Related papers: ChipLight: Cross-Layer Optimization of Chiplet Des…

Overlapping communication with computation is crucial for distributed large-model training, yet optimizing it - especially when computation becomes the bottleneck-remains challenging. We present Lagom, a system that co-tunes communication…

Large language models (LLMs), based on transformer architectures, have revolutionized numerous domains within artificial intelligence, science, and engineering due to their exceptional scalability and adaptability. However, the exponential…

The proliferation of Large Language Models (LLMs) with exponentially growing parameters is making cross-data center (DC) training an inevitable trend. However, viable strategies for extending single-DC training frameworks to multi-DC…

Nowadays, large and complex deep learning (DL) models are increasingly trained in a distributed manner across multiple worker machines, in which extensive communications between workers pose serious scaling problems. In this article, we…

Vision-language pre-training like CLIP has shown promising performance on various downstream tasks such as zero-shot image classification and image-text retrieval. Most of the existing CLIP-alike works usually adopt relatively large image…

The emergence of Superchips represents a significant advancement in next-generation AI hardware. These Superchips employ a tightly coupled heterogeneous architecture that integrates GPU and CPU on the same package, which offers…

The field of integrated circuit (IC) design is highly specialized, presenting significant barriers to entry and research and development challenges. Although large language models (LLMs) have achieved remarkable success in various domains,…

Advanced 2.5D Systems-in-Package (SiPs) compose a growing portion of high-performance systems. While the packaging and interconnect choices play a large role in the overall system design, system architects still lack a suitable framework…

With the recent growth in demand for large-scale deep neural networks, compute in-memory (CiM) has come up as a prominent solution to alleviate bandwidth and on-chip interconnect bottlenecks that constrain Von-Neuman architectures. However,…

In the near future, Internet-of-Things (IoT) is expected to connect billions of devices (e.g., smartphones and sensors), which generate massive real-time data at the network edge. Intelligence can be distilled from the data to support…

Heterogeneous chiplets have been proposed for accelerating high-performance computing tasks. Integrated inside one package, CPU and GPU chiplets can share a common interconnection network that can be implemented through the interposer.…

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,…

The proliferation of large language models (LLMs) is accelerating the integration of multimodal assistants into edge devices, where inference is executed under stringent latency and energy constraints, often exacerbated by intermittent…

Driven by massive investments and consequently significant progresses in optical computing and all-optical signal processing technologies lately, this paper presents a new architectural paradigm for next-generation optical transport…

Mixture-of-Expert (MoE) models outperform conventional models by selectively activating different subnets, named experts, on a per-token basis. This gated computation generates dynamic communications that cannot be determined beforehand,…

Battery-powered wireless IoT devices are now widely seen in many critical applications. Given the limited battery capacity and inaccessibility to external power recharge, optimizing energy efficiency (EE) plays a vital role in prolonging…

Existing studies of training state-of-the-art Contrastive Language-Image Pretraining (CLIP) models on large-scale data involve hundreds of or even thousands of GPUs due to the requirement of a large batch size. However, such a large amount…

The emergence of Large Language Models (LLMs) has necessitated the adoption of distributed training techniques, involving the deployment of thousands of GPUs to train a single model. Unfortunately, the efficiency of large-scale distributed…

Mixture-of-Experts (MoE) architecture offers enhanced efficiency for Large Language Models (LLMs) with modularized computation, yet its inherent sparsity poses significant hardware deployment challenges, including memory locality issues,…

In this work, we propose a novel technique to boost training efficiency of a neural network. Our work is based on an excellent idea that whitening the inputs of neural networks can achieve a fast convergence speed. Given the well-known fact…