Utility Company Bolsters the Employee Experience with GPU-Based Computing

Challenge

Identifying the right technology sets to empower your workforce is challenging for IT departments. For engineering teams in the utility industry, strategizing long-term grid investments and developing 3D models for new smart grid-enabled substations is complicated enough in an office. Now, these activities must happen wherever employees work. Whether at home or in the field, planning teams need access to vast data sets and computing power to execute grid planning initiatives that significantly impact a utility's ability to optimize at the macro and micro levels.

For a Midwest-based electric and gas utility company serving more than three million customers, COVID-19 placed their critical system design team away from corporate offices and the infrastructure capable of quickly accessing and analyzing vast data sets. Working from home, these teams were stuck downloading huge files over home networks and setting up queries to run over the weekend, hoping they would not crash. These challenges limited their analysis depth, negatively impacted the digital employee experience, and hindered overall grid optimization.  

Solution

Frustrated by long query times and buffering image analytics software, the customer's grid engineering team initially requested Graphics Processing Unit (GPU)-enabled laptops from their IT department for $15,000 to $25,000 each. When IT reached out to WWT, our account team advised that the requested hardware would not alleviate collaboration challenges associated with transferring large files over the users' home networks, and our team was confident we could find a more purpose-driven, cost-effective and holistic solution. 

We had recently upgraded the utility company's Virtual Desktop Infrastructure (VDI) environment to Omnissa's Workspace ONE, formerly VMware. With this in mind, we suggested a virtualized solution that addressed their network and compute-related pain points. Working with our strategic partner, NVIDIA, we proposed establishing a GPU-enabled environment within the customer's Workspace ONE backend dedicated to their engineering team. Our goal was to eliminate data leaving the data center and provide a more scalable and cost-effective compute solution aligned with the customer's goals. 

While the customer had initial skepticism and concerns about simplicity, we leveraged our Advanced Technology Center (ATC) to create a fully functioning lab environment for comprehensive testing and solution validation. Working with the client's engineering team, within weeks we validated:

• Baseline performance speeds for pre-determined, complex analyses covering civil, physical and electrical engineering users through a GPU profiler.
• Baseline performance surveys for current state application performance.
• A centralized GPU-based compute architecture and leveraged relevant software for civil, physical and electrical analyses.
• Comparable test performance results for future baselining, including qualitative and quantitative KPIs.

Outcomes and benefits

After a week into the two-week testing period, the customer saw a significant improvement in their employee experience. The customer's engineering team requested additional testing and solution refinement that yielded: 

• Shortened query times for grid demand planning efforts that, in many instances, cut query times from days to minutes.
• Continued improvements in the employee experience, which enabled planners to conduct more comprehensive and in-depth analyses, leading to enhancements in grid planning quality.
• Streamlined 3D computing and planning tools for the customer's grid team.
• Eliminated lengthy file transfers over workers' home networks, simplified workflows and reduced instances of users making conflicting changes to datasets.

Since the initial testing, we've continued to partner with the customer to move forward with installing and configuring GPU-enabled hosts into their on-premises environment with professional enablement and training. This solution improves the employee experience, enables more thorough analysis by the engineering team, and strengthens collaboration between engineering team members regardless of location, ultimately leading to a better-engineered, more efficient, safer, and more affordable and predictable grid. By empowering their engineering teams, this utility provider can continue supporting the needs of their customers.