companies like Nvidia to design, test and verify semiconductors. “Nvidia is built on the foundation of design tools from Synopsys, among others,” Jensen said. The deal allows Synopsys, which completed its acquisition of engineering simulation software maker Ansys earlier this year, to leverage Nvidia’s artificial intelligence platform to deliver computer-modeled design and engineering solutions across multiple industries. Nvidia’s powerful chips called graphics processing units (GPUs) are the gold standard in artificial intelligence. With Monday’s agreement, Nvidia will be positioned to bring GPU-powered accelerated computing to the world’s industrial sector, which represents an addressable market worth tens of trillions of dollars. What makes this possible is that the artificial intelligence we are talking about here obeys the laws of physics; This means it can be trusted to show you how things will actually work in the real world. Synopsys CEO Sassine Ghazi, who was next to Jensen, said that what we are talking about here in practical terms is taking a workload that could take two to three weeks and compressing it into a few hours. Despite the work of Synopsys and other electronic design automation (EDA) providers, Jensen said Nvidia spends billions of dollars “prototyping” products in the physical world. “In the future, we will prototype all these products digitally so that we do not waste money when we produce them physically,” he explained. “We can basically do all the engineering work within a computer in a digital twin before we manufacture it. So the type of products we can invent, the quality we can make it, and the speed at which we can do it will be phenomenal.” Industrial companies that make things, like Nvidia, GM or Boeing, spend hundreds of millions, even billions, of dollars on engineering software tools, Jensen said. However, he stated that the money spent for the prototype could be 10 to 20 times this figure. Therefore, the ability to prototype digitally represents a huge opportunity for industrial companies to reduce costs. Jensen told Jim, “This is the culmination of everything I showed you when you visited Nvidia all those years ago. That’s how long it took us to build the necessary software stack for Synopsys and the rest of EDA. [electronic design automation] Industry to speed up software they have historically run only on CPUs [central processing units]Jim Cramer, who started recommending Nvidia shares in 2009, interviewed Jensen for the first time a year later. The “Mad Money” host even changed his dog’s name to “Nvidia” in 2017 to show his faith in the company. Nvidia shares, which were first purchased from Jim’s Charitable Trust in August 2017 and withdrawn from the market in October 2018, changed the company’s shares to “Nvidia” in 2017. It’s been steady since we restarted in March 2019. More recently, Jim hosted Jensen at the Investment Club’s October Monthly Meeting, where the CEO met many former Nvidia investors who made a lot of money on the stock. In Monday’s interview, Jim also pressed Jensen about Nvidia’s recent concerns about whether the launch of Gemini 3, powered by Google’s custom chips, would hurt Google’s own GPU business. Tensor processing units were co-designed by Broadcom. Complimenting Google on its chips, Jensen dismissed the concerns and brought it back to the potential of the Synopsys investment: “You now see a real, tangible example of an opportunity that we can do with our platform that no one else can do.” Artificial intelligence goes far beyond the chatbots and consumer-facing solutions that capture our attention and is contributing to the pressure on Nvidia shares. Jensen said Monday’s announcement is about revolutionizing the industrial software industry, where the stakes are much higher. On the consumer side, responding to a query with 90% accuracy or recommending an item, movie, or new music with 90% accuracy is a pretty good start; but on the industrial side, “the 10% you don’t get right becomes mission critical,” the CEO added, which is also why the pace of progress is so high. Consumer AI. However, no matter how exciting consumer-focused developments are, the real opportunity is likely on the industrial side. While capital spending by the world’s largest tech companies to support consumer AI has so far been the real driver of AI investment and infrastructure spending, the industry is now reaching the point where we need to see increased spending, whether it’s automakers like Ford and GM or shipbuilders in Korea. Diversifying the spending base will help materially de-risk the customer base of companies like Nvidia, which has seen most of its demand come from a select few customers in recent years. Ultimately, this move marks a significant turning point for Nvidia and its AI business, as it lays the foundation for a material expansion in industrial AI. As we can see, this deal is a strong move for both companies, while Nvidia expands its own ecosystem and helps lay the groundwork for even more. GPU-based accelerated computing infrastructure “Of all the AI opportunities (industrial AI, physical AI), it’s the biggest,” Jensen said on a conference call held by both companies to discuss the deal. “And the reason for that is very clear. The world’s industries represent the vast majority of a $100 trillion industry today. That industry, whether you’re designing cars, whether you’re designing trains, whether you’re designing planes, whether you’re designing computers, they’re all heavily reliant on general-purpose computing. … But we need to go even further, we need to be able to do even more.” “We’ve been preparing for several years to expand the scope of design and engineering so that we can do almost anything in the world in a digital environment, long before we created the physical manifestation, this journey, and our announcement today really turbocharges it.” Jensen concluded his remarks by noting that Synopsys has been the company that has allowed Nvidia to design its own chips since its founding, and that the deal announced Monday “will enable that.” everyone needs to design everything that will physically appear in the future.” (Jim Cramer’s Charitable Trust, long name NVDA, AVGO, BA. See here for a full list of stocks.) When you subscribe to the CNBC Investing Club with Jim Cramer, you receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in the charitable trust’s portfolio. Jim on CNBC about a stock THE ABOVE INVESTMENT CLUB INFORMATION, THE ABOVE INVESTMENT CLUB INFORMATION, WITHOUT LIABILITY OR OBLIGATIONS OR DUTIES SHALL BE CREATED BY OR CREATED BY RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTMENT CLUB. RESULTS OR PROFIT ARE NOT GUARANTEED.
