Compute Unit
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Memristor-Based Spiking Neural Network Accelerator for Bio-inspired Interception Task
arXiv:2605.31299v1 Announce Type: new Abstract: Spiking neural networks (SNNs) provide event-driven and low-power computation inspired by biological neural systems, but current implementations rely on von Neumann graphics processing units (GPUs) and central processing units (CPUs) platforms, where memory and computation bottlenecks limit energy efficiency. To address this challenge, this paper proposes an analog memristor-based spiking neural network (SNN) accelerator that integrates...
Pinterest deepens Amazon partnership with $4 billion cloud deal
Pinterest deepens Amazon partnership with $4 billion cloud deal June 4 : Pinterest said on Thursday it would pay Amazon Web Services $4 billion for cloud services through 2031, as the social media company strengthens a long-term partnership with its largest-ever deal. Shares of Pinterest were up nearly 5 per cent, while those of Amazon rose 1.7 per cent. Amazon.com's cloud computing unit will provide Pinterest its custom chip processors, including Graviton and Trainium, to help scale its AI...
Accelerating Bidiagonalization of Banded Matrices through Memory-Aware Bulge-Chasing on GPUs
arXiv:2510.12705v3 Announce Type: replace Abstract: The reduction of a banded matrix to bidiagonal form is a critical step in the calculation of Singular Values, a cornerstone of scientific computing and AI. Although inherently parallel, this step has traditionally been considered unsuitable for GPUs due to its memory-bound nature. However, recent advances in GPU architectures, such as increased L1 memory per Streaming Multiprocessor or Compute Unit and larger L2 caches, have shifted this...
Google Cloud outage in India after third-party data centre fire triggers shutdown
Google Cloud outage in India after third-party data centre fire triggers shutdown June 9 : Alphabet's Google Cloud said on Tuesday that some customers in India experienced intermittent network disruptions after a fire at a third-party data centre triggered an emergency shutdown of networking equipment. The cloud-computing unit said the fire led to an emergency power shutdown at the facility, isolating a local point of presence in Delhi and reducing network capacity across the metropolitan...
ArrowFlow: Hierarchical Machine Learning in the Space of Permutations
Announce Type: replace Abstract: We introduce ArrowFlow, a machine learning architecture that operates entirely in the space of permutations. Its computational units are ranking filters, learned orderings that compare inputs via Spearman's footrule distance and update through permutation-matrix accumulation, a non-gradient rule rooted in displacement evidence. Layers compose hierarchically: each layer's output ranking becomes the next layer's input, enabling deep ordinal representation...
Heterogeneous Mapping for Analog In-Memory Computing Accelerators: A Unified Workflow
arXiv:2606.02672v1 Announce Type: new Abstract: Analog In-Memory Computing (AIMC) accelerators execute matrix-vector multiplications directly within memory arrays, reducing data movement and improving DNN inference efficiency. Their limited effective precision motivates heterogeneous architectures that combine analog compute tiles with digital processing units. This letter classifies existing methods for partitioning DNN workloads across these resources by mapping granularity, optimization...
HyperParallel-MoE: Multi-Core Interleaved Scheduling for Fast MoE Training on Ascend NPUs
arXiv:2605.23764v2 Announce Type: replace Abstract: Modern Mixture-of-Experts (MoE) models increasingly rely on large-scale AI accelerator clusters for efficient training. Ascend NPUs expose heterogeneous on-chip compute resources, including matrix-oriented AIC units and vector-oriented AIV units with explicit cross-queue synchronization support. However, existing training frameworks largely execute MoE operators in a serialized kernel-by-kernel manner, leaving substantial heterogeneous...
How Much Progress Has There Been in NVIDIA Datacenter GPUs?
Announce Type: replace Abstract: As the role of modern Graphics Processing Units (GPUs) becomes increasingly essential for several computing tasks, analyzing their past and current progress is paramount for determining future constraints on scientific research. This is particularly compelling in the Artificial Intelligence (AI) domain, where rapid technological advancements and fierce global competition have led the United States to recently implement export control regulations limiting...
Accuracy-Configurable Floating-Point Multiplier Design for SRAM-Based Compute-in-Memory
arXiv:2606.08430v1 Announce Type: new Abstract: Digital Compute-in-Memory (DCiM) reduces data movement and has become a promising solution for energy-efficient edge AI. However, most existing DCiM frameworks still primarily target integer or fixed-point arithmetic, and provide limited support for compiler-integrated and accuracy-configurable floating-point computation. Directly integrating conventional IEEE 754 floating-point units into dense SRAM-based DCiM arrays, however, incurs high area...
IN2P3 Computing Center 2024 Workload Dataset
Announce Type: new Abstract: This paper provides and analyzes a dataset detailing the characteristics and execution data of all jobs submitted to the IN2P3 Computing Center (Villeurbanne, France), a national research and support unit of the CNRS, in 2024. The main additional value of this contribution compared to previously available datasets consists in the combination of an extended time interval considered, the inclusion of memory usage data and its recency, on top on improving the...