Cisco IT designed AI-ready infrastructure with Cisco compute, best-in-class NVIDIA GPUs, and Cisco networking that helps AI mannequin coaching and inferencing throughout dozens of use instances for Cisco product and engineering groups.
It’s no secret that the stress to implement AI throughout the enterprise presents challenges for IT groups. It challenges us to deploy new expertise sooner than ever earlier than and rethink how knowledge facilities are constructed to satisfy rising calls for throughout compute, networking, and storage. Whereas the tempo of innovation and enterprise development is exhilarating, it could possibly additionally really feel daunting.
How do you shortly construct the information middle infrastructure wanted to energy AI workloads and sustain with important enterprise wants? That is precisely what our workforce, Cisco IT, was dealing with.
The ask from the enterprise
We had been approached by a product workforce that wanted a option to run AI workloads which can be used to develop and check new AI capabilities for Cisco merchandise. It would ultimately assist mannequin coaching and inferencing for a number of groups and dozens of use instances throughout the enterprise. And they wanted it performed shortly. want for the product groups to get improvements to our prospects as shortly as potential, we needed to ship the new surroundings in simply three months.
The expertise necessities
We started by mapping out the necessities for the brand new AI infrastructure. A non-blocking, lossless community was important with the AI compute material to make sure dependable, predictable, and high-performance knowledge transmission throughout the AI cluster. Ethernet was the first-class alternative. Different necessities included:
- Clever buffering, low latency: Like all good knowledge middle, these are important for sustaining clean knowledge movement and minimizing delays, in addition to enhancing the responsiveness of the AI material.
- Dynamic congestion avoidance for numerous workloads: AI workloads can differ considerably of their calls for on community and compute assets. Dynamic congestion avoidance would make sure that assets had been allotted effectively, forestall efficiency degradation throughout peak utilization, keep constant service ranges, and forestall bottlenecks that would disrupt operations.
- Devoted front-end and back-end networks, non-blocking material: With a aim to construct scalable infrastructure, a non-blocking material would guarantee adequate bandwidth for knowledge to movement freely, in addition to allow a high-speed knowledge switch — which is essential for dealing with massive knowledge volumes typical with AI functions. By segregating our front-end and back-end networks, we might improve safety, efficiency, and reliability.
- Automation for Day 0 to Day 2 operations: From the day we deployed, configured, and tackled ongoing administration, we needed to cut back any handbook intervention to maintain processes fast and decrease human error.
- Telemetry and visibility: Collectively, these capabilities would offer insights into system efficiency and well being, which might permit for proactive administration and troubleshooting.
The plan – with a number of challenges to beat
With the necessities in place, we started determining the place the cluster may very well be constructed. The prevailing knowledge middle services weren’t designed to assist AI workloads. We knew that constructing from scratch with a full knowledge middle refresh would take 18-24 months – which was not an choice. We wanted to ship an operational AI infrastructure in a matter of weeks, so we leveraged an present facility with minor adjustments to cabling and system distribution to accommodate.
Our subsequent considerations had been across the knowledge getting used to coach fashions. Since a few of that knowledge wouldn’t be saved regionally in the identical facility as our AI infrastructure, we determined to duplicate knowledge from different knowledge facilities into our AI infrastructure storage techniques to keep away from efficiency points associated to community latency. Our community workforce had to make sure adequate community capability to deal with this knowledge replication into the AI infrastructure.
Now, attending to the precise infrastructure. We designed the guts of the AI infrastructure with Cisco compute, best-in-class GPUs from NVIDIA, and Cisco networking. On the networking facet, we constructed a front-end ethernet community and back-end lossless ethernet community. With this mannequin, we had been assured that we might shortly deploy superior AI capabilities in any surroundings and proceed so as to add them as we introduced extra services on-line.
Merchandise:
Supporting a rising surroundings
After making the preliminary infrastructure obtainable, the enterprise added extra use instances every week and we added extra AI clusters to assist them. We wanted a option to make all of it simpler to handle, together with managing the swap configurations and monitoring for packet loss. We used Cisco Nexus Dashboard, which dramatically streamlined operations and ensured we might develop and scale for the longer term. We had been already utilizing it in different components of our knowledge middle operations, so it was simple to increase it to our AI infrastructure and didn’t require the workforce to study an extra software.
The outcomes
Our workforce was capable of transfer quick and overcome a number of hurdles in designing the answer. We had been capable of design and deploy the backend of the AI material in beneath three hours and deploy all the AI cluster and materials in 3 months, which was 80% sooner than the choice rebuild.
As we speak, the surroundings helps greater than 25 use instances throughout the enterprise, with extra added every week. This contains:
- Webex Audio: Enhancing codec growth for noise cancellation and decrease bandwidth knowledge prediction
- Webex Video: Mannequin coaching for background substitute, gesture recognition, and face landmarks
- Customized LLM coaching for cybersecurity merchandise and capabilities
Not solely had been we capable of assist the wants of the enterprise right this moment, however we’re designing how our knowledge facilities have to evolve for the longer term. We’re actively constructing out extra clusters and can share extra particulars on our journey in future blogs. The modularity and adaptability of Cisco’s networking, compute, and safety provides us confidence that we are able to preserve scaling with the enterprise.
Extra assets:
Share:
