One of the biggest takeaways to come out of the GTC 2025 conference, to use a statement from NVIDIA CEO Jensen Huang’s keynote: “energy is our most important commodity”.

This resounding message continued throughout the keynote as Huang unveiled new GPU and data center hardware alongside their energy power efficiency stats, specifically throughput of tokens per megawatt. As high performance computing (HPC) and AI workloads demand more power, the energy sector is rapidly evolving to expand production by investing in new sources, methods, and process improvements. As more enterprises depend on hyperscale computing in the cloud, both industry’s end-users and technology leaders are taking a close look at how they can harness new technologies to control rising energy costs while improving the performance of their applications. 

In the GTC session track focused on CAE simulation and design, SLB (FKA Schlumberger) and Rescale’s presentation showcased how advanced simulation technologies, powered by NVIDIA hardware and software, are revolutionizing the energy industry. Rescale’s CEO Joris Poort and SLB’s Simulation Technical Advisor Gocha Chochua led a session on how new GPU hardware combined with specialized HPC and AI software are driving significant gains, in fact, orders of magnitude improvements in speed and cost efficiency for critical engineering analysis. 

Exploring New Technologies for Energy Simulation

Engineering teams in the energy sector are focused on delivering faster, more efficient R&D results using the latest computing methods and technologies. Traditional methods, often constrained by manual and siloed workflows, are no longer sufficient to meet the growing demands. SLB, a global leader in energy technology, has embraced advanced simulation technologies to address these challenges head-on. By leveraging coupled computational fluid dynamics (CFD) and the discrete element modeling (DEM) optimized GPU hardware, SLB accelerates solutions to critical engineering challenges such as scale deposition in downhole equipment, which can severely disrupt oil recovery processes. These coupled or hybrid Eulerian-Lagrangian simulations are extremely computationally and time intensive, because both fluid flow and individual particle interactions are explicitly calculated and influence each other. Getting timely and accurate results from these simulations can make the difference of winning sizable contracts, shaving months off of deployments, and ensuring reliable off-shore operations. 

SLB’s Innovative Approach to Computer Aided Engineering

SLB’s innovative approach involves detailed modeling of subsurface conditions and particle behavior, enabled by NVIDIA GPUs on Rescale’s cloud platform. This combination allows for high-fidelity simulations that enhance oil well productivity, optimize production strategies, and reduce maintenance costs. Particularly in high-salinity environments like Brazil’s pre-salt fields, these simulations set a new standard for efficiency and risk mitigation in hydrocarbon production.

Key Technologies and Use Cases

• GPU Acceleration for CAE: SLB’s use of GPU acceleration has led to significant improvements in simulation speed and efficiency. For instance, coupled CFD-DEM simulations of downhole valves, which previously took 180 days on in-house machines, can now be completed in a fraction of the time using NVIDIA GPUs on Rescale’s cloud-based HPC platform. This not only accelerates the qualification of equipment but also reduces costs and greenhouse gas emissions.
  Joris Poort, Founder & CEO of Rescale, emphasized the importance of GPU acceleration: “The speed up was found to be more than 10 times faster with 20 times cost reduction which is amazing. And then at the same time, can you imagine running 180 days on a computer, how much would it cost not to mention the downtime spent waiting?”

• AI-Enhanced Simulation: AI surrogate models and reduced order models (ROM) are driving faster simulation results, enabling more efficient and sustainable energy production. These models can be deployed on the edge, providing rapid field decision-making capabilities. For example, SLB’s use of ROMs methane leak detection in field operations have achieved a 3,600× speedup, allowing for real-time visualization and early detection of leaks.
  
  SLB’s Chochua highlighted the impact of AI-enhanced simulations: “We can use reduced order models to run numerous simulations which we need to solve this inverse problem of using sensor data to determine the leak’s location. Even a basic computational fluid dynamics simulation on a coarse mesh would take about an hour. This makes it impractical to deploy a computer at the edge in the field for real-time predictions.”

• Real-Time 3D Visualization: NVIDIA Omniverse enables real-time 3D visualization of simulation results, facilitating improved collaboration and decision-making. This technology allows engineers to visualize complex phenomena, such as methane plume simulations, in real-time, enhancing their ability to predict and mitigate emissions.

Chochua explained the benefits of real-time visualization: “Having accurate, real-time visualization is incredibly beneficial. For instance, in this scenario, you can see an offshore platform, the wind direction, and a potential leak source. The resulting plume shape is quite unexpected and wouldn’t be predictable without simulations.”

