Akamai is helping bridge the gap between escalating processing demands and efficient content delivery by introducing NVIDIA RTX 4000 Ada class GPUs optimized for media and gaming workloads. In this episode, Shawn Michels, VP of Product Management, Cloud Computing at Akamai, discusses trends in media and gaming industries and how this specific GPU teamed with Akamai’s backbone and private network provides a high-quality 4K-like experience. He says, “It allows us to more cost-effectively deliver that ubiquitous experience to a broader audience, which is ultimately what gaming companies and media companies are looking for.”

New NVIDIA GPUs for media and gaming, improving cost and performance

• Michels talks about the company’s announcement of the availability of the NVIDIA RTX 4000 Ada class GPUs optimized for media and gaming workloads.
• Michels explains how the new NVIDIA GPUs are helping to drive profitability in these industries by improving the amount of performance or transcoding capabilities which can help reduce preparation costs for content.
• Michels discusses why this specific family of NVIDIA cards is well suited to customers who are looking to do video and audio rendering, enabling end users to deliver the high-quality performance you would expect from a streaming service.

Leveraging GPUs for AI inferencing, video transcoding, and data analysis

• Akamai aims to target specific verticals and use cases with GPU-accelerated computing, potentially expanding to additional classes of NVIDIA cards and chipsets for AI/ML inferencing.
• Akamai is also focusing on the distributed computing capabilities they are bringing to market and getting more density to the edge.
• Akamai has partnered with Neural Magic to bring large language model (LLM) inferencing to the edge while using standard CPU technology. However, Michels remarks that other customers are building their models to work on GPUs.
• Michels highlights the use cases for the GPU-accelerated platform that Akamai is announcing, including video and rendering processing, and AI inferencing.

GPUs for targeted use cases and potential for rolling out additional GPU variants

• Michels discusses the growing demand for compute density and data management in various applications, including consumer-generated content, multiplayer online games, and smart cities.
• Michels explains how Akamai is leveraging lots of Linode’s existing components with Akamai’s updated backbone so that they can provide a global high-performance experience.
• Akamai plans to roll out additional GPU variants to address evolving use cases, and partner with companies like Neural Magic to optimize CPU for AI tasks.
• Michels discusses the rationale behind selecting specific GPU models to appeal to a broad audience at a price point that is accessible to not only large media companies but also average consumers.

Guest: Shawn Michels (LinkedIn)
Company: Akamai (Twitter)
Show: Let’s Talk

This summary was written by Emily Nicholls.