Hyperdimensional Computing (HDC) is a neuro-inspired computational model that represents and manipulates information using high-dimensional distributed representations that are combined and compared using simple and highly parallel vector operations. In this work, we present a highly flexible hardware acceleration unit for HDC learning tasks based on the binary spatter-code model. Integrated into the execution stage of the Klessydra-T03 RISC-V core, the unit accelerates the core arithmetic operations of HDC and can be configured at synthesis time in terms of hardware parallelism, supported operations, and size of the local memories, trading off execution time with hardware resources to match application needs. A custom RISC-V Instruction Set Extension efficiently controls the accelerator, with instructions fully integrated into the GNU Compiler Collection toolchain and exposed to the programmer as intrinsics. Dedicated Control and Status Registers allow specifying the characteristics of the high-dimensional space and the target learning tasks at runtime, controlling the hardware loops of the accelerator and enabling the same hardware architecture to be used for various tasks. The dual flexibility coming from hardware configuration and software programmability sets this work apart from application-specific solutions in the literature, offering a unique, versatile accelerator adaptable to a wide range of applications and learning tasks.
9 December 2025 in IEEE Open Journal of the Computer Society
Hyperdimensional Computing (HDC), or Vector Symbolic Architectures, is a computing paradigm that combines symbolic reasoning with the efficiency of distributed representations. HDC offers energy and computational efficiency, inherent parallelism, and robustness, making it a promising alternative for learning tasks in resource-constrained environments and hardware acceleration. This work introduces Hyperdimensional Contextual Bandits (HD-CB), the first exploration of HDC for modeling and solving Contextual Bandits (CB) tasks. CBs are sequential decision-making problems where, at each iteration, an agent observes a context, selects an action from a finite set, and receives a reward, aiming to maximize cumulative returns over time. The proposed HD-CB framework encodes contextual information in a high-dimensional space and represents each action with a dedicated hypervector. Action selection and learning are performed using HDC arithmetic, replacing the computationally expensive ridge regression used in traditional linear CB algorithms with simple, highly parallel vector operations. We propose four HD-CB variants with different exploration strategies and update rules. Experiments on synthetic datasets and the MovieLens-100k benchmark show that HD-CB achieves comparable or superior accuracy to linear CB methods, improves convergence speed, scalability, memory efficiency, and emerges as a perfect candidate for hardware acceleration, with up to $141.7\times$ speedup.
Hyperdimensional Computing (HDC) offers a robust and energy-efficient paradigm for edge intelligence; however, current hardware accelerators are often proprietary, tailored to the target learning task and tightly coupled to specific CPU microarchitectures, limiting portability and adoption. To address this, and democratize the deployment of HDC hardware, we present a general-purpose, plug-and-play accelerator IP that implements the Binary Spatter Code framework as a standalone, host-agnostic module. The design is compliant with the AMBA AXI4 standard and provides an AXI4-Lite control plane and DMA-driven AXI4-Stream datapaths coupled to a banked scratchpad memory. The architecture supports synthesis-time scalability, enabling high-throughput transfers independently of the host processor, while employing microarchitectural optimizations to minimize silicon area. A multi-layer C++ software (GitHub repository commit 3ae3b46) stack running in Linux userspace provides a unified programming model, abstracting low-level hardware interactions and enabling the composition of complex HDC pipelines. Implemented on a Xilinx Zynq XC7Z020 SoC, the accelerator achieves substantial gains over an ARM Cortex-A9 baseline, with primitive-level speedups of up to . On end-to-end classification benchmarks, the system delivers average speedups of for training and for inference. The complete RTL and software stack are released as open-source hardware to support reproducible research and rapid adoption on heterogeneous SoCs.