Diffusion-based large language models (dLLMs) have recently gained significant attention for their exceptional performance and inherent potential for parallel decoding. Existing frameworks further enhance its inference efficiency by enabling KV caching. However, its bidirectional attention mechanism necessitates periodic cache refreshes that interleave prefill and decoding phases, both contributing substantial inference cost and constraining achievable speedup. Inspired by the heterogeneous arithmetic intensity of the prefill and decoding phases, we propose ODB-dLLM, a framework that orchestrates dual-boundaries to accelerate dLLM inference. In the prefill phase, we find that the predefined fixed response length introduces heavy yet redundant computational overhead, which affects efficiency. To alleviate this, ODB-dLLM incorporates an adaptive length prediction mechanism that progressively reduces prefill overhead and unnecessary computation. In the decoding phase, we analyze the computational characteristics of dLLMs and propose a dLLM-specific jump-share speculative decoding method to enhance efficiency by reducing the number of decoding iterations. Experimental results demonstrate that ODB-dLLM achieves 46-162x and 2.63-6.30x speedups over the baseline dLLM and Fast-dLLM, respectively, while simultaneously mitigating the accuracy degradation in existing acceleration frameworks.

Large Language Models (LLMs) have gained popularity in recent years, driving up the demand for inference. LLM inference is composed of two phases with distinct characteristics: a compute-bound prefill phase followed by a memory-bound decode phase. To efficiently serve LLMs, prior work proposes prefill-decode disaggregation to run each phase on separate hardware. However, existing hardware poorly matches the different requirements of each phase. Current datacenter GPUs and TPUs follow a more-is-better design philosophy that maximizes compute and memory resources, causing memory bandwidth underutilization in the prefill phase and compute underutilization in the decode phase. Such underutilization directly translates into increased serving costs. This paper proposes SPAD (Specialized Prefill and Decode hardware), adopting a less-is-more methodology to design specialized chips tailored to the distinct characteristics of prefill and decode phases. The proposed Prefill Chips have larger systolic arrays and use cost-effective GDDR memory, whereas the proposed Decode Chips retain high memory bandwidth but reduce compute capacity. Compared to modeled H100s, simulations show that the proposed Prefill Chips deliver 8% higher prefill performance on average at 52% lower hardware cost, while the proposed Decode Chips achieve 97% of the decode performance with 28% lower TDP. End-to-end simulations on production traces show that SPAD reduces hardware cost by 19%-41% and TDP by 2%-17% compared to modeled baseline clusters while offering the same performance. Even when models and workloads change, SPAD can reallocate either type of chip to run either phase and still achieve 11%-43% lower hardware costs, demonstrating the longevity of the SPAD design.

The rapid adoption of Large Language Models (LLMs) has driven a growing demand for efficient inference, particularly in latency-sensitive applications such as chatbots and personalized assistants. Unlike traditional deep neural networks, LLM inference proceeds in two distinct phases: the prefill phase, which processes the full input sequence in parallel, and the decode phase, which generates tokens sequentially. These phases exhibit highly diverse compute and memory requirements, which makes accelerator design particularly challenging. Prior works have primarily been optimized for high-batch inference or evaluated only short input context lengths, leaving the low-batch and long context regime, which is critical for interactive applications, largely underexplored. We propose HALO, a heterogeneous memory centric accelerator designed for these unique challenges of prefill and decode phases in low-batch LLM inference. HALO integrates HBM based Compute-in-DRAM (CiD) with an on-chip analog Compute-in-Memory (CiM), co-packaged using 2.5D integration. To further improve the hardware utilization, we introduce a phase-aware mapping strategy that adapts to the distinct demands of the prefill and decode phases. Compute bound operations in the prefill phase are mapped to CiM to exploit its high throughput matrix multiplication capability, while memory-bound operations in the decode phase are executed on CiD to benefit from reduced data movement within DRAM. Additionally, we present an analysis of the performance tradeoffs of LLMs under two architectural extremes: a fully CiD and a fully on-chip analog CiM design to highlight the need for a heterogeneous design. We evaluate HALO on LLaMA-2 7B and Qwen3 8B models. Our experimental results show that LLMs mapped to HALO achieve up to 18x geometric mean speedup over AttAcc, an attention-optimized mapping and 2.5x over CENT, a fully CiD based mapping.

