HAXX (Hamilton's Quaternion Algebra for CXX) is a C++ software infrastructure for the development of efficient scalar and tensorial quaternion algorithms. HAXX can be thought of as two interdependent C++ software libraries:
- The HAXX scalar quaternion class which handles the scalar operations (+,-,etc) over the quaternion numbers
- HBLAS for quaternion and mixed-type linear algebra
HBLAS (Hamilton's Quaternion Basic Linear Algebra Subroutines) provides a BLAS-like interface for matrices and vectors over the quaternion numbers. As HBLAS depends solely on the HAXX scalar quaternion infrastructure, there is no plan currently to release them separately. Currently, HBLAS provides an optimized (serial) software implementation of quaternion matrix operations for AVX and AVX2 microarchitectures (see arXiv:1903.05575 for details).
HAXX is currently a development code which has been hand tuned for a select few microarchitectures to demonstrate the efficacy of such operations on modern computing platforms. The default caching parameters shipped with HAXX are optimized for the Intel(R) Xeon(R) CPU E5-2660 (Sandy Bridge) processor. The API specification in HAXX is very flexible, but most of the flexibility is not directly user-facing. If there is interest in exposing such functionality, please open a GitHub issue.
A primary goal of HBLAS is not only to provide a convenient and efficient interface for quaternion-quaternion linear algebra, but also to efficiently handle mixed-type (quaternion-real, quaternion-complex) linear algebra through their natural embeddings into the quaternion algebra. HAXX is actively being developed with little focus on backwards compatibility with previous versions. The HAXX and HBLAS interfaces are constantly evolving and can (will) change on a regular basis as new, exciting functionality is added.
- A high-level, modern C++ API for scalar quaternion algebra (addition, subtraction, multiplication, division)
- Access to low level optimization and vectorization of the real arithmetic underlying quaternion operations
- Extension of BLAS functionality to quaternion algebra and mixed-type expressions (HBLAS)
- A reusable software framework to enable future scalar and tensorial algorithmic development using the quaternion algebra
David Williams-Young (Computational Research Division / Lawrence Berkeley National Laboratory)
E-Mail: dbwy at lbl dot gov