Related papers: UniCAIM: A Unified CAM/CIM Architecture with Stati…

Large Language Models (LLMs) possess remarkable generalization capabilities but struggle with multi-task adaptation, particularly in balancing knowledge retention with task-specific specialization. Conventional fine-tuning methods suffer…

Large Language Model (LLM) inference uses an autoregressive manner to generate one token at a time, which exhibits notably lower operational intensity compared to earlier Machine Learning (ML) models such as encoder-only transformers and…

Rapid advancements in artificial intelligence have given rise to transformative models, profoundly impacting our lives. These models demand massive volumes of data to operate effectively, exacerbating the data-transfer bottleneck inherent…

Large Language Models (LLMs) are increasingly deployed in scenarios demanding ultra-long context reasoning, such as agentic workflows and deep research understanding. However, long-context inference is constrained by the KV cache, a…

Large Language Models (LLMs) face limitations due to the high demand on GPU memory and computational resources when handling long contexts. While sparsify the Key-Value (KV) cache of transformer model is a typical strategy to alleviate…

Generative large language models (LLMs) have revolutionized multiple domains. Modern LLMs predominantly rely on an autoregressive decoding strategy, which generates output tokens sequentially and employs a key-value cache (KV cache) to…

Computationally hard combinatorial optimization problems (COPs) are ubiquitous in many applications, including logistical planning, resource allocation, chip design, drug explorations, and more. Due to their critical significance and the…

Large Language Model or LLM inference has two phases, the prompt (or prefill) phase to output the first token and the extension (or decoding) phase to the generate subsequent tokens. In this work, we propose an efficient parallelization…

Processing-in-Memory (PIM) architectures offer a promising solution to the memory bottlenecks in data-intensive machine learning, yet often overlook the growing challenge of activation memory footprint. Conventional PIM approaches struggle…

Large Language Models (LLMs) have made remarkable progress in processing extensive contexts, with the Key-Value (KV) cache playing a vital role in enhancing their performance. However, the growth of the KV cache in response to increasing…

Recent reasoning models such as OpenAI-o1 and DeepSeek-R1 have shown strong performance on complex tasks including mathematical reasoning and code generation. However, this performance gain comes with substantially longer output sequences,…

Data movement in memory-intensive workloads, such as deep learning, incurs energy costs that are over three orders of magnitude higher than the cost of computation. Since these workloads involve frequent data transfers between memory and…

As the demand for long-context large language models (LLMs) increases, models with context windows of up to 128K or 1M tokens are becoming increasingly prevalent. However, long-context LLM inference is challenging since the inference speed…

Transformer-based large language models (LLMs) have already achieved remarkable results on long-text tasks, but the limited GPU memory (VRAM) resources struggle to accommodate the linearly growing demand for key-value (KV) cache as the…

Generative reasoning with large language models (LLMs) often involves long decoding sequences, leading to substantial memory and latency overheads from accumulating key-value (KV) caches. While existing KV compression methods primarily…

In the field of instruction-following large vision-language models (LVLMs), the efficient deployment of these models faces challenges, notably due to the high memory demands of their key-value (KV) caches. Conventional cache management…

The quadratic computational complexity of the standard attention mechanism constitutes a fundamental bottleneck for large language models in long-context inference. While existing KV cache compression methods alleviate memory pressure, they…

Unified Virtual Memory (UVM) relieves the developers from the onus of maintaining complex data structures and explicit data migration by enabling on-demand data movement between CPU memory and GPU memory. However, on-demand paging soon…

Generating long sequences of tokens given a long-context input is a very compute-intensive inference scenario for large language models (LLMs). One prominent inference speed-up approach is to construct a smaller key-value (KV) cache,…

The Key-Value (KV) cache reading latency increases significantly with context lengths, hindering the efficiency of long-context LLM inference. To address this, previous works propose retaining a small fraction of KV cache based on token…