High Performance Computing to Optimize an Induced Flow Power Generator Device for Waste Heat Recovery Applications in Data Centers

Spar Energy LLC

Montvale, NJ

Spar Energy LLC will partner with Oak Ridge National Laboratory to improve the design and foster adoption of low-quality heat recovery systems for data centers. If deployed nationwide, this technology could save 0.5 Quads/yr of electricity and reduce CO2 emissions by 75 million tonnes/yr.


Clean, Dispatchable and Affordable CHP using a Novel Argon Power Cycle

Noble Thermodynamic Systems Inc.

Alameda, CA

Noble Thermodynamic Systems Inc. and Argonne National Laboratory will collaborate to improve the design of natural gas CHP system utilizing an Argon power cycle and broaden fuel usage to a 50/50 NG/H2. This could save 0.4 TBtu over 20 years for one plant alone, and many times that if widely adopted.  This technology captures 100% of the CO2 emissions.


AI-Driven Accelerated Inclusion Analysis for Energy Efficient Steelmaking


East Chicago, IN

ArcelorMittal and Lawrence Livermore National Laboratory will work to couple machine learning methods with HPC resources to reduce defects from inclusions in steel manufacturing. The iron and steel industry consumes an estimated 6% (~1470 PJ) of the energy used by the U.S. manufacturing sector. This technology could reduce this number by 1-2% and reduce CO2 emission by ~1.5 million tonnes/yr.


Adsorptive CO2 Removal from Dilute Sources (ACO2RDS)

Total Energies E&P Research & Technology USA LLC

Houston, TX

Total Energies E&P Research & Technology USA LLC and Argonne National Laboratory will use HPC atomistic modeling, process modeling and machine learning to discover novel metal oxide framework adsorbents for CO2 capture from dilute sources. CO2 emissions could be reduced by up to 10 million tonnes/yr.


Combined Electromagnetic and Micromagnetic Simulation of Integrated Magnetic Devices for Improved Integrated Circuit Energy Efficiency

Ferric Inc.

New York, NY

Ferric will continue their collaboration in a follow-on project with Lawrence Berkeley National Laboratory to improve DC-DC voltage regulator energy efficiency and improve manufacturing. This technology could save 50 billion KWh electrical energy per year and reduce CO2 emissions by 20 million tonnes/yr.


Multi-Physics Modeling of Carbon Fiber Oxidation to Enhance Process Energy Efficiency

Collaborative Composite Solutions Corp.

Knoxville, TN

In partnership with Oak Ridge National Laboratory, Collaborative Composite Solutions Corp. will utilize HPC to model oxidation to help increase carbon fiber production rate and energy efficiency by at least 10%.


Effect of Processing on Precipitation Kinetics in Nano Polycrystalline 7075 Aluminum Alloy

Fairmount Technologies LLC.

Wichita, KS

Fairmount Technologies LLC. and Pacific Northwest National Laboratory will use HPC to advance the yield strength of AA7075 by 20% (compared to commercially available 7075-T6 temper sheet) by optimizing thermomechanical processing. Lighter weight cars made using this material could save 6 million gallons of gas per year and reduce CO2 emissions by 4 million tonnes/yr.


Advanced HPC Thermal Simulations for Wafer-Scale Diamond Heat Spreaders

Diamond Foundry

South San Francisco, CA

With the computing expertise of Lawrence Berkeley National Laboratory, Diamond Foundry will optimize high-fidelity thermal modeling of computer chips using diamond heat spreaders. This technology could save about 0.03 Quads of electricity per year and reduce CO2 emissions by about 1 million tonnes/yr.