AMD is gearing up for its annual Advancing AI 2026 summit, scheduled for July 22-23 in San Francisco, where it plans to showcase its latest Instinct GPU platforms designed to take a swing at Nvidia’s stranglehold on the AI accelerator market. Nvidia currently commands somewhere between 80% and 95% of the AI GPU market.
What AMD is bringing to the fight
The centerpiece of AMD’s push is its MI350 series, which includes the MI350X and MI355X GPUs. These were first unveiled in June 2025, and the specs are worth paying attention to.
The MI350X and MI355X pack 288 GB of HBM3e memory. For context, Nvidia’s Blackwell architecture offers 180 GB. That’s a meaningful gap in raw memory capacity, which matters enormously for running large language models where parameter counts keep climbing.
AMD claims the MI350 series delivers up to 4x the performance on Llama 3.1 405B workloads compared to the previous-generation MI300X.
