Bringing Cost-Effective, Scalable, AI Performance to the MassesApril 11, 2017
In case you missed it, the NVIDIA DGX-1 AI supercomputer is celebrating its one-year anniversary this month. In just one year, we’ve seen the momentum building around the disruptive performance and value this platform is bringing to deep learning and data science.
A great example of that impact is the demand we’ve seen from cloud service providers embracing the “AI-Cloud Era” and deploying DGX-1 in their offerings to the enterprise.
This is an important inflection point in the evolution of the AI-accelerated enterprise, enabled by DGX-1. Similar with other mission-critical workloads, every organization is finding its own balance in how to consume AI-compute capacity on premises or from the cloud.
Increasingly, many organizations are going cloud-first for those use cases where it makes most sense, especially for those that:
- need the competitive advantage of faster, accelerated delivery of GPU-powered deep learning, without a lengthy proof-of-concept phase prior to reaping production-level scale and success,
- depend on elasticity of scale, and need the ability to address temporal use cases or seasonality in how they run their deep learning practice,
- want the simplicity of not having to own or operate infrastructure, but simply want GPU-accelerated AI at the ready.
Nimbix Cloud with NVIDIA DGX-1
Nimbix is among the first cloud service provider partners to engage this opportunity. With Nimbix Cloud built on DGX-1, customers get a great approach for quickly and cost-effectively supercharging a deep learning practice, or executing a cloud-hosted PoC before deploying on-premises.
The benefits include:
- Accelerated access to their own DGX-1 in the cloud – with dedicated access to a full node including eight Tesla P100 GPUs, paired with NVIDIA’s container-based, deep learning software stack.
- The economics and elasticity of cloud – at $29.50/hour, customers can untether themselves from infrastructure operations, with the freedom to scale up or down as their business dictates.
- The simplicity of DGX-1 “built-in” – enabling customers to reap the benefits of DGX-1 system’s GPU-optimized stack, including containers, popular deep learning frameworks, drivers and libraries — all performance-engineered by NVIDIA.
Learn more about the Nimbix Cloud built on DGX-1 in this solution brief.