Imagine doubling the processing power of your smartphone, tablet, personal computer, or server using the hardware already in these devices.

Hung-Wei Tseng, UC Riverside associate professor of electrical and computer engineering, presents a paradigm shift in computer architecture in his recent paper titled “Concurrent and Asynchronous Multithreading.”

Today’s computing devices increasingly have graphics processing units (GPUs), hardware accelerators for artificial intelligence (AI) and machine learning (ML), or digital signal processing units as essential components, Tseng explained. These components process information separately, passing information from one processing unit to another, which in effect creates a bottleneck.

In their paper, Tseng and UCR computer science graduate student Kuan-Chieh Hsu introduce what they call “simultaneous and heterogeneous multithreading,” or SHMT. They describe the development of a proposed SHMT framework on an embedded system platform that simultaneously uses a multicore ARM processor, an NVIDIA GPU, and a tensor processing unit hardware accelerator.

The system achieved 1.96 times speedup and reduced energy consumption by 51%.

“You don’t need to add new processors because you already have them,” Tseng said.

The implications are huge.

The simultaneous use of existing processing components can reduce computer hardware costs while also reducing carbon emissions from the energy generated to run servers in warehouse-sized data processing centers. It can also reduce the need for scarce freshwater used to keep servers cool.

However, Tseng’s paper cautions that more investigation is needed to answer several questions about what types of applications are most beneficial, including system implementation, hardware support, code optimization, and other issues.

This paper was presented at the 56th Annual IEEE/ACM International Symposium on Microarchitecture, held in October in Toronto, Canada. The paper won recognition from Tseng’s professional colleagues at the Institute of Electrical and Electronics Engineers, or IEEE, who included it in the group’s “Top Picks from Computer Architecture Conferences” issue to be published this coming summer. Selected as one of the 12 papers.