Related papers: CLSA-CIM: A Cross-Layer Scheduling Approach for Co…

This work introduces MICSim, an open-source, pre-circuit simulator designed for early-stage evaluation of chip-level software performance and hardware overhead of mixed-signal compute-in-memory (CIM) accelerators. MICSim features a modular…

The performance of today's in-memory indexes is bottlenecked by the memory latency/bandwidth wall. Processing-in-memory (PIM) is an emerging approach that potentially mitigates this bottleneck, by enabling low-latency memory access whose…

Large-scale AI training and inference require hundreds of gigabytes to terabytes of DRAM with high peak to average utilization ratios, resulting in overprovisioning. In cloud computing, DRAM constitutes a significant share of the cost. Yet,…

Optimizing CUDA code across multiple generations of GPU architectures is challenging, as achieving peak performance requires an extensive exploration of an increasingly complex, hardware-specific optimization space. Traditional compilers…

Benefiting from the self-attention mechanism, Transformer models have attained impressive contextual comprehension capabilities for lengthy texts. The requirements of high-throughput inference arise as the large language models (LLMs)…

General-purpose compilers abstract away parallelism, locality, and synchronization, limiting their effectiveness on modern spatial architectures. As modern computing architectures increasingly rely on fine-grained control over data…

RRAM-based multi-core systems improve the energy efficiency and performance of CNNs. Thereby, the distributed parallel execution of convolutional layers causes critical data dependencies that limit the potential speedup. This paper presents…

To usher in the next round of client AI innovation, there is an urgent need to enable efficient, lossless inference of high-accuracy large language models (LLMs) and vision language models (VLMs), jointly referred to as xLMs, on client…

CXL has been the emerging technology for expanding memory for both the host CPU and device accelerators with load/store interface. Extending memory coherency to the PCIe root complex makes the codesign more flexible in that you can access…

The full-size MLPs and the projection layers in attention introduce tremendous model sizes of large language models (LLMs), consuming extensive computational resources in pre-training. We empirically observe that the activations of…

The rapid growth of deep neural network (DNN) workloads has significantly increased the demand for large-capacity on-chip SRAM in machine learning (ML) applications, with SRAM arrays now occupying a substantial fraction of the total die…

In the era of LLMs, dense operations such as GEMM and MHA are critical components. These operations are well-suited for parallel execution using a tilebased approach. While traditional GPU programming often relies on low level interfaces…

Compute-in-memory (CIM) based neural network accelerators offer a promising solution to the Von Neumann bottleneck by computing directly within memory arrays. However, SRAM CIM faces limitations in executing larger models due to its cell…

Large language models (LLMs) rely on self-attention for contextual understanding, demanding high-throughput inference and large-scale token parallelism (LTPP). Existing dynamic sparsity accelerators falter under LTPP scenarios due to…

The SIMT execution model is commonly used for general GPU development. CUDA and OpenCL developers write scalar code that is implicitly parallelized by compiler and hardware. On Intel GPUs, however, this abstraction has profound performance…

Processing large-scale graph datasets is computationally intensive and time-consuming. Processor-centric CPU and GPU architectures, commonly used for graph applications, often face bottlenecks caused by extensive data movement between the…

The rapid growth of deep learning has driven exponential increases in model parameters and computational demands. NVIDIA GPUs and their CUDA-based software ecosystem provide robust support for parallel computing, significantly alleviating…

Processing in Memory (PIM) is a computing paradigm that promises enormous gain in processing speed by eradicating latencies in the typical von Neumann architecture. It has gained popularity owing to its throughput by embedding storage and…

Modern online large language model (LLM) services, such as Retrieval-Augmented Generation (RAG) and agent systems, increasingly expose two prominent characteristics: prompt segmentation (e.g., system instructions, retrieved passages, tool…

Crary and Sullivan's Relaxed Memory Calculus (RMC) proposed a new declarative approach for writing low-level shared memory concurrent programs in the presence of modern relaxed-memory multi-processor architectures and optimizing compilers.…