No abstract

Prior work in hardware prefetching has focused mostly on either predicting regular streams with uniform strides, or predicting irregular access patterns at the cost of large hardware structures. This paper introduces the Variable Length Delta Prefetcher (VLDP), which builds up delta histories between successive cache line misses within physical pages, and then uses these histories to predict the order of cache line misses in new pages. One of VLDP's distinguishing features is its use of multiple prediction tables, each of which stores predictions based on a different length of input history. For example, the first prediction table takes as input only the single most recent delta between cache misses within a page, and attempts to predict the next cache miss in that page. The second prediction table takes as input a sequence of the two most recent deltas between cache misses within a page, and also attempts to predict the next cache miss in that page, and so on with additional tables. Longer histories generally yield more accurate predictions, so VLDP prefers to make predictions based on the longest history table that has a matching entry.Using a global history of patterns it has seen in the past, VLDP is able to issue prefetches without having to wait for additional per-page confirmation, and it is even able to prefetch patterns that show no repetition within a physical page. VLDP does not use the program counter (PC) to make its predictions, but our evaluation shows that it out-performs the highest-performing PCbased prefetcher by 7.1%, and the highest performing prefetcher that doesn't employ the PC by 5.8%.

The DRAM main memory system in modern servers is largely homogeneous. In recent years, DRAM manufacturers have produced chips with vastly differing latency and energy characteristics. This provides the opportunity to build a heterogeneous main memory system where different parts of the address space can yield different latencies and energy per access. The limited prior work in this area has explored smart placement of pages with high activities. In this paper, we propose a novel alternative to exploit DRAM heterogeneity. We observe that the critical word in a cache line can be easily recognized beforehand and placed in a low-latency region of the main memory. Other non-critical words of the cache line can be placed in a low-energy region. We design an architecture that has low complexity and that can accelerate the transfer of the critical word by tens of cycles. For our benchmark suite, we show an average performance improvement of 12.9% and an accompanying memory energy reduction of 15%.

Many of the pins on a modern chip are used for power delivery. If fewer pins were used to supply the same current, the wires and pins used for power delivery would have to carry larger currents over longer distances. This results in an "IR-drop" problem, where some of the voltage is dropped across the long resistive wires making up the power delivery network, and the eventual circuits experience fluctuations in their supplied voltage. The same problem also manifests if the pin count is the same, but the current draw is higher. IR-drop can be especially problematic in 3D DRAM devices because (i) low cost (few pins and TSVs) is a high priority, (ii) 3D-stacking increases current draw within the package without providing proportionate room for more pins, and (iii) TSVs add to the resistance of the power delivery network.This paper is the first to characterize the relationship between the power delivery network and the maximum supported activity in a 3D-stacked DRAM memory device. The design of the power delivery network determines if some banks can handle less activity than others. It also determines the combinations of bank activities that are permissible. Both of these attributes can feed into architectural policies. For example, if some banks can handle more activities than others, the architecture benefits by placing data from high-priority threads or data from frequently accessed pages into those banks. The memory controller can also derive higher performance if it schedules requests to specific combinations of banks that do not violate the IR-drop constraint.We first define an IR-drop-aware scheduler that encodes a number of activity constraints. This scheduler, however, falls short of the performance of an unrealistic ideal PDN

No abstract