AWS Announces New Amazon AI Chips and Partners With NVIDIA

Amazon Web Services (AWS) is diversifying its cloud services with in-house artificial intelligence (AI) chips and access to NVIDIA’s latest GPUs. This two-pronged strategy of creating its own chips and offering NVIDIA’s sought-after GPUs could give AWS a lead over Microsoft. Revealed at the Reinvent conference in Las Vegas, these Amazon AI chips are designed to give AWS a competitive edge.

“Silicon underpins every customer workload, making it a critical area of innovation for AWS. By focusing our chip designs on real workloads that matter to customers, we’re able to deliver the most advanced cloud infrastructure to them. Graviton4 marks the fourth generation we’ve delivered in just five years, and is the most powerful and energy efficient chip we have ever built for a broad range of workloads. And with the surge of interest in generative AI, Tranium2 will help customers train their ML models faster, at a lower cost, and with better energy efficiency.”

David Brown, vice president of Compute and Networking at AWS, via Amazon Press Release

The demand for GPUs skyrocketed with the release of OpenAI’s ChatGPT chatbot, leading to a chip shortage. AWS also addressed this during the Reinvent conference, stating that they would offer access to NVIDIA’s H200 AI graphics processing units. AWS simultaneously introduced its new Trainium2 AI chip and the general-purpose Graviton4 processor.

The Amazon AI Trainium2 chips, specifically designed for training AI models, promise a phenomenal four-fold performance increase over their predecessor. Several startups have already expressed interest in using these chips. Additionally, AWS’s new Graviton4 processor, inheriting Arm architecture, is a cost-effective, energy-efficient solution that delivers 30% better performance than the Graviton3. With over 50,000 customers already using Graviton chips, AWS is expecting a positive response amidst today’s challenging economy.

According to CNBC, “As part of its deepening relationship with NVIDIA, AWS said it will operate more than 16,000 NVIDIA GH200 Grace Hopper Superchips, which contain NVIDIA GPUs and NVIDIA’s Arm-based general-purpose processors. NVIDIA’s own research and development group and AWS customers will both be able to take advantage of this infrastructure.”

While AWS has not revealed when NVIDIA H200 chips and Trainium2 will be available, it has confirmed that customers can start testing the Graviton4 now, hinting toward its fast-approaching launch.