CN101 —— 用于生成式AI的数字热力学计算机

中文摘要 AI 辅助

热力学计算是一种新兴的硬件范式，其中随机物理动力学直接作为计算的基本单元。生成式AI的近期爆发式增长，使得人们对替代计算方法的探索愈发迫切，而本研究表明，热力学计算非常适合这一领域。一类重要的方法将函数实现为遍历随机过程的平稳期望：答案编码在平衡轨迹的时间平均统计中。迄今为止，这类平衡式方法仅通过朗之万动力学实现，将其实现局限于模拟基底及相关工程挑战。本研究提出了平衡式表述的一种与基底无关的形式化方法，其中唯一的设计对象是任意遍历过程的动力学生成元L*。该形式化方法明确了表述的三个硬件级特性：结果的精度是可调节的参数，由动力学运行时长决定；样本平均值可在独立轨迹间分解；计算的依赖阶段可并行而非串行操作，这一特性被称为顺序并行性。我们通过制造名为CN101的原型数字热力学计算芯片，实例化了该形式化方法，该芯片基于标准CMOS工艺，利用随机计算原理，通过离散累加器动力学实现了该表述。我们在以变分自编码器（VAE）和流匹配形式呈现的传统生成式AI工作负载上，对CN101在图像生成及科学问题上的表现进行了表征。综上，该形式化方法及其数字实例表明，平衡式表述与基底无关，且其计算特性可在标准数字硬件上得到利用。

英文摘要

Thermodynamic computing is an emerging hardware paradigm, in which stochastic physical dynamics serve as the direct computational primitive. The recent explosion of generative AI has only sharpened the search for alternative approaches to compute, and, as we show in this work, thermodynamic computing turns out to be well suited to this space. An important class of methods realises a function as the stationary expectation of an ergodic stochastic process: the answer is encoded in the time-averaged statistics of an equilibrating trajectory. To date, this equilibration-style class has been formulated exclusively through Langevin dynamics, restricting its implementations to analogue substrates and the engineering challenges those bring. In this work, we propose a substrate-independent formalisation of the equilibration-style formulation, in which the only object of design is the dynamical generator L* of an arbitrary ergodic process. The formalisation makes three hardware-level properties of the formulation explicit: the precision of a result is a knob set by how long the dynamics are run, sample averages decompose across independent trajectories, and dependent stages of a computation operate concurrently rather than serially, a property we call sequential parallelism. We instantiate the formalisation by fabricating a prototype digital thermodynamic computing chip, named CN101, that implements the formulation through discrete accumulator dynamics on standard CMOS using stochastic computing principles. We characterise CN101's success across conventional generative AI workloads in the form of VAEs and flow matching, applied to both image generation and scientific problems. Together, the formalisation and its digital instantiation show that the equilibration-style formulation is substrate-independent, and that its computational properties can be exploited on standard digital hardware.

发表机构

• Normal Computing Corporation（诺玛计算公司）
• University of Oxford（牛津大学）
• University of Amsterdam（阿姆斯特丹大学）
• National Taiwan University（台湾大学）
• Academia Sinica（中央研究院）

机构由 AI 辅助整理，请以论文原文为准。