We show that ABACuS's performance scales well with the number of DRAM banks. At a future RowHammer threshold of 125, ABACuS performs very similarly to (within 0.38% of the performance of) the best prior performance- and energy-efficient RowHammer mitigation mechanism while requiring 22.72× smaller chip area. At the RowHammer threshold of 1000, the best prior lowarea-cost mitigation mechanism incurs 1.80% higher average performance overhead than ABACuS, while ABACuS requires 2.50× smaller chip area to implement. At a nearfuture RowHammer threshold of 1000, ABACuS incurs only 0.58% (0.77%) performance and 1.66% (2.12%) DRAM energy overheads, averaged across 62 single-core (8-core) workloads, requiring only 9.47 KiB of storage per DRAM rank. We compare ABACuS to four state-of-the-art mitigation mechanisms. Our comprehensive evaluations show that ABACuS securely prevents RowHammer bitflips at low performance/energy overhead and low area cost. Unlike state-of-the-art RowHammer mitigation mechanisms that implement a separate row activation counter for each DRAM bank, ABACuS implements fewer counters (e.g., only one) to track an equal number of aggressor rows. Based on this observation, ABACuS's key idea is to use a single shared row activation counter to track activations to the rows with the same row address in all DRAM banks. We observe that both benign workloads and RowHammer attacks tend to access DRAM rows with the same row address in multiple DRAM banks at around the same time. We introduce ABACuS, a new low-cost hardware-counterbased RowHammer mitigation technique that performance-, energy-, and area-efficiently scales with worsening RowHammer vulnerability. We evaluate CORA in SCML training tasks, and our results show that CORA can reduce communication cost by up to 11x and achieve 1.2x - 4.2x end-to-end speedup over TCP in SCML training. CORA can be readily deployed with existing SCML frameworks such as Piranha with its socket-like interface. CORA exploits the chance that the SCML task is determined before execution and the pattern is largely input-irrelevant, so that CORA can plan message destinations on remote hosts at compile time. By using a protocol-aware design, CORA identifies the protocol used by the SCML program and sends messages directly to the remote party's protocol buffer, improving the efficiency of message exchange. We present CORA1 to implement SCML over RDMA. ![]() ![]() While modern datacenters offer high-bandwidth and low-latency networks with Remote Direct Memory Access (RDMA) capability, existing SCML implementation remains to use TCP sockets, leading to inefficiency. Secure Collaborative Machine Learning (SCML) suffers from high communication cost caused by secure computation protocols.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |