AMD is ramping up its AI efforts to counter Nvidia’s dominance in this space. Synovus Trust Senior Portfolio His Manager Daniel Morgan talks to Yahoo Finance Live about AMD’s challenges and opportunities.
video transcript
Rachel Akufo: have understood. Well, AMD faces off against Nvidia. Earlier this week, the semiconductor company unveiled its most advanced GPU for AI to date. And now AMD is poised to become a major player in the coming AI wave. Investors are interested in which other chip makers are best positioned to counter Nvidia’s dominance.
Now let’s hear from Daniel Morgan, Senior Portfolio Manager at Synovus Trust. So how tough has this made him to be competitive for Nvidia?
Daniel Morgan: Well, Rachel, we have to remember that Nvidia was the first company to enter the AI market. The company was one of the first chip companies to seriously embrace AI before AI became a big hot word. They have an 80% market share. Nvidia said he had around 40% market share in 2019, so we can see that the market share has actually doubled for him.
And now there are many companies. Apparently on Tuesday AMD announced his AI blueprint as you mentioned. They are working on some new things. I heard about the deal with Microsoft (codename Athena).
They also had another chip, the Instinct MI300, which they advertise directly against Nvidia. They also have an architecture called Alveo that they are pushing in that area as well. That’s where they’re trying to grow. But keep in mind, Rachel, AMD only has about 3% market share in the server data center space. So they have a lot of work to do.
Rachel Akufo: Sure, they’re at least getting a start. Enjoy the big splash here. What do you think are the potential headwinds in this space for AMD?
Daniel Morgan: It’s just that everyone is using their own chips because Nvidia was the first to hit the market. And what came later is actually kind of like Intel was the dominant player in his CPU market and AMD was always trying to take market share against them. So they are no strangers to being underdogs.
But the biggest challenge is sort of breaking and proving that their chip is better—that it can run faster. These are more efficient than what people are already using with Nvidia. So it’s always hard to be the number two player when the number one player holds his 80% of the market.
Rachel Akufo: And especially when we see other companies trying to build their own chips in-house to advance more advanced technologies, it raises another question. Of course, many of these chips are in China, but in fact the majority of these chips are in China. Chips are sourced from In this space we have AMD, in this space he has Nvidia, but really Chinese and Taiwanese semiconductors are at the center of everything. How will this move affect how investors should approach this?
Daniel Morgan: Right, Rachel. Namely, there was an earlier announcement that China banned some of the Hopper 100 Nvidia chips. Now that’s a big tip for them. I think there are probably some components that you feel are a security risk. It is a risk in terms of a backlash from China. About a month ago it was announced that they would be banning some of Micron Technology’s DRAMs used in certain products.
Then, as you said, the risks of Taiwan Semiconductor, Samsung, and Global Foundries. Taiwan Semiconductor produces almost all of Nvidia’s chips. They are in a region of the world that can be very hostile to China. Could there be more escalating issues that could adversely affect chip production capacity?
So now, not only is China repelling various US-made chips made by TSM, but there is also the risk that the dispute between Taiwan and China could somehow escalate. It may affect TSM’s chip production capacity. So there is definitely some risk in it.
Rachel Akufo: So clearly AMD and Nvidia are still in the early stages of this space. But just as we’re seeing other tech companies dive into this, we’re really seeing some astronomical predictions here. You can see AI dotting everyone’s earnings call. But what do you think about some of the predictions you’re seeing about where generative AI and larger language model predictions will go in the future?
Daniel Morgan: Rachel, by the Nvidia report you and I talked about a month ago, we’re already starting to see positive effects. There hasn’t been much evidence that AI is actually having a real, large, positive impact on these semiconductor companies. I know Nvidia came out. They raised their targets for next quarter. I think the total revenue was $11 billion.
This is a significant increase over what Street was asking for. So finally we saw some evidence. But looking at these other companies, we’re starting to see some evidence. Marvell Technology expects its chip, called the PAM4 DSP chip, to bring in about $400 million in revenue next year. that’s a positive thing.
Other companies have emerged to say they are seeing real positive effects from some of the AI chips they are launching. But keep in mind, Rachel, what’s happening right now is that there’s been a long-term slowdown in data center capex spending. The numbers for Azure, AWS, and GCP show that as well. To combat that, spending on AI is increasing.
The key is whether it’s enough for these chipmakers to offset the long-term downtrend we’re seeing right now. A small flicker on the data center screen. And so far he’s the only one who’s been able to really show it is Nvidia.
Rachel Akufo: What they have to do there is certainly going to be a delicate balance. It’s always great to have you. Daniel Morgan, Synovus Trust Senior Portfolio Manager. Thank you for joining us this morning.
