Compress objects, not cache lines: an object-based compressed memory hierarchy
The Morning Paper
MAY 23, 2019
Last time out we saw how Google have been able to save millions of dollars though memory compression enabled via zswap. Looking across a set of eight Java benchmarks, we find that only two of them are array dominated, the rest having between 40% to 75% of the heap footprint allocated to objects, the vast majority of which are small.
Let's personalize your content