AMD hosted its first AI Developer Day event in Shanghai on May 23, marking a significant expansion of its presence outside the United States. The company used this occasion to unveil new hardware and software initiatives tailored for developers working with artificial intelligence models.
Free public cloud platform built on Radeon graphics processors announced alongside new hardware
The centerpiece of the announcement is a free public cloud platform built on Radeon graphics processors. AMD also showcased the Mini AI Workstation Cube01, which runs on the Ryzen AI Max+ 395 processor and includes 128GB of unified memory. This device integrates with the Zero-One Thing AI agent platform to support advanced inference tasks.
Server configurations for enterprise environments feature dual EPYC 9334 processors paired with eight Radeon Pro W7900 graphics cards. Each server unit contains 512GB of RAM and two 3.84TB solid-state drives to handle intensive workloads. AMD highlighted its full-stack approach, combining CPUs, GPUs, FPGAs, and the ROCm software platform.
Dr. Lisa Su stated that CPU to GPU deployment ratios will shift from eight-to-one or four-to-one down to two-to-one or even one-to-one as AI agents become more prevalent. She predicted that fifty billion people will rely on artificial intelligence daily within five years, a massive increase from the current user base. AMD considers its ecosystem in China a core driver for future product roadmaps.
The company has not confirmed specific launch windows for all announced products beyond immediate availability for certain cloud services and over thirty-five existing mini PC models based on the Ryzen AI Max+ 395 processor.
Discussion
0 comments
Log in to join the thread with a thoughtful take, question, or correction.