Qualcomm (QCOM) is ramping up the AI chip competitors, saying two new chips, one for PCs and one for smartphones. Although these are totally different from the sorts of chips Nvidia (NVDA) makes, it’ll put stress on rivals to ship their very own AI-focused semiconductors. Yahoo Finance’s Dan Howley and Allie Garfinkle talk about the state of the AI chip race.
For extra skilled perception and the most recent market motion, click on right here to observe this full episode of Yahoo Finance Dwell.
Video Transcript
JOSH LIPTON: Qualcomm out with various key AI bulletins, together with new chips, as the corporate faces off in opposition to different opponents on this chip race. They’re right here to debate Yahoo Finance’s Allie Garfinkle and Dan Howley. Allie, let’s begin with you. What’s the newest we bought right here from Qualcomm?
ALLIE GARFINKLE: Josh, Qualcomm’s popping out right here swinging, actually attempting to make the case for itself as an AI play. Now the method they’re taking is what they name on-device AI, AI that’s meant to be extra environment friendly in your gadgets, whether or not that is your smartphone or your laptop computer. And there are two new chips to this finish that they’ve introduced as we speak. The primary one is known as the Qualcomm Oryon CPU.
Now, it is a chip particularly for PCs, and it’s hooked up to what they’re calling their Snapdragon X Elite compute platform. Now, an important factor to know in regards to the Oryon is that Qualcomm’s making some fairly spectacular claims about it– some fairly hefty claims. What it is saying is that the processing energy is quicker than quite a lot of Arm-based competitors– that is a gaggle that features Google and TSMC– and that it hits the height efficiency of chips from firms like Intel and Apple.
In order that’s going to be one thing that you simply’re– you have to be anticipating them to see them actually make a case about sooner somewhat than later, and it is one thing they’ll be speaking about quite a bit. Now, the opposite chip they introduced is known as the Snapdragon 8 Gen 3. Now, it is a chip for smartphones, and it was constructed with generative AI in thoughts. The aim is definitely that you would be able to run massive language fashions like ChatGPT or Meta’s Llama 2 in your smartphone.
Now, that is particularly for Android smartphones. And it is vital to recollect right here that these chips, whereas they’re AI chips within the sense that they’re meant to work with AI, they don’t seem to be the AI chips that we’re speaking about from NVIDIA. These are CPUs. NVIDIA’s GPUs are what are getting used to coach AI fashions. So ultimately, do not count on Qualcomm to return in and remedy that large chip scarcity that we have all been speaking about.
However I’ll say that the success of those chips, I feel, goes to be actually key for Qualcomm’s case as an AI play shifting ahead.
JULIE HYMAN: And Dan, let’s deliver you into this dialog, proper? There’s quite a bit happening on this planet of semiconductors proper now. We have now these bulletins from Qualcomm, there was a report yesterday that NVIDIA was creating CPUs. Put all of it collectively for us and kind of paint the aggressive panorama for us.
AND HOWLEY: Yeah, so the thought right here with the NVIDIA CPUs is that these would even be for PCs. They might be for Home windows PCs competing alongside of those Qualcomm chips, in addition to Intel chips and AMD chips. So actually, that is opening up the panorama, I bought a caveat that the NVIDIA half with, they did not wish to remark. This was a report from Reuters. However the, , panorama now’s going to be infinitely extra aggressive for Intel.
I feel it is vital to level out that not like Apple, which has an Arm-based CPU and runs them because the M1, the M2, and the anticipated M3, Qualcomm, NVIDIA, they do not run the software program that powers the PCs– that is Microsoft. Apple has the good thing about working the software program, so that they have soup-to-nuts management over the system. For NVIDIA and Qualcomm, it may be slightly bit harder to maybe really catch as much as what Intel and AMD are doing in the case of total efficiency.
Now, the businesses within the case of Qualcomm cite total efficiency in comparison with Intel chips, however, , it actually comes right down to whenever you get them into shoppers’ fingers. Do they match up? And, , Qualcomm has been on this area earlier than. That they had labored with Microsoft on a Floor gadget working a Qualcomm chip, so we simply must see if that truly equates to a giant quantity of energy and a giant quantity of battery life.
That is actually the important thing right here, I feel, total is, , Apple managed to place out the M1, the M2, the M3. They will final a very long time and have loads of energy. Can Qualcomm, can Intel, can AMD, can Nvidia do this with a Home windows PC? And that is going to be a giant deal if they will.
JULIE HYMAN: After which, can they collect quite a lot of income and market share from it? Ultimately, we’ll see what occurs, you guys. Thanks a lot for placing this story in perspective for us, our Allie Garfinkle and Dan Howley. And we’ll proceed this dialog, by the way in which. Remember to tune in to Yahoo Finance tomorrow. We will likely be talking with Qualcomm CEO and President Cristiano Amon. We do not wish to miss that dialog.
