Search NASA⌕ Search

SEARCH · Search NASA

Results for “Field-Programmable Gate Arrays”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

169 records · Page 10

Distance-Weighted Graph Neural Networks on FPGAs for Real-Time Particle Reconstruction in High Energy Physics

Graph neural networks have been shown to achieve excellent performance for several crucial tasks in particle physics, such as charged particle tracking, jet tagging, and clustering. An important domain for the application of these networks is the FGPA-based first layer of real-time data filtering at the CERN Large Hadron Collider, which has strict latency and resource constraints. We discuss how to design distance-weighted graph networks that can be executed with a latency of less than one μs on an FPGA. To do so, we consider a representative task associated to particle reconstruction and identification in a next-generation calorimeter operating at a particle collider. We use a graph network architecture developed for such purposes, and apply additional simplifications to match the computing constraints of Level-1 trigger systems, including weight quantization. Using the hls4ml library, we convert the compressed models into firmware to be implemented on an FPGA. Performance of the synthesized models is presented both in terms of inference accuracy and resource usage.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Accelerating Genome Sequencing 100-1000X with FPGAs

The performance of FPGAs on Cray XD1 and SGI/RASC systems (with Virtex-II Pro 50 and Virtex-4 LX160 FPGAs) was evaluated for human genome sequencing using FASTA1 and BLAST2, respectively. Scalable speedups of 100X for a Virtex-4 FPGA (and exceeding 1000X for multiple FPGAs) over a 2.2 GHz Opteron were achieved. Similar FPGA speedups were achieved using BLAST on a SGI/RASC system at Oak Ridge National Laboratory.

42 ENGINEERING↗

High-Performance Mixed-Precision Linear Solver for FPGAs

Compared to higher-precision data formats, lower-precision data formats result in higher performance for computational intensive applications on FPGAs because of their lower resource cost, reduced memory bandwidth requirements, and higher circuit frequency. On the other hand, scientific computations usually demand highly accurate solutions. This paper seeks to utilize lower-precision data formats whenever possible for higher performance without losing the accuracy of higher-precision data formats by using mixed-precision algorithms and architectures. First, we analyze the floating-point performance of different data formats on FPGAs. Second, we introduce mixed-precision iterative refinement algorithms for linear solvers and give error analysis. Finally, we propose an innovative architecture for a mixed-precision direct solver for reconfigurable computing. Our results show that our mixed-precision algorithm and architecture significantly improve the performance of linear solvers on FPGAs.

97 MATHEMATICS AND COMPUTING↗

Experiences on 64 and 150 FPGA Systems

Four FPGA systems were evaluated: the Cray XD1 system with 6 FPGAs at ORNL and Cray, the Cray XD1 system with 150 FPGAs at NRL* and the 64 FPGAs on Edinburgh s Maxwell . Their hardware and software architectures, programming tools and performance on scientific applications are discussed. FPGA speedup (over a 2.2 GHz Opteron) of 10X was typical for matrix equation solution, molecular dynamics and weather/climate codes and upto 100X for human genome DNA sequencing. Large genome comparisons requiring 12.5 years for an Opteron took less than 24 hours on NRL s Cray XD1 with 150 Virtex FPGAs for a 7,350X speedup. pipeline so each query and database character are compared in parallel, resulting in a table of scores. Genome Sequencing Results: FPGA timing results (for up to 150 FPGAs) were obtained and compared with up to 150 Opterons for sequences of varying size and complexity (e.g. 4GB openfpga.org human DNA benchmark and 155M human vs. 166M mouse DNA). 1 FPGA: Bacillus_anthracis DNA compare: Genomes

59 BASIC BIOLOGICAL SCIENCES↗

Computational Speed-Up of Complex Durability Analysis of Large-Scale Composite Structures

The analysis of modern structures for aerospace, infrastructure, and automotive engineering applications necessitates the use of larger and larger computational models for accurate prediction of structural response. The ever-increasing size of computational structural mechanics simulations imposes a pressing need for commensurate increases in computational speed to keep costs and computation times in check. Innovative methods are needed to expedite the numerical analysis of complex structures while minimizing computational costs. The need for these methodologies is even more critical when performing durability and damage tolerance evaluation as the computation is repeated a number of times for various loading conditions. This paper describes a breakthrough for efficient and accurate predictive methodologies that are amenable to the analysis of progressive failure, reliability, and optimization of large-scale composite structures or structural components.

42 ENGINEERING↗

Compiling Quantum Circuits for Dynamically Field-Programmable Neutral Atoms Array Processors

Dynamically field-programmable qubit arrays (DPQA) have recently emerged as a promising platform for quantum information processing. In DPQA, atomic qubits are selectively loaded into arrays of optical traps that can be reconfigured during the computation itself. Leveraging qubit transport and parallel, entangling quantum operations, different pairs of qubits, even those initially far away, can be entangled at different stages of the quantum program execution. Such reconfigurability and non-local connectivity present new challenges for compilation, especially in the layout synthesis step which places and routes the qubits and schedules the gates. In this paper, we consider a DPQA architecture that contains multiple arrays and supports 2D array movements, representing cutting-edge experimental platforms. Within this architecture, we discretize the state space and formulate layout synthesis as a satisfiability modulo theories problem, which can be solved by existing solvers optimally in terms of circuit depth. For a set of benchmark circuits generated by random graphs with complex connectivities, our compiler OLSQ-DPQA reduces the number of two-qubit entangling gates on small problem instances by 1.7x compared to optimal compilation results on a fixed planar architecture. To further improve scalability and practicality of the method, we introduce a greedy heuristic inspired by the iterative peeling approach in classical integrated circuit routing. Using a hybrid approach that combined the greedy and optimal methods, we demonstrate that our DPQA-based compiled circuits feature reduced scaling overhead compared to a grid fixed architecture, resulting in 5.1X less two-qubit gates for 90 qubit quantum circuits. These methods enable programmable, complex quantum circuits with neutral atom quantum computers, as well as informing both future compilers and future hardware choices.

Physics↗