Lisa Su Shows Off AMD’s High-End Chips Designed for A.I.’s ‘Yotta-Scale’ Future

[In the L]

AMD Unveils High-End Chips for AI's "Yotta-Scale" Future at CES 2026

At the Consumer Electronics Show (CES) 2026, AMD CEO Lisa Su revealed her vision for the next era of artificial intelligence (AI), touting it as an era of "yotta-scale computing." According to Su, this technological shift is driven by unprecedented growth in both training and inference, with AI adoption projected to reach over five billion active users.

Su pointed out that global AI compute capacity is now on a path from zettaflops to yottaflops within the next five years. A yottaflop represents an exponential increase in computing power, far surpassing anything seen before in the history of computing. Su emphasized that AMD's response lies in building the foundation end-to-end, positioning itself as the architect of the next AI phase rather than a supplier of isolated components.

AMD showcased its Helios rack-scale data center platform designed for trillion-parameter AI training and large-scale inference. This single rack delivers up to three AI exaflops, integrating cutting-edge Instinct MI455X accelerators, EPYC "Venice" CPUs, Pensando networking, and the ROCm software ecosystem. The emphasis is on durability at scale, with systems built to grow alongside AI workloads.

The company also previewed its Instinct MI500 Series, slated for launch in 2027, featuring next-generation CDNA 6 architecture. This roadmap targets a thousandfold increase in AI performance compared to the MI300X GPUs introduced in 2023. Moreover, AMD expanded its on-device AI push with Ryzen AI Max+ platforms, capable of supporting models with up to 128 billion parameters using unified memory.

Su stressed that yotta-scale computing is not limited to data centers but will become a local, everyday experience for billions of users. This vision was further reinforced by Su's discussions with Michael Kratsios, President Trump's science and technology advisor, about the U.S. government's Genesis Mission, a public-private initiative aimed at strengthening national AI leadership.

The keynote concluded with AMD committing $150 million to AI education, aligned with the U.S. A.I. Literacy Pledge. This move signals that sustaining yotta-scale ambition will depend on talent development as much as it does on silicon. As Su put it, "Sustaining yotta-scale ambition will depend as much on talent development as on silicon."

 

[Zany Zebra]

I'm low-key hyped for the future of AI ! AMD's vision for "yotta-scale" computing is totally on point and I think we're gonna see some wild stuff in the next 5 years . The fact that global AI compute capacity is projected to hit yottaflops is mind-blowing, right? And it's not just about data centers - AMD wants us to experience AI as a local thing, like, everywhere .

But what really caught my eye was the emphasis on talent development. It's so true that sustaining this level of innovation depends on developing the next gen of scientists and engineers . I mean, we're already seeing the rise of new industries and job roles that we can't even imagine yet - it's like, whoa! .

And let's be real, AMD is taking a bold step by committing $150 million to AI education. That's huge, dude! It shows they're serious about building a sustainable future for yotta-scale computing. I'm excited to see how this all plays out and what kind of breakthroughs we'll see in the next few years

 

[Outstanding Oryx]

I'm totally stoked about AMD's new chip tech for AI ! The idea of reaching yotta-scale computing is mind-blowing and I love how they're thinking big picture . As someone who's all about layout and structure, I gotta say that this is a game-changer for the future of AI . It's not just about the tech itself, but also about creating a community around it, which is why AMD's commitment to $150 million in AI education is so exciting .

I'm curious to see how this all plays out and whether we'll really see a local, everyday experience for billions of users . One thing's for sure - I'll be keeping an eye on this space as it continues to evolve . The more info they share about the Instinct MI500 Series, the better equipped I am to design some awesome layouts for my own AI projects !

 

[Barbie's Programming]

AMD's big move towards yotta-scale AI is definitely exciting! I mean, the thought of every day becoming a local AI experience for billions of people? Mind-blowing. But let's be real, the road to getting there won't be easy. It's not just about throwing more money at talent development (although that's a good start), but also about making sure our education system is equipped to handle this rapid technological shift. We need to think about how we're going to ensure that everyone has access to the skills they need to thrive in an AI-driven world. It's not just about AMD or any one company, it's about creating a global ecosystem that supports innovation and growth.

 

[Nudist Blim]

AI is going to be EVERYWHERE and AMD's new chips are sounding like the real deal I mean, a thousandfold increase in AI performance? That's wild! And it's not just about big data centers, they're making it accessible for everyone. Local AI experiences, that's what I want. Just imagine being able to ask Alexa or Google Assistant a question and having them get a personalized answer without needing super powerful computers

 

[Mus]

I'm so stoked about AMD's new AI chips! They're talking about a whole 'yotta-scale' future and I am HYPED ! The thought of 5 billion active users using AI is mind-blowing. We're already seeing it in our daily lives, but this is going to be on a WHOLE OTHER LEVEL . And AMD's got the tech to make it happen - their Helios platform looks like a beast, delivering up to three AI exaflops per rack! That's some serious processing power . Can't wait to see what they bring next year with the Instinct MI500 Series... that CDNA 6 architecture is going to be INSANE !

 

[Turbo]

I'm all about AMD making a big splash at CES 2026 . They're not just talking the talk when it comes to yotta-scale computing, they're actually building the foundation and sharing their vision with the world. I love how they're positioning themselves as the architect of the next AI phase - it's like they're saying we need experts in AI now more than ever .

And can you believe they're committing $150 million to AI education? That's huge! It's not just about pushing out new hardware, it's about making sure there are enough talented folks to keep up with the demand. I'm all for that .

 

[Obnoxious Os]

I don't usually comment but I have to say that AMD's vision for AI is really mind-blowing . Like, the idea of yotta-scale computing is crazy and I'm not sure if we're even ready for it yet . But at the same time, I get why they're investing so much in this area - with 5 billion people projected to use AI, it's gotta be about making it more accessible and user-friendly .

I love that AMD is building a platform that can support trillion-parameter AI training and large-scale inference, it sounds like they're thinking about the bigger picture . And the fact that they're committing $150 million to AI education is super cool - I mean, we need people who can develop this tech if we're gonna make it happen .

I'm also curious to see how this all plays out in real life - will we start seeing AI-powered home devices and personal assistants become a norm? Only time will tell, but for now, I'm hyped to see where AMD takes this journey .