LANL Solicits Bids for 18 MW Crossroads Supercomputer for Delivery in 2021

Print Friendly, PDF & Email

The next big supercomputer is out for bid. An RFP was released today for Crossroads, an 18 Megawatt system that will support the nation’s Stockpile Stewardship Program. The RFP is a joint effort of the New Mexico Alliance for Computing at Extreme Scale (ACES), a collaboration between Los Alamos and Sandia National Laboratories.

Los Alamos National Laboratory is proud to serve as the home of Crossroads. This high-performance computer will continue the Laboratory’s tradition of deploying unique capabilities to achieve our mission of national security science,” said Thom Mason, director of Los Alamos National Laboratory.

Responses to the RFP are due on March 18, 2019. After evaluation of proposals, the goal is to award the Crossroads contract to the selected vendor by the end of calendar year 2019. The Crossroads platform is projected to be installed at Los Alamos by the fall of 2021.

In 2021, the first AT system, Trinity, will be nearing the end of its useful lifetime. Crossroads is a Trinity replacement, a tri-lab computing resource for existing simulation codes, and a larger resource for ever-increasing computing requirements to support the weapons program. Crossroads will provide a large portion of the AT system resources for the NNSA ASC tri-lab simulation community: Lawrence Livermore National Laboratory (LLNL), Los Alamos National Laboratory (LANL), and Sandia National Laboratories (SNL), during the FY22-26 timeframe.

One of the primary goals of the Crossroads system is to improve efficiency in three key areas: application performance, workflow, and application development. Performance efficiency is the achieved performance of the application once ported to the proposed platform. Workflow efficiency is the efficiency that a complete NNSA workflow executes on the proposed platform. Development efficiency is the ease in which NNSA mission codes can be transitioned to the proposed architecture with minimal modifications.

The focus on efficiency is to optimize the complete weapons simulation life-cycle, rather than focusing on a single aspect such as system cost or power usage. The Crossroads system will support the highest fidelity of next-generation weapons simulations and will meet NNSA Defense Programs’ mission needs. It will also allow applications to explore and exploit upcoming technologies.

Since the era of the Manhattan project, it has been essential to underwrite the performance of nuclear weapons using the most powerful computing available. This is particularly true since the cessation of testing,” said Bob Webster, the deputy laboratory director of the Weapons program at Los Alamos National Laboratory.

Los Alamos, Sandia, and Lawrence Livermore national laboratories will use Crossroads to run some of the largest and most demanding simulations for Stockpile Stewardship with the goal of assuring the safety, security and effectiveness of the United States nuclear stockpile.

As part of NNSA’s mission to assure the health of the U.S. nuclear stockpile, the Advanced Simulation and Computing (ASC) Program continues to provide NNSA with leading-edge, high-end computing and simulation capabilities so that NNSA meets nuclear weapons assessment, certification, and qualification requirements.

In addition to meeting NNSA Defense Programs’ mission needs, the architecture of Crossroads will keep ASC moving forward by allowing applications to fully explore and exploit upcoming technologies. These new technologies include processors, memory, input/output (I/O), interconnect, power efficiency, and reliability. Some of the largest and most demanding capability-class jobs will be able to run on Crossroads. These capabilities are required for supporting stockpile stewardship certification and assessments to ensure that the Nation’s nuclear stockpile is safe, secure, and reliable.

Crossroads will establish a validated predictive capability for key physical phenomena, quantifying and aggregating uncertainties in simulation tools, and providing mission-responsive computational environments. Simulation fidelity improvements include:

  • Physics fidelity—replacing calibrations with physics models to improve confidence of extrapolation away from the nuclear test base;
  • Numerical fidelity—Significant Finding Investigations (SFIs) are sensitive to fine scale 3D features, requiring much finer resolution; and
  • Geometric fidelity—neglected features need to be included; many are inherently 3D.

Crossroads will continue to push the cutting edge of what is possible on current systems with its’ 3D simulations capabilities. By 2021, the goal of running high-resolution 3D simulations with full physics and geometric features will be attainable. Ultimately, Crossroads will accelerate industry’s R&D technology roadmap for the ASC program’s future exascale-class computational needs.

In related news, catch our recent interview with LANL’s Gary Grider on the new Efficient Mission-Centric Computing Consortium.

Sign up for our insideHPC Newsletter