Communications of the ACM
For Better Computing, Liberate CPs From Garbage Collection
A team of University of California, Berkeley researchers have devised a compact accelerator that can relieve a central processing unit from needing to identify and delete redundant or irrelevant data from apps to conserve memory.
Credit: Ivan Kuznetsov
Among other things, central processing units (CPUs) manage a "garbage collection" process of identifying and deleting redundant or irrelevant data from apps so computers have more memory available, taking up 10% or more of the total time a CPU spends on an app.
Former University of California, Berkeley researcher Martin Maas and colleagues have engineered a compact accelerator to relieve the CPU of this garbage collection task.
The accelerator, which can be added to the CPU, runs on a small amount of chip area and power, and can handle garbage collection more efficiently by performing a larger number of memory operations concurrently.
The device far outperformed a CPU core, and cut the amount of power needed for garbage collection by 15%.
Maas said, "This means that you could build a system where the software does not have to worry about garbage collection at all and just keeps using the available memory."
From IEEE Spectrum
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA
No entries found