AI for Everyone: Intel's Breakthrough in Accessible AI Computing

0:00I think right now the tech world is really really interested as Intel's foray into the world of AI PCs inches closer with the advent of the Meteor Lake. So while many might ask why this matters, I think Intel's ambitious vision and its potential billion dollar value might be the answer. So previously known as Core Ultra, Intel's 14th gen chip is Meteor Lake, and it integrates NPU AI inference engines directly into its processors. So this development has not gone unnoticed. Robert Hallock, who's a former AMD stalwart and now spearheading technical CPU marketing for Intel's microprocessors, kind of highlighted that Intel envisions shipping a staggering 100 million AI PCs by 2025.

0:47I know I, for one, 100 % would buy this. Everyone on my team that is currently working on my software startups, AI Box and everything else, we 100 % would be buying these PCs. We want to be able to test, download, run AI models on our own computers. This makes it so much faster and more exciting to test things in real time. So the term AI PC gained traction this year when Intel CEO Pat Gensler started its kind of promotion. Timed with the start of the Intel Innovation Conference in San Jose, Intel unveiled the architecture behind the Core Ultra chip. So this event served as a platform for Intel to envision what the future might look like with AI at its core, particularly if Intel is the one powering it.

1:33One key distinction of Intel's core Ultra is the transition to the Intel 4 process technology. With this shift, Intel has reimagined its kind of classic two-die design, and it's breaking it into four distinct, you know, quote-unquote has led to a 50 % reduction in power compared to its predecessor, which is really, really impressive, and the 13th gen Raptor Lake. So thanks in part to the novel Low Power E-Cores, this is all possible. I think the kind of crown jewel of this whole evolution is the NPU, which is kind of touted to democratize AI access. So Pat Jensling Singer, during his inaugural keynote, kind of described the AI PC as, quote, a sea change in tech innovation.

2:25So I think the conviction he has is definitely like very palpable across Intel's innovation conference. And the core message I feel like they were really kind of getting out there is everything AI. They're really, really focusing at Intel on making and enabling AI. I think they're seeing right, obviously, that NVIDIA passed a billion dollar valuation. They're seeing how that company absolutely skyrocketed and benefited from AI and kind of focusing their marketing and their tone towards that. And so I think Intel is really kind of trying to follow in that in their footsteps. And I think Intel's aspirations to pioneer a novel AI PC market is definitely evident in response to a query from different tech journals about this potential market evolution.

3:08Their CEO said this is category creation at its finest. To be honest, I completely agree. This is not something that's very possible. you know our cto ben who works at ai box like he has a very top of the line computer that he built himself um he's got all of the top of the line gpus and cpus and stuff and so he can um run different open source ai models on there it's you know it's pretty kind of tricky to set a lot of this stuff up i think this stuff is going to become easier and easier to run and as these um pcs are built with all of the needed hardware inside of them this is absolutely going to democratize access to these AI models, which, you know, we're already starting to find them on hugging faces and other places.

3:49So they're accessible, but it's like, we don't have the hardware to really make them accessible. So I think this is incredibly exciting. Traditionally, AI's domain has been kind of confined to the cloud or smartphones providing functionalities like portrait mode or, you know, filters like that that you're going to see on Facebook or, you know, a bunch of different like social apps. That's kind of where you see it. In contrast, PCs have seen really limited AI integrations, mostly limited to kind of like Windows Studio effects. Intel, however, envisions local PC AI that offers a more personal touch without the need for back and forth data trips to the cloud, right?

4:23Because right now, if you're using AI on your PC, like chances are, it's because it's hooked up to a cloud that's actually doing the processing. It's not processing on your computer. So I think the roadmap to realizing this envision seems multifaceted. To begin with, Intel has meticulously designed the hardware showcased recently in Malaysia. The design promises that the NPU would be more energy efficient for AI tasks than kind of standalone CPUs or GPUs. Additionally, Intel has given glimpses of its future AI innovations with revelations like the Lunar Lake, Arrow Lake, and Panther Lake chips that they are currently working on.

5:00And I think furthermore, AI emphasized the current and future AI applications such as DeepRender, Rewind.ai, and Fabolitics, and collaborations with FitMatch AI. I think these kind of demonstrations serve a dual purpose. Number one, they're illustrating the potential for AI while also kind of subtly addressing consumer questions around the tangible benefits of local PC AI. They're really showing what you can actually do with this. So Michael Johnson, Holathaus, who is Intel's executive vice president and head for Intel's client computing group, recently said, quote, I'm going to deliver a cadence of products yearly.

5:37And that really opens the ecosystem up. This is exciting, right? This isn't just them saying, hey, we got this cool thing. They're really committed to like a cadence of annual products that are constantly being upgraded. I think this is really what consumers want. And this is going to really accelerate AI adoption. And I think while the promise is there, the journey of AI's integration into PCs, right, is obviously at its infancy. They're announcing these things, but we don't have them in our hands yet. I think, as they kind of mentioned, it's, you know, AI is still in its kind of fine-tuning phase.

6:07And I think a lot remains to be seen regarding benchmarks, performance metrics, and how Intel's AI is going to actually compare against industry rivals like AMD and Qualcomm. But I think despite all of this uncertainty, and definitely there's a lot of kind of uncharted terrain out there with this, one thing is definitely clear. And that is that AI's rapid assimilation into our digital lives has been transformative. It was just this year that ChatGBT was launched. And now, you know, we're looking at the fact that in less than a year, every single tech startup is transforming into an AI startup.

6:36Every single company is focusing on looking at ways that they can integrate AI. The ones that are not are frankly being left behind and will probably slowly die. and I think a lot of we're going to see a lot of winners a lot of losers and probably a lot of people that integrate it and continue to just see healthy growth in their company this is going to be really interesting as this entire pc era with ai kind of comes to the forefront and definitely something we'll be excited to watch as time passes