Uncertainty quantification measures the prediction uncertainty of a neural network facing out-of-training-distribution samples. Bayesian Neural Networks (BNNs) can provide high-quality uncertainty quantification by in...
详细信息
ISBN:
(纸本)9781665420273
Uncertainty quantification measures the prediction uncertainty of a neural network facing out-of-training-distribution samples. Bayesian Neural Networks (BNNs) can provide high-quality uncertainty quantification by introducing specific noise to the weights during inference. To accelerate BNN inference, ReRAM processing-in-memory (PIM) architecture is a competitive solution to provide bothhigh-efficient computing and in-situ noise generation at the same time. However, there normally exists a huge gap between the generated noise in PIM hardware and that required by a BNN model. We demonstrate that the quality of uncertainty quantification is substantially degraded due to this gap. To solve this problem, we propose a holistic framework called W2W-PIM. We first introduce an efficient method to generate noise in ReRAM PIM design according to the demand of a BNN model. In addition, the PIM architecture is carefully modified to enable the noise generation and evaluate uncertainty quality. Moreover, a calibration unit is further introduced to reduce the noise gap caused by imperfection of the noise model. Comprehensive evaluation results demonstrate that W2W-PIM framework can achieve high-quality uncertainty quantification and high energy-efficiency at the same time.
Modern hybrid storage systems typically use SSD as the cache layer between memory and primary storage, and adopts inline deduplication to eliminate redundant I/Os and storage, thus delivering better performance and ca...
详细信息
Production high-performancecomputing (HPC) systems are adopting and integrating GPUs into their design to accommodate artificial intelligence (AI), machine learning, and data visualization workloads. To aid withthe ...
详细信息
ISBN:
(纸本)9781665420273
Production high-performancecomputing (HPC) systems are adopting and integrating GPUs into their design to accommodate artificial intelligence (AI), machine learning, and data visualization workloads. To aid withthe design and operations of new and existing GPU-based large-scale systems, we provide a detailed characterization of system operations, job characteristics, user behavior, and trends on a contemporary GPU-accelerated production HPC system. Our insights indicate that the pre-mature phases in modern AI workflow take up significant GPU hours while underutilizing GPUs, which opens up the opportunity for a multi-tier system. Finally, we provide various potential recommendations and areas for future investment for system architects, operators, and users.
As computerperformance advances and deep learning models proliferate, their applications expand, but concerns arise due to their opaque nature and security issues. Edge intelligence requires more from these models bu...
详细信息
the rapid development of intelligent transportation technology has promoted the progress of multiple trains cooperative technology. this paper proposes an online cooperative cruise control method based on improved par...
详细信息
the emergence of quantum computers as a new computational paradigm has been accompanied by speculation concerning the scope and timeline of their anticipated revolutionary changes. While quantum computing is still in ...
详细信息
ISBN:
(纸本)9781665420273
the emergence of quantum computers as a new computational paradigm has been accompanied by speculation concerning the scope and timeline of their anticipated revolutionary changes. While quantum computing is still in its infancy, the variety of different architectures used to implement quantum computations make it difficult to reliably measure and compare performance. this problem motivates our introduction of SupermarQ, a scalable, hardware-agnostic quantum benchmark suite which uses application-level metrics to measure performance. SupermarQ is the first attempt to systematically apply techniques from classical benchmarking methodology to the quantum domain. We define a set of feature vectors to quantify coverage, select applications from a variety of domains to ensure the suite is representative of real workloads, and collect benchmark results from the IBM, IonQ, and AQT@LBNL platforms. Looking forward, we envision that quantum benchmarking will encompass a large cross-community effort built on open source, constantly evolving benchmark suites. We introduce SupermarQ as an important step in this direction.
>A multiprocessor system is a computer system with multiple processors that are connected to solve specific problems. However, processor failures are inevitable in multiprocessor systems, which seriously affect the...
详细信息
Withthe development of network technology and the blockchain, the metaverse as an emerging network paradigm has received extensive attention. In metaverse, users can create multiple virtual avatars to obtain differen...
详细信息
this work presents the development of a new ultra-compact implantable two-element MIMO antenna with radiation diversity for bio-telemetry communication in scalp-based implants. the design consists of two separated mir...
详细信息
computer vision has been increasingly applied in the study of fish behavior. However, accurate and robust detection and tracking of fish schools remains a challenging problem due to appearance changes and occlusion du...
详细信息
暂无评论