相关论文: BlockBPE: Parallel BPE Tokenization

The challenge of mapping AI architectures to GPU hardware is creating a critical bottleneck in AI progress. Despite substantial efforts, hand-written custom kernels fail to meet their theoretical performance thresholds, even on…

As large language models gain widespread adoption, running them efficiently becomes crucial. Recent works on LLM inference use speculative decoding to achieve extreme speedups. However, most of these works implicitly design their algorithms…

Efficiently processing long sequences with Transformer models usually requires splitting the computations across accelerators via context parallelism. The dominant approaches in this family of methods, such as Ring Attention or DeepSpeed…

Subword tokenization schemes are the dominant technique used in current NLP models. However, such schemes can be rigid and tokenizers built on one corpus do not adapt well to other parallel corpora. It has also been observed that in…

Decoding the linguistic intricacies of the genome is a crucial problem in biology, and pre-trained foundational models such as DNABERT and Nucleotide Transformer have made significant strides in this area. Existing works have largely hinged…

Bloom filters are a fundamental data structure for approximate membership queries, with applications ranging from data analytics to databases and genomics. Several variants have been proposed to accommodate parallel architectures. GPUs,…

The success of large pretrained Transformers is closely tied to tokenizers, which convert raw input into discrete symbols. Extending these models to graph-structured data remains a significant challenge. In this work, we introduce a graph…

This paper addresses the problem of universal synchronization primitives that can support scalable thread synchronization for large-scale many-core architectures. The universal synchronization primitives that have been deployed widely in…

The task of integrating a large number of independent ODE systems arises in various scientific and engineering areas. For nonstiff systems, common explicit integration algorithms can be used on GPUs, where individual GPU threads…

The pervasive "memory wall" bottleneck is significantly amplified in modern large-scale Mixture-of-Experts (MoE) architectures. MoE's inherent architectural sparsity leads to sparse arithmetic compute and also introduces substantial…

Fast, scalable decoding architectures that operate in a block-wise parallel fashion across space and time are essential for real-time fault-tolerant quantum computing. We introduce a scalable AI-based pre-decoder for the surface code that…

By supporting the access of multiple memory words at the same time, Bit-line Computing (BC) architectures allow the parallel execution of bit-wise operations in-memory. At the array periphery, arithmetic operations are then derived with…

Mixture of Experts (MoE) models have emerged as the de facto architecture for scaling up language models without significantly increasing the computational cost. Recent MoE models demonstrate a clear trend towards high expert granularity…

The transition from standard generative AI to \emph{reasoning-centric architectures}, exemplified by models capable of extensive Chain-of-Thought~(CoT) processing, marks a fundamental paradigm shift in system requirements. Unlike…

Massive multi-threading in GPU imposes tremendous pressure on memory subsystems. Due to rapid growth in thread-level parallelism of GPU and slowly improved peak memory bandwidth, the memory becomes a bottleneck of GPU's performance and…

Inter-GPU communication has become a major bottleneck for modern AI workloads as models scale and improvements in hardware compute throughput outpace improvements in interconnect bandwidth. Existing systems mitigate this through…

Efficient Graph processing is challenging because of the irregularity of graph algorithms. Using GPUs to accelerate irregular graph algorithms is even more difficult to be efficient, since GPU's highly structured SIMT architecture is not a…

Transformer-based large language models (LLMs) have achieved remarkable success as model sizes continue to grow, yet their deployment remains challenging due to significant computational and memory demands. Quantization has emerged as a…

Existing Machine Translation (MT) research often suggests a single, fixed set of hyperparameters for word segmentation models, symmetric Byte Pair Encoding (BPE), which applies the same number of merge operations (NMO) to train tokenizers…

Automatic parallelization remains a challenging problem in software engineering, particularly in identifying code regions where loops can be safely executed in parallel on modern multi-core architectures. Traditional static analysis…