Low-Latency VMs and Containers Leveraging NVIDIA GPUs Now Available
StackPath, the industry-leading edge computing platform announced the addition of NVIDIA GPU-Accelerated Instances to its Virtual Machine (VM) and Container product options.
The new instances utilize NVIDIA A2 Tensor Core and NVIDIA A16 GPUs to deliver the computational power required by workloads such as deep learning algorithms, intense graphical processing, and other parallel architectures, which are key to innovative and emerging technologies ranging from artificial intelligence (AI) and machine learning (ML) to augmented reality (AR) and virtual reality (VR).
Recommended: The Ethics of Generative AI: Navigating New Responsibilities
At launch, the instances are available in StackPath Dallas, San Jose, and Frankfurt locations and will be added across the StackPath platform throughout 2024. With StackPath facilities’ proximity to sources and destinations of data, the GPU-Accelerated Instances will pair exceptional computational power and efficiency with incredible data ingress and egress speed.
“Our GPU-Accelerated Instances are exactly what new and next-generation workloads—like AI inference, computer vision, and natural language processing—really need to succeed,” said Tom Reyes, Chief Product Officer for StackPath. “These are real-time applications. So, as much as they need high computational power, they also need exceptionally low latency. The physical location of our platform minimizes the number of hops in and out of our instances, so the advantages provided by a GPU aren’t undermined by geographic distance.”
SP// Edge Compute VM GPU-Accelerated Instances are available in the following configurations:
- 1 NVIDIA A2/16 GPU x 12vCPUs x 48GiB RAM x 25GiB Root Disk
- 2 NVIDIA A2/16 GPU x 24vCPUs x 96GiB RAM x 25GiB Root Disk
- 4 NVIDIA A2/16 GPU x 48vCPUs x 192GiB RAM x 25GiB Root Disk
Recommended: Artificial Intelligence and the Trust Deficit: A Call for Greater Transparency
SP// Edge Compute Container GPU-Accelerated Instances are available in the following configurations:
- 1 NVIDIA A2/16 GPU x 12vCPUs x 48GiB RAM x 40GiB Root Disk
- 2 NVIDIA A2/16 GPU x 24vCPUs x 96GiB RAM x 40GiB Root Disk
- 4 NVIDIA A2/16 GPU x 48vCPUs x 192GiB RAM x 40GiB Root Disk
StackPath edge compute instances are provisioned on demand through the StackPath Customer Portal or API. Instances are billed by the hour and volume of data transferred. Additional options include forming virtual private clouds, leveraging built-in L3-L4 DDoS protection, persistent storage, image capture and deployment, private IP addresses,
Recommended: Modernizing Logistics Operations with Online Freight Forwarding
[To share your insights with us, please write to sghosh@martechseries.com]
Comments are closed.