Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.02658

Annotation-Driven Migration of CUDA Programs to Tenstorrent Blackhole

Ayumi Ohno·Shinya Takamaeda-Yamazaki
View original source ↗Read source document ↗Export metadata ↓

Abstract

Tenstorrent Blackhole combines distributed local memories, explicit inter-core communication, and Tensix cores decoupling data movement from computation. CUDA offers a substantial HPC software base but leaves physical data placement and scheduling largely implicit. Migrating CUDA HPC kernels to Blackhole requires spatial mapping across cores and per-core coordination of compute and data-movement kernels. We present an MLIR-based compiler deriving data placement, inter-core communication, and tile computation from a statically shaped affine CUDA subset. Declarative annotations express choices not fixed by the source, including compute/DM operation placement and streaming granularity. We evaluate it on Gaussian elimination, five-point stencil, and symmetric rank-k update. Relative to our compiler's default realizations, the best measured configurations achieve a 4.2x speedup for BF16 Gaussian and 2.1-2.2x for FP32 Gaussian and Jacobi. A choice's performance impact can reverse with surrounding policies, precision, and loop schedule, motivating comparison of alternative per-core realizations rather than independent policy selection.

Explore related subjects

cs.DC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Ayumi Ohno, Shinya Takamaeda-Yamazaki. 2026-10-02. Annotation-Driven Migration of CUDA Programs to Tenstorrent Blackhole. https://arxiv.org/abs/2610.02658

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

DIKTAMO: Extending CXL for Resilience to CPU Failures

Compute Express Link (CXL) 3.0 and beyond allows the compute nodes of a cluster to share data with hardware cache coherence and at the granularity of a cache line. This enables shared-memory semantics for distributed computing, but introduces a new resilience challenge: a node failure leads to the loss of the dirty data in its caches, corrupting application state. Sadly, the CXL specification does not consider processor failures. Moreover, when a component fails, the specification tries to isolate it and continue application execution; there is no attempt to bring the application to a consistent state -- a step required to recover a shared-memory program. To address these limitations, this paper extends CXL to be resilient to node failures, and to correctly recover the application after node failures. We call the system DIKTAMO. To survive node failures, DIKTAMO augments the coherence transaction of a write with messages that propagate the update to a small set of other nodes (i.e., Replicas). Replicas save the update in a local hardware Logging Unit. Replication ensures resilience to node failures. Then, at regular intervals, the Logging Units dump the compressed updates to memory. After a node failure, recovery involves using the logs to bring the directory and memory to a correct state. Our evaluation with 16 4-core nodes shows that DIKTAMO enables fault-tolerant execution with 30% slowdown with 3 replicas (or 27% with 2 replicas) over a platform without fault-tolerance support. DIKTAMO is 2.82x faster than ensuring fault tolerance by using a write-through protocol.

cs.DC↗
arXiv2026

EmuGEMM: Fused Tensor Core Kernels for Precision Emulation in Matrix Multiplication

Modern GPUs devote an increasing silicon budget to low-precision matrix-multiplication units, widening the precision-throughput gap for scientific computing workloads. Ozaki Schemes I and II offer an alternative by reconstructing high-precision general matrix multiplication (GEMM) from low-precision operations, yet existing implementations leave substantial performance untapped. In particular, intermediate results are repeatedly materialized in global memory, making data movement the dominant bottleneck. We present EmuGEMM, fused integer Tensor Core kernels for NVIDIA Hopper and Blackwell GPUs that eliminate redundant memory round-trips in both Ozaki schemes. Using Scheme I, EmuGEMM sustains up to 1,639 Top/s on Hopper (83% of INT8 peak) and 3,654 Top/s on Blackwell (81%). For large matrices, EmuGEMM surpasses cuBLAS TF32 throughput by up to 1.4x on Hopper and 1.7x on Blackwell, at comparable accuracy. Using Scheme II, EmuGEMM extends to complex arithmetic and outperforms cuBLAS ZGEMM by up to 2.3x on Hopper and 5.5x on Blackwell.

cs.DC↗
arXiv2026

Lightweight and Resource-Efficient Perception for Robotic Guide Dogs

Robotic guide dogs should understand their surroundings, objects, and potential risks. Prior research has focused on raw sensor data from cameras and 2D or 3D LiDAR, which precisely measure distance points rather than provide a semantic understanding of the scene. While these physical measurements are effective for robot-centric collision avoidance and robot safety, they are not suitable for human-centric guidance. The system should recognize the type and relevance of obstacles and explain them, clearly and actionably, in terms of their spatial relation to the user. We present complete on-device perception modules that fuse a 360 camera and a 2D LiDAR for reliable collision avoidance, with moving-object detection and tracking for human-centric guidance. Finally, in walking-impossible situations, a vision--language model delivers pathway explanations as a safety mechanism to reduce user anxiety. In experiments, verification of fused 360 camera--LiDAR depth shows reliable near-range perception but inherent mid-range bias, while the system as a whole sustained real-time performance under 55 W. On the real-world egocentric GuideDogQA benchmark, our system achieved 83.8\% accuracy, compared with 67.1\% for GPT-4o. These results demonstrate that practical human-centric guidance with real-time on-device inference is feasible even on quadrupeds.

cs.DC↗