MC-CIM: Compute-in-Memory With Monte-Carlo Dropouts for Bayesian Edge Intelligence
MC-CIM: Compute-in-Memory With Monte-Carlo Dropouts for Bayesian Edge Intelligence
复制标题
MC-CIM：具有蒙特卡罗辍学的内存计算用于贝叶斯边缘智能
DOI：
10.1109/tcsi.2022.3224703
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Trivedi, Amit Ranjan
中科院分区：
文献类型：
--
作者：
Shukla, Priyesh;Nasrin, Shamma;Darabi, Nastaran;Gomes, Wilfred;Trivedi, Amit Ranjan
我们提出了MC-CIM，一个计算内存（CIM）框架的强大，但低功耗，贝叶斯边缘智能。具有确定性权重的深度神经网络（DNN）无法表达其预测的不确定性，因此对预测错误的后果是致命的应用（如手术机器人）构成了关键风险。为了解决这一限制，DNN的贝叶斯推理得到了关注。使用贝叶斯推理，不仅可以提取预测本身，还可以提取预测置信度，用于规划风险感知行动。然而，DNN的贝叶斯推理在计算上是昂贵的，不适合实时和/或边缘部署。使用蒙特卡罗丢弃（MC-丢弃）对贝叶斯DNN的近似已经显示出高鲁棒性沿着低计算复杂度。为了提高该方法的计算效率，我们讨论了一种新的CIM模块，除了在内存中的权重输入标量积，可以执行内存中的概率丢弃，以支持该方法。我们还提出了一个计算重用的MC-Dropout的重新表述，其中每个连续的实例可以利用前一次迭代的乘积和计算。此外，我们还讨论了如何利用组合优化方法对随机实例进行优化排序，以最大限度地减少整体MC-Dropout工作量。提出的基于CIM的MC-Dropout执行的应用程序进行了讨论的MNIST字符识别和视觉里程计（VO）的自主无人机。该框架可靠地提供了预测的信心，在MC-CIM施加的非理想性在很大程度上。提出的MC-CIM采用SRAM阵列，0.85 V电源，16 nm低待机功耗（LSTP）技术，在其最佳计算和外围配置下，30个MC-Dropout概率推理实例消耗32 pJ，与典型执行相比节省%的能源。
We propose MC-CIM, a compute-in-memory (CIM) framework for robust, yet low power, Bayesian edge intelligence. Deep neural networks (DNN) with deterministic weights cannot express their prediction uncertainties, thereby pose critical risks for applications where the consequences of mispredictions are fatal such as surgical robotics. To address this limitation, Bayesian inference of a DNN has gained attention. Using Bayesian inference, not only the prediction itself, but the prediction confidence can also be extracted for planning risk-aware actions. However, Bayesian inference of a DNN is computationally expensive, ill-suited for real-time and/or edge deployment. An approximation to Bayesian DNN using Monte Carlo Dropout (MC-Dropout) has shown high robustness along with low computational complexity. Enhancing the computational efficiency of the method, we discuss a novel CIM module that can perform in-memory probabilistic dropout in addition to in-memory weight-input scalar product to support the method. We also propose a compute-reuse reformulation of MC-Dropout where each successive instance can utilize the product-sum computations from the previous iteration. Even more, we discuss how the random instances can be optimally ordered to minimize the overall MC-Dropout workload by exploiting combinatorial optimization methods. Application of the proposed CIM-based MC-Dropout execution is discussed for MNIST character recognition and visual odometry (VO) of autonomous drones. The framework reliably gives prediction confidence amidst non-idealities imposed by MC-CIM to a good extent. Proposed MC-CIM withSRAM array, 0.85 V supply, 16nm low-standby power (LSTP) technology consumes 32 pJ for 30 MC-Dropout instances of probabilistic inference in its most optimal computing and peripheral configuration, saving% energy compared to typical execution.
登录
查看更多内容
MC2RAM：SRAM 中的马尔可夫链蒙特卡罗采样用于快速贝叶斯推理
DOI：
--
发表时间：
2020
期刊：
International Symposium on Circuits and Systems
影响因子：
--
作者：
Priyesh Shukla;A. Shylendra;Theja Tulabandhula;A. Trivedi
通讯作者：
A. Trivedi
DOI：
10.1109/tcsi.2021.3064033
发表时间：
2021-01
期刊：
IEEE Transactions on Circuits and Systems I: Regular Papers
影响因子：
--
作者：
Shamma Nasrin;Diaa Badawi;A. Cetin;Wilfred Gomes;A. Trivedi
通讯作者：
Shamma Nasrin;Diaa Badawi;A. Cetin;Wilfred Gomes;A. Trivedi
DOI：
10.1109/iccad45719.2019.8942149
发表时间：
2019-11
期刊：
2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD)
影响因子：
--
作者：
Yannan Nellie Wu;J. Emer;V. Sze
通讯作者：
Yannan Nellie Wu;J. Emer;V. Sze
基于 SRAM 的多位内存矩阵矢量乘法器，其精度可在面积、时间和功耗上线性缩放
影响因子：
2.8
作者：
R. Khaddam;P. Francese;L. Benini;E. Eleftheriou
通讯作者：
E. Eleftheriou
影响因子：
7.8
作者：
Chukewad, Yogesh M.;James, Johannes;Fuller, Sawyer
通讯作者：
Fuller, Sawyer