Related papers: KunServe: Parameter-centric Memory Management for …

LLM serving is increasingly multi-tenant: the same deployment must handle latency-critical interactive requests and more relaxed background workloads under a fixed GPU budget. This creates a tiered-SLO setting where maximizing overall…

The continued growth of the computational capability of throughput processors has made throughput processors the platform of choice for a wide variety of high performance computing applications. Graphics Processing Units (GPUs) are a prime…

The widespread adoption of LLMs has driven an exponential rise in their deployment, imposing substantial demands on inference clusters. These clusters must handle numerous concurrent queries for different LLM downstream tasks. To handle…

The rapid increase in LLM model sizes and the growing demand for long-context inference have made memory a critical bottleneck in GPU-accelerated serving systems. Although high-bandwidth memory (HBM) on GPUs offers fast access, its limited…

Inference for Large Language Models (LLMs) is computationally demanding. To reduce the cost of auto-regressive decoding, Key-Value (KV) cache is used to store intermediate activations, which significantly lowers the computational overhead…

The context window of large language models (LLMs) is rapidly increasing, leading to a huge variance in resource usage between different requests as well as between different phases of the same request. Restricted by static parallelism…

Training LLMs larger than the aggregated memory of multiple GPUs is increasingly necessary due to the faster growth of LLM sizes compared to GPU memory. To this end, multi-tier host memory or disk offloading techniques are proposed by state…

We propose a server-based approach to manage a general-purpose graphics processing unit (GPU) in a predictable and efficient manner. Our proposed approach introduces a GPU server that is a dedicated task to handle GPU requests from other…

The rise of LLMs has driven demand for private serverless deployments, characterized by moderate-sized models and infrequent requests. While existing serverless solutions follow exclusive GPU allocation, we take a step back to explore…

The rapid scaling of Large Language Models (LLMs) has pushed training workloads far beyond the limits of single-node analysis, demanding a deeper understanding of how these models behave across large-scale, multi-GPU systems. In this paper,…

Large Language Model (LLM) serving systems remain fundamentally fragile, where frequent hardware faults in hyperscale clusters trigger disproportionate service outages in the software stack. Current recovery mechanisms are prohibitively…

High throughput serving of large language models (LLMs) requires batching sufficiently many requests at a time. However, existing systems struggle because the key-value cache (KV cache) memory for each request is huge and grows and shrinks…

Offline LLM inference seeks to maximize request processing under fixed budgets, making commodity GPU servers a promising choice. However, prior work typically considers offloading and parallelism in isolation, resulting in suboptimal…

Large Language Model (LLM) adapters enable low-cost model specialization, but introduce complex caching and scheduling challenges in distributed serving systems where hundreds of adapters must be hosted concurrently. While prior work has…

Training large language models (LLMs) in the cloud faces growing memory bottlenecks due to the limited capacity and high cost of GPUs. While GPU memory offloading to CPU and NVMe has made large-scale training more feasible, existing…

Large Language Models (LLMs) are wildly popular today and it is important to serve them efficiently. Existing LLM serving systems are stateless across requests. Consequently, when LLMs are used in the common setting of multi-turn…

The increasing adoption of large language models (LLMs) necessitates inference serving systems that can deliver both high throughput and low latency. Deploying LLMs with hundreds of billions of parameters on memory-constrained GPUs exposes…

Long-context LLM serving is bottlenecked by the cost of attending over ever-growing KV caches. Dynamic sparse attention promises relief by accessing only a small, query-dependent subset of the KV state per decoding step and extending the KV…

LLMs are widely adopted in production, pushing inference systems to their limits. Disaggregated LLM serving (e.g., PD separation and KV state disaggregation) improves scalability and cost efficiency, but it also turns KV into an explicit…

KV cache management is essential for efficient LLM inference. To maximize utilization, existing inference engines evict finished requests' KV cache if new requests are waiting. This policy breaks for agentic workloads, which interleave LLM…