Adaptive Bulk Search: Solving Quadratic Unconstrained Binary Optimization Problems on Multiple GPUs

Yasudo, Ryota; Nakano, Koji; Ito, Yasuaki; Tatekawa, Masaru; Katsuki, Ryota; Yazane, Takashi; Inaba, Yoko

doi:10.1145/3404397.3404423

Cited by 12 publications

(12 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, since this is not realistic, the actual search rate is likely to be much lower. On the other hand, the search rate of the GPU system 7 is approximately 1.5 times higher than that of the proposed FPGA implementation. Since the utilized devices differ, it is difficult to directly compare FPGA's performance and the GPU.…”

Section: Ta B L E 1 the Relation Between D I And U I Whenmentioning

confidence: 83%

“…D‐Wave Systems developed the first commercial system that solves the QUBO problem by quantum annealing 4,5 . In fact, the system with quantum annealing is used for solving Ising models , but we can convert instances of the QUBO and the Ising model to each other easily 6,7 . However, since the present system has a limitation of the size of inputs with up to 124 bits, 5 we cannot use the system to the practical size of problems.…”

Section: Introductionmentioning

confidence: 99%

“…4,5 In fact, the system with quantum annealing is used for solving Ising models, but we can convert instances of the QUBO and the Ising model to each other easily. 6,7 However, since the present system has a limitation of the size of inputs with up to 124 bits, 5 we cannot use the system to the practical size of problems. Goto et al show a simulated bifurcation that models a combinatorial optimization problem with a magnetic material model for solving the Ising model.…”

mentioning

confidence: 99%

“…10 The FPGA implementation can solve a 1024-bit QUBO problem with a 20.4G search rate, where the search rate is a metric of the throughput. 1,7,10 It shows the number of solutions searched by a system per second. The circuit design has some parts in common with that of our proposed hardware implementation.…”

mentioning

confidence: 99%

“…Yasudo et al proposed an efficient GPU implementation of solving QUBO problems. 7 The implementation performs massively parallel local searches by cooperation with the host CPU and multiple GPUs. The host CPU performs the global search with a genetic algorithm while GPUs simultaneously execute the local searches with the thousands of processors on the GPUs.…”

mentioning

confidence: 99%

See 4 more Smart Citations

High‐throughput FPGA implementation for quadratic unconstrained binary optimization

Kagawa

Ito

Nakano

et al. 2021

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Quadratic unconstrained binary optimization (QUBO) is a combinatorial optimization problem. Since various NP‐hard problems such as the traveling salesman problem can be formulated as a QUBO instance, QUBO is used with a wide range of applications. The main contribution of this article is to propose high‐throughput FPGA implementations for the QUBO solver. We perform the local search using different bit‐selection strategies based on the simulated annealing in the proposed implementation. The hardware is a pipeline structure with no pipeline hazards using multiple instances, where the bit‐flip operation is always performed every clock cycle. We implemented the proposed circuit on Xilinx UltraScale+ FPGA VU9P. The implementation result shows that the circuit can search 4.30×1011 solutions per second. Besides, by sharing the block RAM that stores a weight matrix, we implemented a dual annealer architecture that has two QUBO solvers into the FPGA. As a result, the dual annealer architecture can search 6.76×1011 solutions per second.

show abstract

Section: Ta B L E 1 the Relation Between D I And U I Whenmentioning

confidence: 83%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

High‐throughput FPGA implementation for quadratic unconstrained binary optimization

Kagawa

Ito

Nakano

et al. 2021

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

Bit duplication technique to generate hard quadratic unconstrained binary optimization problems with adjustable sizes

Li,

Nakano,

Ito

et al. 2023

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SummaryQuadratic unconstrained binary optimization (QUBO) is a combinatorial optimization to find an optimal binary solution vector that minimizes the energy value defined by a quadratic formula of binary variables in the vector. The main contribution of this article is to propose the bit duplication technique that can specify the number of duplicated bits, so that it can generate hard QUBO problem with adjustable sizes. The idea is to duplicate specified number of bits and then to give constraints so that the corresponding two bits take the same binary values. By this technique, any QUBO problem with bits is converted to a hard QUBO problem with bits . We use random QUBO problems, N‐Queen problems, traveling salesman problem and maximum weight matching problems for experiments. The performance of QUBO solvers including Gurobi optimizer, Fixstars Amplify AE, OpenJij with SA, D‐Wave samplers with SA, D‐Wave hybrid and ABS2 QUBO solver are evaluated for solving these QUBO problems. The experimental results show that only a small scale of duplicated bits can make QUBO problems harder. Hence, the bit duplication technique is a potent method to generate hard QUBO problems and generated QUBO problems can be used as benchmark problems for evaluating the search performance of QUBO solvers.

show abstract

GPU-accelerated scalable solver with bit permutated cyclic-min algorithm for quadratic unconstrained binary optimization

Yasudo

Nakano

Ito

et al. 2022

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Adaptive Bulk Search: Solving Quadratic Unconstrained Binary Optimization Problems on Multiple GPUs

Cited by 12 publications

References 18 publications

High‐throughput FPGA implementation for quadratic unconstrained binary optimization

High‐throughput FPGA implementation for quadratic unconstrained binary optimization

Bit duplication technique to generate hard quadratic unconstrained binary optimization problems with adjustable sizes

GPU-accelerated scalable solver with bit permutated cyclic-min algorithm for quadratic unconstrained binary optimization

Contact Info

Product

Resources

About