https://hgpu.org/?p=27794
SaLoBa: Maximizing Data Locality and Workload Balance for Fast Sequence Alignment on GPUs