Related papers: One Pool, Two Caches: Adaptive HBM Partitioning fo…

Mixture-of-Expert (MoE) models enable efficient inference by employing smaller experts and activating only a subset of them per token. MoE serving engines distribute experts across multiple GPUs and route tokens to appropriate GPUs at…

The rapid adoption of large language models (LLMs) is pushing AI accelerators toward increasingly powerful and specialized designs. Instead of further complicating software development with deeply hierarchical scratchpad memories (SPMs) and…

The significant computational demands of pretrained language models (PLMs), which often require dedicated hardware, present a substantial challenge in serving them efficiently, especially in multi-tenant environments. To address this, we…

Sparse Matrix-Matrix Multiplication (SpMM) is a fundamental operation in graph computing and analytics. However, the irregularity of real-world graphs poses significant challenges to achieving efficient SpMM operation for graph data on…

Existing large language model (LLM) serving systems fall into two categories: 1) a unified system where prefill phase and decode phase are co-located on the same GPU, sharing the unified computational resource and storage, and 2) a…

We study the problem of load balancing in distributed stream processing engines, which is exacerbated in the presence of skew. We introduce Partial Key Grouping (PKG), a new stream partitioning scheme that adapts the classical "power of two…

Retrieval-Augmented Generation (RAG) relies on large-scale Approximate Nearest Neighbor Search (ANNS) to retrieve semantically relevant context for large language models. Among ANNS methods, IVF-PQ offers an attractive balance between…

Production vLLM fleets typically provision each instance for the worst-case context length, leading to substantial KV-cache over-allocation and under-utilized concurrency. In practice, 80-95% of requests are short, yet are served under…

With the rapid explosion of data volume from mobile networks, edge caching has received significant attentions as an efficient approach to boost content delivery efficiency by bringing contents near users. In this article, cache-enabled…

In the competitive landscape of sponsored search, balancing retrieval quality with production latency is a critical challenge. While large retrieval models based on Small Language Models (SLMs) such as Qwen3-Embedding-4B/8B set strong upper…

Large language models (LLMs) with different architectures and sizes have been developed. Serving each LLM with dedicated GPUs leads to resource waste and service inefficiency due to the varying demand of LLM requests. A common practice is…

Meeting the strict Quality of Service (QoS) requirements of terminals has imposed a signiffcant challenge on Multiaccess Edge Computing (MEC) systems, due to the limited multidimensional resources. To address this challenge, we propose a…

Large language model (LLM)-based inference workloads increasingly dominate data center costs and resource utilization. Therefore, understanding the inference workload characteristics on evolving CPU-GPU coupled architectures is crucial for…

Improving time-to-solution in molecular dynamics simulations often requires strong scaling due to fixed-sized problems. GROMACS is highly latency-sensitive, with peak iteration rates in the sub-millisecond, making scalability on…

Today's high-speed switches employ an on-chip shared packet buffer. The buffer is becoming increasingly insufficient as it cannot scale with the growing switching capacity. Nonetheless, the buffer needs to face highly intense bursts and…

The memory-for-computation paradigm of KV caching is essential for accelerating large language model (LLM) inference service, but limited GPU high-bandwidth memory (HBM) capacity motivates offloading the KV cache to cheaper external storage…

Cloud-device collaborative recommendation partitions computation across the cloud and user devices: the cloud provides semantic user modeling, while the device leverages recent interactions and cloud semantic signals for privacy-preserving,…

The hidden Markov model (HMM) is a widely-used generative model that copes with sequential data, assuming that each observation is conditioned on the state of a hidden Markov chain. In this paper, we derive a novel algorithm to cluster HMMs…

The increasing memory demand of the Key-Value (KV) cache poses a significant bottleneck for Large Language Models (LLMs) in long-context applications. Existing low-rank KV compression methods reduce this footprint by modifying model…

Large language models (LLMs) are widely used but expensive to run, especially as inference workloads grow. To lower costs, maximizing the request batch size by managing GPU memory efficiently is crucial. While PagedAttention has recently…