TPU-Gen: LLM-Driven Custom Tensor Processing Unit Generator

• URL: http://arxiv.org/abs/2503.05951v1
• Date: Fri, 07 Mar 2025 21:41:42 GMT
• Title: TPU-Gen: LLM-Driven Custom Tensor Processing Unit Generator
• Authors: Deepak Vungarala, Mohammed E. Elbtity, Sumiya Syed, Sakila Alam, Kartik Pandit, Arnob Ghosh, Ramtin Zand, Shaahin Angizi,
• Abstract summary: This paper introduces TPU-Gen, the first Large Language Model (LLM) based framework designed to automate the exact and approximate TPU generation process.<n>TPU-Gen is supported with a meticulously curated, comprehensive, and open-source dataset that covers a wide range of spatial array designs and approximate multiply-and-accumulate units.
• Score: 4.479077825955557
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The increasing complexity and scale of Deep Neural Networks (DNNs) necessitate specialized tensor accelerators, such as Tensor Processing Units (TPUs), to meet various computational and energy efficiency requirements. Nevertheless, designing optimal TPU remains challenging due to the high domain expertise level, considerable manual design time, and lack of high-quality, domain-specific datasets. This paper introduces TPU-Gen, the first Large Language Model (LLM) based framework designed to automate the exact and approximate TPU generation process, focusing on systolic array architectures. TPU-Gen is supported with a meticulously curated, comprehensive, and open-source dataset that covers a wide range of spatial array designs and approximate multiply-and-accumulate units, enabling design reuse, adaptation, and customization for different DNN workloads. The proposed framework leverages Retrieval-Augmented Generation (RAG) as an effective solution for a data-scare hardware domain in building LLMs, addressing the most intriguing issue, hallucinations. TPU-Gen transforms high-level architectural specifications into optimized low-level implementations through an effective hardware generation pipeline. Our extensive experimental evaluations demonstrate superior performance, power, and area efficiency, with an average reduction in area and power of 92\% and 96\% from the manual optimization reference values. These results set new standards for driving advancements in next-generation design automation tools powered by LLMs.

