ASCR Highlights

Researchers demonstrated a new quantum error mitigation approach by simulating the evolution of a chain of six spins (top). The mitigated results are much closer to the exact results than the original measured data from the quantum computer (bottom).

May 4, 2022

New Error Mitigation Approach helps Quantum Computers Level Up

Noise estimation circuits, in conjunction with other error mitigation methods, produce reliable results for quantum computer-based materials simulations.

Quantum equipment in the "Alice” laboratory, where the researchers stored the photon source and the first node in the three-node quantum network.

April 21, 2022

New Quantum Network Shares Information at a Scale Practical for Future Real-World Applications

Researchers enable real-time adjustments to communication among three remote nodes in a quantum network.

Snapshot of the simulated landfall of an atmospheric river along the west coast of North America on February 11, 2020. Grey tones depict water vapor. Colors indicate precipitation intensity from blue (light rain) to green (very strong precipitation).

April 7, 2022

Better Clouds than Ever with New Exascale Computing-Ready Atmosphere Model

Scientists demonstrate the value of a new global atmosphere model for the Energy Exascale Earth System Model.

A silicon quantum processor in which an electron spin qubit (blue) enables communication between two phosphorus nuclear spin qubits (red). Researchers used gate set tomography to show that the processor’s logic gates surpass 99 percent fidelity.

March 24, 2022

Probing the Inner Workings of High-Fidelity Quantum Processors

Scientists use gate set tomography to discover and validate a silicon qubit breakthrough.

Left, an overview of the new evolutionary approach. Right, comparison of the new approach with another evolutionary approach to SNNs (EONS) showing the new approach reaches comparable performance to EONS, but does so more gradually, which can indicate better generalization ability of the resulting SNN.

April 2, 2021

Evolution Sets the Stage for More Powerful Spiking Neural Networks

Indirect encoding of neural network components creates a pathway to complex applications for neural networks.

Overview of the 3D Coded SUMMA process showing the role of matrix multiplication.

March 30, 2021

New Approach to Fault Tolerance Means More Efficient High-Performance Computers

3D Coded SUMMA replaces traditional fault tolerances methods with coded computation-based matrix multiplication.

Visualization of how the research team’s multitask convolutional neural network classifies primary cancer sites.

August 11, 2020

“Multitasking” AI Tool Extracts Cancer Data in Record Time

A designed coiled-coil heterodimer, with halves colored green and purple. A DNA double-helix (gold), for which particular sequences have the same property of forming specific pairs, is superimposed to show scale.

January 30, 2020

Breaking Through Computational Barriers to Create Designer Proteins

Using advanced computing, scientists designed protein pairs that perfectly complement each other.

The illustrations show how the correlation between lattice distortion and proton binding energy in a material affects proton conduction in different environments.

January 30, 2020

Simulations Identify Importance of Atomic-Level Distortions in Certain Fuel Cell Materials

Researchers determine how to design better materials for energy storage.