Imagine the possibility of doubling the processing power of your smartphone, tablet, personal computer, or server using the existing hardware already in these devices. This might soon become a reality, thanks to a recent breakthrough in computer architecture.
Associate Professor Hung-Wei Tseng, from UC Riverside's Department of Electrical and Computer Engineering, has proposed a novel paradigm shift in computer architecture, capable of greatly enhancing processing speeds. The basis of this groundbreaking discovery revolves around the concept of "simultaneous and heterogeneous multithreading” or SHMT.
As computer devices increasingly utilize components such as graphics processing units (GPUs), hardware accelerators for artificial intelligence (AI) and machine learning (ML), and digital signal processing units, these components process information separately, transitioning information from one processing unit to the subsequent one. This sequential transition effectually creates a bottleneck.
To overcome this hurdle, Tseng, in collaboration with UCR computer science graduate student Kuan-Chieh Hsu, introduced their concept of "simultaneous and heterogeneous multithreading," or SHMT. The duo described the development of their proposed SHMT framework on an embedded system platform. Surprisingly, the system simultaneously uses a multi-core ARM processor, an NVIDIA GPU, and a Tensor Processing Unit hardware accelerator, essentially bypassing any bottleneck situation.
Quite impressively, this system achieved a 1.96 times speedup and a noteworthy 51% reduction in energy consumption. "You don't have to add new processors because you already have them," Tseng pointed out, highlighting the potential for cost savings with the new approach.
The implications of this innovation are vast and transformative. Simultaneous utilization of existing processing components not only promises a significant cut in computer hardware cost, but it might also lead to a decrease in carbon emissions from the energy produced to keep servers running in vast data processing centers. Additionally, the need for the increasingly scarce freshwater used extensively to cool servers could also be curtailed.
However, Tseng's paper does warn that more research is required. Questions regarding system implementation, hardware support, and code optimization are still to be addressed. Also, it remains to be seen which types of applications stand to gain most from this breakthrough.
This compelling paper was acknowledged at the 56th Annual IEEE/ACM International Symposium on Microarchitecture in Toronto, Canada. Tseng's paper receiving recognition among his professional peers in the Institute of Electrical and Electronics Engineers (IEEE), who selected it as one of 12 significant papers in the "Top Picks from the Computer Architecture Conferences" issue set for publication in summer 2024.
Disclaimer: The above article was written with the assistance of AI. The original sources can be found on ScienceDaily.
