14,842 papers · page 8 of 743
Decoupled Access-Execute (DAE) architectures separate memory accesses from computation in two specialized units. This design is becoming increasingly popular among hyperscalers to accelerate irregular embedding lookups in recommendation models. In this paper, we first broaden the…
Modern shared-memory parallel programming models, such as OpenMP and Cilk, enable developers to encode a parallel execution plan within their code. Existing compilers, including Clang and GCC, directly lower or add additional compatible parallelism on top of the developers’ plan.…
Operator fusion is essential for accelerating FHE-based DNN inference because it reduces multiplicative depth and, in turn, lowers the cost of ciphertext operations by keeping them at lower ciphertext levels. Existing approaches either rely on manual optimizations, which miss cro…
Modern processors are breaking a fundamental rule: backward compatibility within their own ISA families. We term this Generational ISA Fragmentation (GIF), where newer processors cannot execute instructions supported by prior generations within the same ISA family. This phenomeno…
We describe LEGO, a new approach to optimizing data movement whereby code is expressed as a layout-independent computation and composed with layouts for data and computation. This code generator organization derives complex indexing expressions associated with hierarchical parall…
Code optimization is a crucial task that aims to enhance code performance. However, this process is often tedious and complex, highlighting the necessity for automatic code optimization techniques. Reinforcement Learning (RL) has emerged as a promising approach for tackling such …
Python dominates AI development and is the most widely used dynamic programming language, but synthesizing its polymorphic functions into hardware remains challenging. Existing HLS solutions support only static subsets of Python, forcing CPU offload with costly communication over…
Android utilizes Ahead-of-Time (AOT) compilation technology to precompile applications and stores the compiled code in OAT files, thereby improving app performance. When the Android system is updated, the old OAT files become invalidated. Applications will fall back to interprete…
The rapid growth in the size of deep learning models strains the capabilities of dense computation paradigms. Leveraging sparse computation has become increasingly popular for training and deploying large-scale models, but existing deep learning frameworks lack extensive support …
High-performance numerical quantum compilers rely on classical optimization, but are limited by slow numerical evaluations and a design that makes extending them with new instructions a difficult, error-prone task for domain experts. This paper introduces OpenQudit, a compilation…
The growing demand for domain-specific accelerators in fields such as machine learning, graph analytics, and scientific computing has highlighted the need for productive and efficient hardware design methodologies. High-level synthesis (HLS) offers an attractive solution by gener…
Efficient GPU programming is crucial for achieving high performance in deep learning (DL) applications. The performance of GPU programs depends on how data is parallelized across threads and arranged within memory subsystems. The mapping functions describing tensors on GPUs are k…
Tensor accelerators now represent a growing share of compute resources in modern CPUs and GPUs. However, they are hard to program, leading developers to use vendor-provided kernel libraries that support tensor accelerators. As a result, the usage of tensor accelerators is limited…
GPUs power modern scientific and AI applications, but their limited memory capacity restricts scalability. Buying GPUs with larger HBM is prohibitively expensive and still bounded by market limits. Existing solutions either exploit application-specific knowledge through out-of-co…
Sparse Matrix-Matrix Multiplication (SpMM) on GPU is critical to applications ranging from scientific simulations to Graph Neural Networks (GNNs) and Deep Neural Networks (DNNs). Modern GPUs offer diverse processing units, such as CUDA cores, Tensor Cores, and Sparse Tensor Cores…
Domain-Specific languages (DSLs) such as Triton enable developers to write high-performance GPU kernels in a Python-friendly manner; however, profiling these kernels with existing tools often incurs runtime and storage overhead while failing to deliver actionable insights for bot…
We present an intrinsic representation of type theory in the proof assistant Cubical Agda, inspired by Awodey’s natural models of type theory. The initial natural model is defined as quotient inductive-inductive-recursive types, leading us to a syntax accepted by Cubical Agda wit…
Step-indexing and the later modality ▷ P are widely used in program logics. A key challenge in proofs in step-indexed logics is turning ▷ P into P, coined the later elimination problem. Later elimination cannot be done unconditionally, and has traditionally been linked one-to-one…
We design a Rocq library about adhesive categories, using Hierarchy Builder (HB). It is built around two hierarchies. The first is for categories, with usual categories at the bottom and adhesive categories at the top, with weaker variants of adhesive categories in between. The s…
Verifying cache coherence protocols is a notoriously difficult problem. At the intersection between distributed protocol and computer architecture, it has long served as a premier target for formal methods. Current verification approaches hinge on the challenging discovery of lar…