Connect with us

Tech

Nvidia, AMD’s new chips spice up AI race for data centers

Published

on

Nvidia, AMD’s new chips spice up AI race for data centers

Nvidia (NVDA) and Advanced Micro Devices (AMD) introduced their latest generation of AI chips — Nvidia’s Ruben and AMD’s MI325X accelerator — as the chip giants accelerate into a new product cycle. I/O Fund Lead Tech Analyst Beth Kindig joins The Morning Brief to comment on the market share these chip leaders have room to acquire as they roll out more of these products in the tightening AI race.

“So in this AI battle, this AI competition between AMD and Nvidia, what we’re seeing is something that has never been done before at the data center level,” Kindig tells Yahoo Finance. “They’re basically going to be releasing these every year, as you said, there’s Blackwell by the end of this year, Blackwell Ultra, Ruben, Ruben Ultra. And within that, there’s additional variations.”

Catch Yahoo Finance’s full interview with Nvidia CEO Jensen Huang.

For more expert insight and the latest market action, click here to watch this full episode of Morning Brief.

This post was written by Luke Carberry Mogan.

Video Transcript

Well, the chips war is heating up once again this time, A MD and NVIDIA both pushing higher after announcing new next generation A I chips over the weekend, Nvidia’s Ceo Jensen Wong touting a new chip called Ruben, which he expects to release in 2026 A MD announcing a next generation A I chip for laptops with its CEO Lisa Su saying that A I is the company’s quote number one priority for more on what this means for the space and some of those investment opportunities we wanna bring in Beth Kendig, I owe a fund lead tech analyst, Beth, it’s, it’s good to talk to you here.

So first, when we see some of this excitement once again playing out in uh early uh market action here this morning, both A MD and NVIDIA shares pushing to the upside.

What do you make of the announcements that we got this weekend and where it really tells us about the fact that we’re still, it sounds like very early within the cycle.

Yes, early in the cycle is a great way of putting it.

Thanks for having me on uh basically a couple of weeks ago.

We had written that um in a Forbes analysis that uh the product road map is the first line of defense is really important to understand that by speeding up the product release cycle from two years to one that companies that were previously very cyclical, uh you had to be very careful about your timing are now becoming more secular.

So in this A I battle this A I competition between A MD and NVIDIA.

Uh what we’re seeing is something that has never been done before.

At the data center level, they’re basically going to be releasing these every year.

As you said, there’s Blackwell by the end of this year, Blackwell Ultra Ruben Ruben ultra.

And within that, there’s additional variations within Blackwell, there’s A B 100 A B 200 A GB 200 NVL 72.

So within even that product release cycle of every year, there’s a additional variations that are coming out.

What that means for investors to understand is that previously cyclical companies are becoming more secular.

And then when you add on the software layer that’s really gonna address at home secular gains.

Yeah.

And that, that certainly mirrors kind of the tone that we heard at Computex from the NVIDIA Ceo Jensen Huang where he was talking about accelerating everything.

And so how does the mindset and an operation of accelerate everything translate through to market cap and valuation for the company from your perspective and, and really kind of ramp up where this could continue to move towards.

And this astronomic growth we continue to see we have a $10 trillion market cap uh target for NVIDIA by 2030.

And I’ll be Candid and tell you, I think it’s too low.

Uh The reason is that already we see right now about 60,000 GP US in the data center.

Uh Jensen Wang said on stage that we will go toward a million GP U data center at some point.

So 10 trillion could even be too low.

And here’s why one generation which is Hopper gave us a 1.5 trillion push.

So now we have four additional generations that they’ve laid out for you over the next 4 to 5 years.

Now, you got to add on software and you have to add on something like the Omniverse platform, which is how all robotics likely a near monopoly of robotics will be developed because they’re the simulation company coming from gaming routes.

They basically have an engine and a software platform that will allow uh robotics, that is so far ahead that other people, other companies are going to have a hard time competing and that comes from uh you know, simulation routes already.

So when you combine now you have software to think of, you have automotive, to think of, you have robotics to think of A a and that’s already, you know, pushing us easily past 10 trillion by 2030 of course, when but just quickly on, what kind of market share are we talking here?

Because I mean, there is a competitive landscape where we gotta imagine that, you know, none of the other chip companies are just gonna fall in a ditch.

I estimate that A MD is capable of taking somewhere between 10% to maximum 20% of the market.

This is based on the gaming competition.

These two companies have had for some time Intel, they are moving well past 20% Intel is an easier competitor than NVIDIA.

I think NVIDIA clearly is going to be a tougher competitor for a MD.

So I would say, you know, 10 to 20% A MD on GP US when you take a look, Beth, at least Nvidia’s valuation right now, the massive run up that we’ve seen in the near term.

Do you see, I, I know you’re seeing obviously a huge opportunity here over the next several years, but in the near term, do you see any sort of the fact that the valuation could be stretched at all?

That is probably the most important question, which is not, where will NVIDIA be in 2030?

But are we buying now?

And it’s not actually a valuation concern because we’re talking a company that has gone up, you know, 6 700% on the bottom line.

So evaluation actually is quite reasonable because of all these raises and beats what NVIDIA is going to need is broad market participation.

And if you look at the broad market, things are a bit weak.

So what the the best time to buy a stock in our opinion is when broad market is participating, which means can NVIDIA withhold or can NVIDIA hold off on a broad market, sell off?

Can it um sustain?

And that’s the question that tech investors need to understand is that at times it has nothing to do with the stock, it has nothing to do with the individual management team or these amazing fundamentals or the valuation which we are calling eerily low.

What it has to do with is broad market and sentiment.

All right, Beth Kay, great to have you.

We look forward to hopefully having you back here on Yahoo Finance again soon.

I will find a lead tech analyst.

Thanks so much, Beth.

Thank you.

Continue Reading