Amazon's most advanced chip helps power the AI behind Prime 360—here's how

Key takeaways Amazon's custom Trainium chip helps power the AI behind Prime 360‘s cinematic highlights. Prime 360 highlights are featured within select NBA on Prime broadcasts. Trainium runs the billions of computations each frame requires fast enough for use within a live game broadcast.

During select NBA on Prime broadcasts this season, a breakthrough technology called Prime 360 will deliver cinematic highlights of actual game footage to present on-court action to fans at home like never before.

Built through a unique collaboration between Prime Video producers, engineers, and AI scientists, the proprietary AI-powered technology also relies on diffusion models, a class of AI that demands enormous computing power. To run those models fast enough to work within live broadcasts, Amazon turned to Trainium—its custom-designed AI chip.

How the next generation of AI researchers is using Amazon chips to accelerate discovery University researchers are using Amazon’s Trainium chip to push the boundaries of their work—and making the chip better for all developers in the process.

Trainium, which Amazon designed for workloads that process large blocks of data at high speed, is part of the engine behind Prime 360‘s highlight technology. Prime 360 utilizes an array of high-resolution cameras positioned around the court to capture the action.

That footage is processed through AI models that construct the immersive highlights frame by frame. The real technical challenge isn't capturing the footage, but rather delivering broadcast-quality highlights fast enough to be employed within live game coverage. Why do in-game highlights demand so much computing power?

The visual fidelity of Prime 360 comes from a class of AI models called a diffusion model. It's the same foundational technology behind popular image-generation tools: the model iteratively refines visual output, transforming noise into a polished frame. Diffusion models are powerful, but they are also exceptionally compute-intensive.

Every frame of footage must pass through the model before it's ready for air. Each iteration through the model involves billions of mathematical operations—matrix multiplications, attention computations, and iterative denoising steps—all of which must complete fast enough to keep pace with a live broadcast.

That makes the diffusion process a critical step within the Prime 360 pipeline. The models that generate the final output need to keep up with the fast pace of live television. Solving that challenge requires both software optimization and innovation at the chip level. What makes Trainium fast enough? That's where Trainium, Amazon's custom AI chip, comes in.

Trainium was designed from the ground up for exactly these kinds of workloads—even though this type of application was just a moonshot idea a few years ago. Its architecture follows what Trainium engineers behind Amazon's chip design effort describe as a “tile in, tile out” design: optimized for workflows whose inputs and outputs are large tensors.

Prime 360‘s diffusion model processes each frame by dividing it into dozens of overlapping tiles, each of which passes through the model independently. The inputs and outputs at every step are large tensors (massive grids of numbers that AI models use to represent images, motion, and depth).

The result is diffusion-model inference that reduces the latency by more than 40%, making it fast enough to generate these highlights for in-game use and enabling producers to enhance storytelling and the viewing experience for fans.

Amazon’s custom AI-powered Prime Insights take fans inside the game Real-time analytics powered by custom Amazon AI models bring expert-level insights to Thursday Night Football, NASCAR, and NBA viewers. What does Trainium offer for the future?

Amazon has been building custom chips for years, but most of that work has powered cloud infrastructure behind the scenes—making computing faster and more efficient for the businesses and developers who build on AWS. With Prime 360, that investment is showing up in something millions of fans will see on screen.

It's an exciting example of Amazon using its own custom silicon to power a game-changing AI experience for fans watching NBA on Prime. The takeaway is simpler: those cinematic highlights that traditional cameras cannot replicate are being generated by Amazon-designed hardware.

It's a direct line from chip design to what you see on your TV—and it's just one example of what Amazon's growing custom chip portfolio makes possible. Read the story of how Amazon became one of the world's top chip companies.

Trending news and stories An update from Amazon engineering leadership on the role of the developer in the age of AI Amazon introduces Echo Time, a smarter way to start and end your day Amazon introduces the all-new Echo Show 16—our most advanced, versatile smart display yet Amazon unveils its newest lineup of Al