Akamai rolls out cloud infrastructure and services powered by NVIDIA, optimised for video processing

Akamai Technologies, Inc. has added a new media-optimised offering based on NVIDIA GPUs to its growing cloud portfolio. With the NVIDIA RTX 4000 Ada Generation GPU, the new cloud-based service provides better productivity and economics for companies in the media and entertainment industry that are challenged with processing video content faster and more efficiently.

Internal benchmarking conducted by Akamai demonstrated that GPU-based encoding using the NVIDIA RTX 4000 processes frames per second (FPS) 25x faster than traditional CPU-based encoding and transcoding methods, which presents a significant advancement in the way streaming service providers address their typical workload challenges.

Using Akamai’s offering, media and entertainment companies can build scalable, resilient architectures and deploy workloads that will be faster, more reliable, and portable while taking advantage of the world’s most distributed cloud platform and integrated content delivery and security services.

“Media companies need low-latency, reliable compute resources that maintain the portability of the workloads they create,” said Shawn Michels, Vice President of Cloud Products at Akamai. “NVIDIA GPUs provide superior price performance when deployed on Akamai’s global edge platform. Together with our Qualified Compute Partners and open platform, we give our customers the capability to architect their next-gen workloads to be cloud agnostic and support multicloud architectures.”

The need for industry optimised GPUs

In a market hyper-focused on using NVIDIA GPUs to support large language modeling, Akamai’s media-tailored GPU service homes in on an industry underserved by current industry offerings, which can be expensive. Building on its rich heritage and deep experience in the space, Akamai fine-tuned its new GPU offering to meet the demanding and specific requirements of the media and entertainment industry.

Use cases

The NVIDIA RTX 4000 GPU achieves the speed and power efficiency necessary to tackle demanding creative, design, and engineering workflows for digital content creation, 3D modeling, rendering, inferencing, and video content and streaming.

Media-specific use cases include:

Video transcoding and live video streaming: GPUs can perform faster-than-real-time transcoding of live video streams, improving the streaming experience by reducing buffering and even playback, while GPU-based encoding improves efficiency and reduces processing times compared with traditional CPU-based transcoding. The NVIDIA RTX 4000 GPU is equipped with the latest-generation NVIDIA NVENC and NVDEC hardware, which enables additional capacity for simultaneous encoding and decoding tasks. This is critical for applications requiring high-throughput video processing, such as live streaming. The eighth-generation NVENC engines provide support for the latest video codecs, including the highly efficient AV1 codec, which enables higher-quality video at lower bitrates.

VR and AR content: VR and AR applications require the rendering of 3D graphics and multimedia content in real-time. GPUs are ideal for processing such content. While Akamai optimised the new solution for the media market, the new offering also has applicability for developers and companies looking to build apps tied to several other industry use cases, including:

GenAI/ML: One of the primary applications of GPU cloud computing is in generative AI/ML. GPUs are well-suited for tasks such as training and inference with neural networks, as they can perform many calculations in parallel, which allows faster and more efficient training of new models, which can lead to better accuracy and performance. The NVIDIA RTX 4000 GPU harnesses the NVIDIA Ada Lovelace architecture to deliver exceptional performance in inferencing tasks. A total of 192 fourth-generation Tensor Cores accelerate more data types and include a new Fine-Grained Structured Sparsity feature for up to 4x the throughput for tensor matrix operations when compared with the previous generation. The inclusion of 20 GB of GDDR6 memory provides extensive capacity for large models and datasets.

Data analysis and scientific computing: GPU cloud computing is also commonly used in data analysis and scientific computing because of the nature of its tasks, which often involve processing large amounts of data. These tasks are time-consuming and computationally intensive. GPUs can help accelerate these tasks by processing large amounts of data in parallel, which enables faster and more efficient analysis and simulation.
Gaming and graphics rendering: GPUs are widely used in the gaming industry, mainly for graphics rendering and other tasks related to video game development. This is because GPUs are designed to handle complex graphics processing and can provide fast, high-quality rendering of 3D graphics.

High-performance computing: GPU-enabled cloud computing is commonly used for high-performance computing applications, such as modeling and simulation, that require fast and efficient processing of large amounts of data. GPUs can also be used to accelerate simulations, calculations, and other computationally intensive tasks, which leads to faster results and better performance.

“To support a wide range of workloads, you need a wide array of compute instances,” continued Michels. “What we’re doing with industry-optimized GPUs is one of many steps we’re taking for our customers to increase instance diversity across the entire continuum of compute to drive and power edge native applications.”

AICloudGenAIITtechnology
Comments (0)
Add Comment