As large language models become increasingly prevalent, memory bandwidth constraints significantly limit inference throughput, motivating post-training quantization (PTQ). In this paper, we propose FireQ, a co-designed PTQ framework and an INT4-FP8 matrix multiplication kernel that accelerates LLM inference across all linear layers. Specifically, FireQ quantizes linear layer weights and key-values to INT4, and activations and queries to FP8, significantly enhancing throughput. Additionally, we introduce a three-stage pipelining for the prefill phase, which modifies the FlashAttention-3 kernel, effectively reducing time-to-first-token in the prefill phase. To minimize accuracy loss from quantization, we develop novel outlier smoothing techniques tailored separately for linear and attention layers. In linear layers, we explicitly use per-tensor scaling to prevent underflow caused by the FP8 quantization scaling factor of INT4 quantization, and channel-wise scaling to compensate for coarse granularity of INT4. In attention layers, we address quantization challenges posed by rotary positional embeddings (RoPE) by combining pre-RoPE and post-RoPE scaling strategies. FireQ significantly outperforms state-of-the-art methods, achieving 1.68x faster inference in feed-forward network layers on Llama2-7B and 1.26x faster prefill phase performance on Llama3-8B compared to QServe, with negligible accuracy loss.

Serving large language models (LLMs) to millions of users requires efficient resource allocation and parallelism strategies. It is a labor intensive trial-and-error process to find such a strategy. We present BestServe, a novel framework for ranking serving strategies by estimating goodput under various operating scenarios. Supporting both collocated and disaggregated architectures, BestServe leverages an inference simulator built on an adapted roofline model and CPU-GPU dispatch dynamics. Our framework determines the optimal strategy in minutes on a single standard CPU, eliminating the need for costly benchmarking, while achieving predictions within a $20\%$ error margin. It appeals to be practical for rapid deployment planning because of its lightweight design and strong extensibility.

The rapid advancement of Large Language Models (LLMs) has driven the need for more efficient serving strategies. In this context, efficiency refers to the proportion of requests that meet their Service Level Objectives (SLOs), particularly for Time To First Token (TTFT) and Time Between Tokens (TBT). However, existing systems often prioritize one metric at the cost of the other. We present Ascendra, an LLM serving system designed to meet both TTFT and TBT SLOs simultaneously. The core insight behind Ascendra is that a request's urgency evolves as it approaches its deadline. To leverage this, Ascendra partitions GPU resources into two types of instances: low-priority and high-priority. Low-priority instances maximize throughput by processing requests out of arrival order, but at the risk of request starvation. To address this, Ascendra employs a performance model to predict requests at risk of missing their SLOs and proactively offloads them to high-priority instances. High-priority instances are optimized for low-latency execution and handle urgent requests nearing their deadlines. This partitioned architecture enables Ascendra to effectively balance high throughput and low latency. Extensive evaluation shows that Ascendra improves system throughput by up to 1.7x compared to vLLM and Sarathi-Serve while meeting both TTFT and TBT SLOs.

--Generative large language models (LLMs) have garnered significant attention due to their exceptional capabilities in various AI tasks. Traditionally deployed in cloud datacenters, LLMs are now increasingly moving towards more accessible edge platforms to protect sensitive user data and ensure privacy preservation. The limited computational resources of individual edge devices, however, can result in excessively prolonged inference latency and overwhelmed memory usage. While existing research has explored collaborative edge computing to break the resource wall of individual devices, these solutions yet suffer from massive communication overhead and under-utilization of edge resources. Furthermore, they focus exclusively on optimizing the prefill phase, neglecting the crucial autoregressive decoding phase for generative LLMs. T o address that, we propose Jupiter, a fast, scalable, and resource-efficient collaborative edge AI system for generative LLM inference. Jupiter introduces a flexible pipelined architecture as a principle and differentiates its system design according to the differentiated characteristics of the prefill and decoding phases. For prefill phase, Jupiter submits a novel intra-sequence pipeline parallelism and develops a meticulous parallelism planning strategy to maximize resource efficiency; For decoding, Jupiter devises an effective outline-based pipeline parallel decoding mechanism combined with speculative decoding, which further magnifies inference acceleration. Extensive evaluation based on realistic implementation demonstrates that Jupiter remarkably outperforms state-of-the-art approaches under various edge environment setups, achieving up to 26. 1 end-to-end latency reduction while rendering on-par generation quality. I NTRODUCTION The emergence of generative large language models (LLMs) has attracted widespread attention from both industry and academia owing to their exceptional capabilities in a wide range of artificial intelligence (AI) tasks. These models, widely deployed in cloud datacenters equipped with powerful server-grade GPUs, have driven increasing intelligent edge applications such as ChatBot [1] and smart-home AI agent [2].

