The global energy company Santos, with operations across Australia, Papua New Guinea, Timor-Leste and the United States, supplies oil and liquid natural gas while keeping its eye on the future of renewables and lower-carbon fuels. Headquartered in Adelaide, with key locations in Brisbane and Perth, Santos has powered residential, commercial, and industrial customers for more than 70 years.
In addition to being an innovator in energy, Santos was an innovator and early adopter of cloud computing for its benefits in the energy exploration and production (E&P) sector. Santos is a longtime user of Amazon Web Services and is hosting remote workstations in AWS for its geoscientists. Their Microsoft Services Provider License Agreement was also coming to an end, giving Santos a timely opportunity to go all-in with AWS and get even more value from their Amazon relationship.
Darren Stanton, Exploration Service Delivery Lead for Santos, was working on transitioning more than 300 users to WorkSpaces Core Managed Instances to deliver Microsoft 365 applications and Amazon’s own DCV remote protocol, which does not require purchasing a license to use in Amazon EC2 or WorkSpaces Core Managed Instances (CMI). With WorkSpaces CMI, organizations have the flexibility to host virtual desktop infrastructure in the cloud, including Amazon EC2, while DCV offers a secure, performant remote display and streaming protocol that’s capable of keeping pace even with high-performance, low-latency applications common in scientific computing.
“With users scattered around the countryside with thousands of kilometers in between, there’s obviously no thought of having workstations under desks anymore,” said Stanton. “It’s about being able to connect by a thin client to a centralized location so that our scientists are as close to the data as possibly we can get them.”
With these upgrades, the only missing piece was an equally secure, performant orchestration layer to manage and provision the remote desktop/workstation environment, such as end-user connections, assigning resources, and scaling up or down as needed. Leostream’s powerful Remote Desktop Access Platform simplifies administration and helps optimize cloud costs while ensuring virtual/remote desktops and workstations take full advantage of AWS’ infrastructure power, including GPU processors for heavy compute loads.
For Santos, those workloads and resource demands can vary from one morning to another, or one moment to another. Some geoscientists running simulations may use their machines 24/7, while others may connect only two to three times a week to check their seismic output. With the Leostream platform, Stanton can hard-assign GPU-enabled compute to those always-on users, and place those connecting less frequently into a shared pool, rather than provisioning individual GPU resources for each user. A shared pool and improved load management has proven to be far more efficient, and far more economical. With Leostream’s ability to power down GPU instances when they’re idle, instead of leaving the meter running on unused cloud GPUs, Santos can save even more.
Stanton also used the Leostream platform to create pools tailored to different departments. A “Geo pool” makes 50 machines available to those using subsurface E&P applications like Petrel, while geographic information system have a “GIS pool” offering machines for GIS-specific workflows like ArcGIS and MapInfo. Santos’ use of pools makes it simpler to meet the requirements of different users, track resource usage, and manage performance.
Another selling point was the Leostream platform’s ability to work with Santos’ preferred Virtual Network Computing (VNC) protocol, TurboVNC, which is capable of the substantial scale required by 3D applications. Part of Santos’ commercial edge is the technical skill of its geoscientists, and many prefer to run applications from Linux command prompts, write their own scripts, or employ Python and run applications from there. Leostream’s vendor- and protocol-agnostic features support VNC connections as well as more typical connections like PCoIP, and since Stanton can easily set up another pool in Leostream, he can assign different resources to those writing and running their own code to segregate those user requirements.
The combination of AWS WorkSpaces CMI and Leostream has helped ensure required performance by keeping data close to the users who need it. Santos, like the rest of its industry, relies on massive data loads: four and a half petabytes, a volume that continues to grow, with file sizes measured in terabytes rather than megabytes and gigabytes.
“Data and proximity to data is absolutely critical for geoscience applications to work successfully, so on AWS we can co-locate the desktop or workstation with the workload,” said Andrew DeFoe, Principal Product Manager, AWS.
WorkSpaces CMI offers enterprise features and capabilities with state-of-the-art flexibility and availability, and with the Leostream platform, Santos gains additional power and options for their virtual desktops/workstations. The integration allows AWS customers to deliver multiple OS flavors, GPU-accelerated workloads, multi-user sessions, and more to best suit their users — all within the AWS customer’s account. WorkSpaces CMI fully supports Microsoft 365 Apps for enterprise, and Leostream ensures end users have a reliable, consistent, simplified experience accessing and using 365 apps on top of AWS infrastructure. The Leostream platform’s deep integration with Amazon DCV delivers secure, responsive access for users of demanding, graphics-intense applications for low-latency experiences including high-performance GPU processing.