CES 2026: AMD Just Showed Off ‘Helios,’ the Hardware That Will Power the AI Content in Your Feeds
When you come across an AI video on Instagram or watch ChatGPT answer your query, do you ever think about how that content was generated? Beyond the actual programs and prompts, generative AI requires a huge amount of computation, especially as its popularity skyrockets. As such, AI companies are looking for more power than ever, which of course means turning to those who make the hardware.
AMD calls Helios “the world’s best AI rack”
In a keynote Monday evening, AMD CEO Dr. Lisa Su showed off the hardware that will soon power everything from ChatGPT to the AI videos that will take over your feeds. Su introduced “Helios” to dramatic music, the company’s upcoming AI rack, which packs an impressive amount of computing power into a rack weighing nearly 7,000 pounds.
Each “cross section” of these racks, if you will, is powered by four key pieces of AMD hardware: the company’s new AMD Instinct MI455X GPU, the new AMD EPYC “Veince” processor, the AMD Pensando “Vulcano” 800 AI NIC, and the AMD Pensando “Salina” 400 DPU. There are some staggering stats here: Helios is capable of 2.9 exaflops of AI computing and comes with 31TB of HBM4 memory. It offers a scalable bandwidth of 43 TB per second and is developed with 2nm and 3nm architecture. The rack has 4,600 “Zen 6” processor cores and 18,000 GPU compute units. In other words, this is no ordinary hardware.
Su’s argument is that the AI industry needs this extra computing power. She notes how the world used one ZettaFlop of computing power in 2022 for AI technology, compared to 100 ZettaFlops in 2025. (For those curious, one ZettaFlop has a value of 10 to the power of 21.) This is no surprise: AI is everywhere, and many of us are using it, whether we know it or not. Some of us use it openly, generating AI videos or running chatbots on a daily basis. But others use AI discretely integrated into functions, like live translation.
Su hosted representatives from OpenAI, creator of ChatGPT, and Luma AI, which creates generative AI video content, to explain how additional computation helps their programs. But during Luma AI’s demonstration of its generations of hyperrealistic videos, all I could think about was how this type of content already tricks people into believing it’s real, even though it’s entirely fabricated, not to mention its impact on human artists. AMD is optimistic about AI and the data centers that power it, but critics have pushed back, citing concerns about the impacts on the communities in which the companies build these data centers.
What do you think of it so far?
Helios will likely be a major success for AMD, but it comes at an interesting time for technology and AI in general. AI is more popular than ever, but it is also more controversial than ever. I see hardware like Helios just fueling the fire back and forth.
AMD Ryzen AI 400 Series
In addition to Helios, Su announced the AMD Ryzen AI 400 series. These newest chips come with 12 “Zen 5” CPU cores and 24 threads, 16 RDNA 3.5 GPU cores, a 60 TOPS XDNA 2 NPU, and memory speeds of 8,533 MT/s. AMD claims that the Ryzen AI 400 series is 1.7x faster at content creation and 1.3x faster at multitasking compared to the Intel Core Ultra 9 288V.
These new chips will soon be available in a number of major PC brands, including Acer, Asus, Dell, HP, Lenovo, Beelink, Colorful, Gigabyte, LG, Mechrevo, MSI and NEC.