Related papers

• LIFT: LLM-Based Pragma Insertion for HLS via GNN Supervised Fine-Tuning [38.679497621876926]
  LIFT is a large language model (LLM)-based coding assistant for HLS that automatically generates performance-critical pragmas. We fine-tune the LLM by tightly integrating and supervising the training process with a graph neural network (GNN)
  arXiv  Detail & Related papers  (2025-04-29T21:42:59Z)
• ZeroLM: Data-Free Transformer Architecture Search for Language Models [54.83882149157548]
  Current automated proxy discovery approaches suffer from extended search times, susceptibility to data overfitting, and structural complexity. This paper introduces a novel zero-cost proxy methodology that quantifies model capacity through efficient weight statistics. Our evaluation demonstrates the superiority of this approach, achieving a Spearman's rho of 0.76 and Kendall's tau of 0.53 on the FlexiBERT benchmark.
  arXiv  Detail & Related papers  (2025-03-24T13:11:22Z)
• IMPROVE: Iterative Model Pipeline Refinement and Optimization Leveraging LLM Agents [17.301758094000125]
  Large language model (LLM) agents have emerged as a promising solution to automate the development of computer vision models.<n>We introduce Iterative Refinement, a novel strategy for LLM-driven ML pipeline design.<n>Iterative Refinement improves stability, interpretability, and overall model performance.
  arXiv  Detail & Related papers  (2025-02-25T01:52:37Z)
• Tackling the Dynamicity in a Production LLM Serving System with SOTA Optimizations via Hybrid Prefill/Decode/Verify Scheduling on Efficient Meta-kernels [12.77187564450236]
  We introduce XY-Serve, a versatile, Ascend native, end-to-end production large language model (LLM) serving system.<n>The core idea is an abstraction mechanism that smooths out the workload variability by decomposing computations into fine-grained meta primitives.<n>For GEMM, we introduce a virtual padding scheme that adapts to dynamic shape changes while using highly efficient GEMM primitives with assorted fixed tile sizes.
  arXiv  Detail & Related papers  (2024-12-24T02:27:44Z)
• Automatically Learning Hybrid Digital Twins of Dynamical Systems [56.69628749813084]
  Digital Twins (DTs) simulate the states and temporal dynamics of real-world systems. DTs often struggle to generalize to unseen conditions in data-scarce settings. In this paper, we propose an evolutionary algorithm ($textbfHDTwinGen$) to autonomously propose, evaluate, and optimize HDTwins.
  arXiv  Detail & Related papers  (2024-10-31T07:28:22Z)
• Task-Oriented Real-time Visual Inference for IoVT Systems: A Co-design Framework of Neural Networks and Edge Deployment [61.20689382879937]
  Task-oriented edge computing addresses this by shifting data analysis to the edge. Existing methods struggle to balance high model performance with low resource consumption. We propose a novel co-design framework to optimize neural network architecture.
  arXiv  Detail & Related papers  (2024-10-29T19:02:54Z)
• EPS-MoE: Expert Pipeline Scheduler for Cost-Efficient MoE Inference [49.94169109038806]
  This paper introduces EPS-MoE, a novel expert pipeline scheduler for MoE that surpasses the existing parallelism schemes.<n>Our results demonstrate at most 52.4% improvement in prefill throughput compared to existing parallel inference methods.
  arXiv  Detail & Related papers  (2024-10-16T05:17:49Z)
• POMONAG: Pareto-Optimal Many-Objective Neural Architecture Generator [4.09225917049674]
  Transferable NAS has emerged, generalizing the search process from dataset-dependent to task-dependent. This paper introduces POMONAG, extending DiffusionNAG via a many-optimal diffusion process. Results were validated on two search spaces -- NAS201 and MobileNetV3 -- and evaluated across 15 image classification datasets.
  arXiv  Detail & Related papers  (2024-09-30T16:05:29Z)
• Search for Efficient Large Language Models [52.98684997131108]
  Large Language Models (LLMs) have long held sway in the realms of artificial intelligence research. Weight pruning, quantization, and distillation have been embraced to compress LLMs, targeting memory reduction and inference acceleration. Most model compression techniques concentrate on weight optimization, overlooking the exploration of optimal architectures.
  arXiv  Detail & Related papers  (2024-09-25T21:32:12Z)
• Flex-TPU: A Flexible TPU with Runtime Reconfigurable Dataflow Architecture [0.0]
  The work herein consists of developing a reconfigurable dataflow TPU, called the Flex-TPU, which can dynamically change the dataflow per layer during run-time. The results show that our Flex-TPU design achieves a significant performance increase of up to 2.75x compared to conventional TPU, with only minor area and power overheads.
  arXiv  Detail & Related papers  (2024-07-11T17:33:38Z)
• Mechanistic Design and Scaling of Hybrid Architectures [114.3129802943915]
  We identify and test new hybrid architectures constructed from a variety of computational primitives. We experimentally validate the resulting architectures via an extensive compute-optimal and a new state-optimal scaling law analysis. We find MAD synthetics to correlate with compute-optimal perplexity, enabling accurate evaluation of new architectures.
  arXiv  Detail & Related papers  (2024-03-26T16:33:12Z)
• Energy-efficient Task Adaptation for NLP Edge Inference Leveraging Heterogeneous Memory Architectures [68.91874045918112]
  adapter-ALBERT is an efficient model optimization for maximal data reuse across different tasks. We demonstrate the advantage of mapping the model to a heterogeneous on-chip memory architecture by performing simulations on a validated NLP edge accelerator.
  arXiv  Detail & Related papers  (2023-03-25T14:40:59Z)
• Hardware-Efficient Deconvolution-Based GAN for Edge Computing [1.5229257192293197]
  Generative Adversarial Networks (GAN) are cutting-edge algorithms for generating new data samples based on the learned data distribution. We proposed an HW/SW co-design approach for training quantized deconvolution GAN (QDCGAN) implemented on FPGA using a scalable streaming dataflow architecture. Various precisions, datasets, and network scalability were analyzed for low-power inference on resource-constrained platforms.
  arXiv  Detail & Related papers  (2022-01-18T11:16:59Z)
• Reconfigurable Intelligent Surface Assisted Mobile Edge Computing with Heterogeneous Learning Tasks [53.1636151439562]
  Mobile edge computing (MEC) provides a natural platform for AI applications. We present an infrastructure to perform machine learning tasks at an MEC with the assistance of a reconfigurable intelligent surface (RIS) Specifically, we minimize the learning error of all participating users by jointly optimizing transmit power of mobile users, beamforming vectors of the base station, and the phase-shift matrix of the RIS.
  arXiv  Detail & Related papers  (2020-12-25T07:08:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.