Rescale and SLB’s Collaboration and Results

SLB’s success with Rescale’s platform is evident in several key projects:

• Downhole Flow Control Valves: SLB’s simulations of valve performance in harsh deepwater conditions have led to significant improvements in equipment qualification and production efficiency. The use of GPU-accelerated simulations has reduced compute time from 180 days to just 18 days, with a 30× reduction in costs.

Chochua shared the impact of this project: “We faced a challenge in offshore Brazil with a pre-salt formation and very high flow rates. The two main issues were erosion and scale deposition. While erosion is a more common problem that we have experience with, this particular case lacked any testing facilities worldwide capable of handling such conditions. Therefore, we relied on advanced coupled CFD-DEM simulations to find a solution.”

• Methane Emissions Solutions: SLB’s deployment of AI-enhanced simulations for methane leak detection has enabled rapid field decision-making and reduced greenhouse gas emissions. The use of reduced order models allows for real-time predictions and visualizations, enhancing the effectiveness of emissions mitigation strategies.
  
  Chochua explained the significance of this solution: “Methane is a significantly more potent greenhouse gas than CO2 over a 20-year period, contributing about 30% to the overall greenhouse gas effect from methane emissions. When a sensor detects a leak, the data is sent to a solar tower. The crucial question then becomes: how do we pinpoint the leak’s location? To address this, we’re exploring plume transport models, like the reduced order model shown here.”
• Geology-Informed Subsurface Simulation: SLB’s use of surrogate modeling for subsurface simulations has achieved a 10× speedup, enabling more accurate predictions of subsurface properties and accelerating exploration and production.
  
  Chochua described the benefits of this approach: “In subsurface analysis, properties like porosity and permeability aren’t always precisely defined. While direct measurement can be challenging, we have historical production data, such as water and gas cut. The process involves history matching – adjusting formation parameters in our models until they closely align with the observed historical data.”

Rescale’s Role in Empowering Engineering Innovators

Rescale’s cloud-based platform plays a crucial role in empowering energy innovators like SLB. By providing scalable, agile, and cost-efficient HPC solutions, Rescale enables SLB to overcome hardware limitations and focus on innovation. The platform’s ability to orchestrate compute, manage data, and develop domain-specific AI in a complementary loop unlocks new methods that accelerate product breakthroughs.

Chochua emphasized the importance of Rescale’s platform: “Our partnership with Rescale began just last year. Their platform is software-agnostic, meaning it’s not tied to any specific code or vendor, which facilitates collaboration. Since last year, we’ve effectively eliminated hardware limitations. Previously, we constantly had to ask if we had enough computing power for a simulation. Now, it feels like we have unlimited access to the resources we need.”

The Virtuous Cycle of Computing and Energy Innovation

The collaboration between SLB and Rescale exemplifies a virtuous cycle of innovation. Cloud-based computing on Rescale supports SLB’s engineering efforts, which in turn power energy production to meet global demand. This cycle of innovation is further enhanced by GPU optimization, which drives performance and cost improvements, and AI surrogate models, which enable faster and more efficient simulations.

Joris Poort highlighted the importance of this cycle: “As we develop new, better, more sustainable, more reliable energy sources, we’re able to fuel that into the way we power data centers. Data centers themselves are being built at a very rapid pace. And again, the rate-limiting factor here for engineers historically is being compute. And compute is rate-limited by energy.”

Rescale Impact on Energy Innovation

SLB’s transformative use of advanced simulation technologies, powered by NVIDIA GPUs on Rescale’s cloud platform, is setting new standards for efficiency and risk mitigation in the energy sector. By embracing GPU acceleration, AI-enhanced methods, and real-time 3D visualization, SLB is driving innovation and sustainability in energy production. Rescale’s enabling technologies are empowering energy innovators to overcome challenges and unlock new methods that enrich humanity. Watch the full presentation from NVIDIA GTC for the full story. 

As we look to the future, the collaboration between SLB, Rescale, and NVIDIA will continue to drive advancements in energy technology, paving the way for more sustainable and efficient energy solutions.

If you are interested in the technologies and methods explored in this blog, Rescale’s simulation and HPC/AI experts look forward to connecting with you. Schedule time with us to learn more and estimate some of the benefits for your organization.