In large language model (LLM) serving systems, executing each request consists of two phases: the compute-intensive prefill phase and the memory-intensive decoding phase. To prevent performance interference between the two phases, current LLM serving systems typically adopt prefill-decoding disaggregation, where the two phases are split across separate machines. However, we observe this approach leads to significant resource underutilization. Specifically, prefill instances that are compute-intensive suffer from low memory utilization, while decoding instances that are memory-intensive experience low compute utilization. To address this problem, this paper proposes Adrenaline, an attention disaggregation and offloading mechanism designed to enhance resource utilization and performance in LLM serving systems. Adrenaline's key innovation lies in disaggregating part of the attention computation in the decoding phase and offloading them to prefill instances. The memory-bound nature of decoding-phase attention computation inherently enables an effective offloading strategy, yielding two complementary advantages: 1) improved memory capacity and bandwidth utilization in prefill instances, and 2) increased decoding batch sizes that enhance compute utilization in decoding instances, collectively boosting overall system performance. Adrenaline achieves these gains through three key techniques: low-latency decoding synchronization, resource-efficient prefill colocation, and load-aware offloading scheduling. Experimental results show that Adrenaline achieves 2.28x higher memory capacity and 2.07x better memory bandwidth utilization in prefill instances, up to 1.67x improvements in compute utilization for decoding instances, and 1.68x higher overall inference throughput compared to state-of-the-art systems.

Abstract--Recent advancements in large language models (LLMs) have catalyzed a substantial surge in demand for LLM services. While traditional cloud-based LLM services satisfy high-accuracy requirements, they fall short in meeting critical demands for low delay and enhanced privacy . T o address these limitations, we propose HA T, a novel device-cloud collaborative inference framework that leverages the complementary strengths of U-shaped inference and speculative decoding. HA T partitions the LLM into three submodels, and the input and output submodels, stacked with a lightweight adapter network, are deployed as a small language model (SLM) on each end device. Meanwhile, the middle submodel, encompassing the majority of the LLM's decoder layers, is hosted in the cloud to perform speculative decoding with on-device SLMs. During inference, HA T exchanges hidden states (rather than raw tokens) of input or draft tokens between devices and the cloud, thereby incurring substantial communication delays. Besides, processing hidden states of long prompts will exacerbate computation delays in the cloud, further compromising inference efficiency . T o improve efficiency, we introduce a prompt chunking mechanism that segments long prompts into shorter chunks, enabling parallel transmission and processing. Furthermore, HA T is implemented to dynamically determine optimal chunk sizes for devices handling long prompts, thereby improving overall inference speed. Extensive experiments are conducted on a physical testbed comprising 30 NVIDIA Jetson devices and a server with 8 NVIDIA A6000 GPUs. Experimental results demonstrate that HA T achieves promising performance improvements, reducing TTFT by 41% to 54% and TBT by 41% to 77% compared to the baselines. Recent advancements in large language models (LLMs) have revolutionized the field of natural language processing, demonstrating unprecedented capabilities across various tasks and triggering exponential growth of LLM services [1], [2]. For instance, OpenAI's ChatGPT provides various services, e.g., chat-based interaction, and automated writing, to approximately 180 million users, and processes over 1.6 billion requests monthly [3]. The underlying architecture of LLM services mainly operates through an autore-gressive process, which involves a prefill phase followed by a decode phase. In prefill phase, the LLM processes all input prompt tokens simultaneously, leveraging parallel computation to generate the initial output token.

Continual learning has become a promising solution to refine large language models incrementally by leveraging user feedback. In particular, online continual learning - iteratively training the model with small batches of user feedback - has demonstrated notable performance improvements. However, the existing practice of separating training and serving processes forces the online trainer to recompute the intermediate results already done during serving. Such redundant computations can account for 30%-42% of total training time. In this paper, we propose Alchemist, to the best of our knowledge, the first online continual learning system that efficiently reuses serving activations to increase training throughput. Alchemist introduces two key techniques: (1) recording and storing activations and KV cache only during the prefill phase to minimize latency and memory overhead; and (2) smart activation offloading and hedging. Evaluations with inputs of varied token length sampled from ShareGPT dataset show that compared with a separate training cluster, Alchemist significantly increases training throughput by up to 1.72x, reduces up to 47% memory usage during training, and supports up to 2x more training tokens - all while maintaining negligible impact on serving latency.