Benchmark of Bloqade

Methodology

the following aspects are reflected by our benchmark:

• ODE algorithm: different ODE algorithm may result in different performance on different problem setup.
• ODE solver stepping: adaptive stepping method can make the simulation faster in some cases.
• hardware acceleration
• abstraction overheads

measurement of the time: each of the benchmark experiment was measured multiple times and the miminum value of the time is picked as a stable estimator to eliminated noise from machine environment.

Note, this method does not count the just-in-compile (JIT) time into the benchmark, which can be a large overhead for small tasks.

Platform Info

CPU & System:

Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  CPU: Intel(R) Xeon(R) CPU E5-2686 v4 @ 2.30GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, broadwell)

CUDA:

Unknown NVIDIA driver, for CUDA 11.6
CUDA driver 11.6

Libraries: 
- CUBLAS: 11.8.1
- CURAND: 10.2.9
- CUFFT: 10.7.0
- CUSOLVER: 11.3.2
- CUSPARSE: 11.7.1
- CUPTI: 16.0.0
- NVML: missing
- CUDNN: 8.30.2 (for CUDA 11.5.0)
- CUTENSOR: 1.4.0 (for CUDA 11.5.0)

Toolchain:
- Julia: 1.7.3
- LLVM: 12.0.1
- PTX ISA support: 3.2, 4.0, 4.1, 4.2, 4.3, 5.0, 6.0, 6.1, 6.3, 6.4, 6.5, 7.0
- Device capability support: sm_35, sm_37, sm_50, sm_52, sm_53, sm_60, sm_61, sm_62, sm_70, sm_72, sm_75, sm_80

0: Tesla V100-SXM2-16GB (sm_70, 15.483 GiB / 15.782 GiB available)

Package Specifications and Evnironments

All non-specified options are default, e.g Julia compiler optimization level is set to the default O2, qutip/pulser is setup using the default build from pip.

Environment Variables

export OMP_NUM_THREADS=1
export MKL_NUM_THREADS=1
export MKL_DOMAIN_NUM_THREADS=1
export JULIA_NUM_THREADS=1

Bloqade:

Dependencies:

• Bloqade.jl: v0.1.5

For more detailed dependencies, please refer to Manifest.toml.

QuTiP/Pulser:

Dependencies:

• pulser: 0.6.0
• qutip: 4.7.0
• numpy: 1.21.5
• scipy: 1.8.0

Results