Zobrazeno 1 - 10
of 230
pro vyhledávání: '"Keckler, Stephen W."'
Recently, large language models (LLMs) have shown surprising performance in task-specific workloads as well as general tasks with the given prompts. However, to achieve unprecedented performance, recent LLMs use billions to trillions of parameters, w
Externí odkaz:
http://arxiv.org/abs/2406.13868
This paper focuses on mitigating DRAM Rowhammer attacks. In recent years, solutions like TRR have been deployed in DDR4 DRAM to track aggressor rows and then issue a mitigative action by refreshing neighboring victim rows. Unfortunately, such in-DRAM
Externí odkaz:
http://arxiv.org/abs/2404.16256
Exploiting sparsity in deep neural networks (DNNs) has been a promising area to meet the growing computation need of modern DNNs. However, in practice, sparse DNN acceleration still faces a key challenge. To minimize the overhead of sparse accelerati
Externí odkaz:
http://arxiv.org/abs/2403.07953
Autor:
Hsiao, Yu-Shun, Hari, Siva Kumar Sastry, Sundaralingam, Balakumar, Yik, Jason, Tambe, Thierry, Sakr, Charbel, Keckler, Stephen W., Reddi, Vijay Janapa
High-dimensional motion generation requires numerical precision for smooth, collision-free solutions. Typically, double-precision or single-precision floating-point (FP) formats are utilized. Using these for big tensors imposes a strain on the memory
Externí odkaz:
http://arxiv.org/abs/2310.07854
Autor:
Sreedhar, Kavya, Clemons, Jason, Venkatesan, Rangharajan, Keckler, Stephen W., Horowitz, Mark
Publikováno v:
2024 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)
State-of-the-art deep learning models for computer vision tasks are based on the transformer architecture and often deployed in real-time applications. In this scenario, the resources available for every inference can vary, so it is useful to be able
Externí odkaz:
http://arxiv.org/abs/2212.02687
Autor:
Hsiao, Yu-Shun, Hari, Siva Kumar Sastry, Filipiuk, Michał, Tsai, Timothy, Sullivan, Michael B., Reddi, Vijay Janapa, Singh, Vasu, Keckler, Stephen W.
The processing requirement of autonomous vehicles (AVs) for high-accuracy perception in complex scenarios can exceed the resources offered by the in-vehicle computer, degrading safety and comfort. This paper proposes a sensor frame processing rate (F
Externí odkaz:
http://arxiv.org/abs/2205.03347
As GPUs scale their low precision matrix math throughput to boost deep learning (DL) performance, they upset the balance between math throughput and memory system capabilities. We demonstrate that converged GPU design trying to address diverging arch
Externí odkaz:
http://arxiv.org/abs/2104.02188
Autor:
Ghodsi, Zahra, Hari, Siva Kumar Sastry, Frosio, Iuri, Tsai, Timothy, Troccoli, Alejandro, Keckler, Stephen W., Garg, Siddharth, Anandkumar, Anima
Extracting interesting scenarios from real-world data as well as generating failure cases is important for the development and testing of autonomous systems. We propose efficient mechanisms to both characterize and generate testing scenarios using a
Externí odkaz:
http://arxiv.org/abs/2103.07403
The ability of Convolutional Neural Networks (CNNs) to accurately process real-time telemetry has boosted their use in safety-critical and high-performance computing systems. As such systems require high levels of resilience to errors, CNNs must exec
Externí odkaz:
http://arxiv.org/abs/2006.04984
Autor:
Hari, Siva Kumar Sastry, Rech, Paolo, Tsai, Timothy, Stephenson, Mark, Zulfiqar, Arslan, Sullivan, Michael, Shirvani, Philip, Racunas, Paul, Emer, Joel, Keckler, Stephen W.
High-performance and safety-critical system architects must accurately evaluate the application-level silent data corruption (SDC) rates of processors to soft errors. Such an evaluation requires error propagation all the way from particle strikes on
Externí odkaz:
http://arxiv.org/abs/2005.01445