Modern GPUs reach record analytical performance only when a query's working set fits within its GPU memory budget, and that budget is under growing pressure: GPUs are partitioned and shared, working sets are outgrowing high-bandwidth memory (HBM), and HBM is a fixed on-device resource whose per-dollar capacity has stag...
Antonio Boffa, Ananya Gupta, Hamish Nicholson et al.· 0 citations
The growing disparity between processor core scaling and memory bandwidth has exposed the physical and economic limits of processor-centric database architectures. While Compute Express Link (CXL) and other emerging technologies enable a necessary shift toward memory-centric, disaggregated topologies, it also introdu...
Yi Jiang, Hamish Nicholson, Anastasia Ailamaki· Datenbank-Spektrum· 0 citations
It is found that hardware offload is not a universal replacement for CPU compression, and compression should be scheduled dynamically: route blocks by codec, operation, size, and accelerator load; cap per-device sub-mission concurrency; and fall back to software when offload is unsupported or saturated.
Yi Jiang, Antonio Boffa, Hamish Nicholson et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.