paper-with-me

Papers

AI and Memory Wall

2024-03-21 · Amir Gholami, Zhewei Yao, Sehoon Kim, Coleman Hooper, Michael W. Mahoney, Kurt Keutzer

The availability of unprecedented unsupervised training data, along with neural scaling laws, has resulted in an unprecedented surge in model size and compute requirements for serving/training LLMs. However, the main performance bottleneck is increasingly shifting to memory bandwidth. Over the past 20 years, peak server hardware FLOPS has been scaling at 3.0x/2yrs, outpacing the growth of DRAM and interconnect bandwidth, which have only scaled at 1.6 and 1.4 times every 2 years, respectively. This disparity has made memory, rather than compute, the primary bottleneck in AI applications, particularly in serving. Here, we analyze encoder and decoder Transformer models and show how memory bandwidth can become the dominant bottleneck for decoder models. We argue for a redesign in model architecture, training, and deployment strategies to overcome this memory limitation.

📄 PDF Abstract BibTeX arXiv:2403.14123

Code (0)

등록된 구현이 없습니다.

Tasks

Decoder

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Multi-Head Attention 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…

Similar Papers 제목 키워드 기반

Memory Wall is not gone: A Critical Outlook on Memory Architecture in Digital Neuromorphic Computing

2026-04-09 · Amirreza Yousefzadeh, Sameed Sohail, Ana Lucia Varbanescu arxiv

The rapid advancement of neuromorphic technology aims to address the memory wall challenge inherent in conventional von Neumann architectures. This paper critically examines current digital neuromorphic processors and th…

Breaking On-device Training Memory Wall: A Systematic Survey

2023-06-17 · Shitian Li, Chunlin Tian, Kahou Tam, Rui Ma 외

On-device training has become an increasingly popular approach to machine learning, enabling models to be trained directly on mobile and edge devices. However, a major challenge in this area is the limited memory availab…

NavigateSurvey

In-memory Implementation of On-chip Trainable and Scalable ANN for AI/ML Applications

2020-05-19 · Abhash Kumar, Jawar Singh, Sai Manohar Beeraka, Bharat Gupta

Traditional von Neumann architecture based processors become inefficient in terms of energy and throughput as they involve separate processing and memory units, also known as~\textit{memory wall}. The memory wall problem…

MoEBlaze: Breaking the Memory Wall for Efficient MoE Training on Modern GPUs

2026-01-08 · Jiyuan Zhang, Yining Liu, Siqi Yan, Lisen Deng 외 arxiv

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…

Energy-efficiency Limits on Training AI Systems using Learning-in-Memory

2024-02-21 · Zihao Chen, Johannes Leugering, Gert Cauwenberghs, Shantanu Chakrabartty

Learning-in-memory (LIM) is a recently proposed paradigm to overcome fundamental memory bottlenecks in training machine learning systems. While compute-in-memory (CIM) approaches can address the so-called memory-wall (i.…