skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Multi-Level Memory: What You Add Is More Important Than What You Take Out.


Abstract not provided.

; ;
Publication Date:
Research Org.:
Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)
OSTI Identifier:
Report Number(s):
DOE Contract Number:
Resource Type:
Resource Relation:
Conference: Proposed for presentation at the MemSys held October 3-6, 2016 in Washington DC, DC.
Country of Publication:
United States

Citation Formats

Rodrigues, Arun F., Voskuilen, Gwendolyn Renae, and Hammond, Simon David. Multi-Level Memory: What You Add Is More Important Than What You Take Out.. United States: N. p., 2016. Web.
Rodrigues, Arun F., Voskuilen, Gwendolyn Renae, & Hammond, Simon David. Multi-Level Memory: What You Add Is More Important Than What You Take Out.. United States.
Rodrigues, Arun F., Voskuilen, Gwendolyn Renae, and Hammond, Simon David. 2016. "Multi-Level Memory: What You Add Is More Important Than What You Take Out.". United States. doi:.
title = {Multi-Level Memory: What You Add Is More Important Than What You Take Out.},
author = {Rodrigues, Arun F. and Voskuilen, Gwendolyn Renae and Hammond, Simon David},
abstractNote = {Abstract not provided.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = 2016,
month = 9

Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • In 1962, the U.S. Atomic Energy Commission (AEC) began removing radioactive waste from defense facilities across the nation with the first disposal of low-level waste. Thirty-seven years later, transuranic (TRU) waste began making its way toward permanent isolation in the excavated salt drifts at the Waste Isolation Pilot Plant (WIPP) repository. Ever since, close tallies have been assigned to track the volume removed, shipped, and disposed at WIPP. In 2007, DOE will move past the 45-year mark of progress toward meeting a much improved environmental stewardship mission. At the end of January 2007, WIPP's contribution to this mission totals 45,214more » m{sup 3} and 5,413 shipments of TRU waste. This equates to approximately 8.5 m{sup 3} per shipment and an average of seven hundred shipments per year since opening. Considering that the actual annual rates have consistently climbed, this appears to be very good progress since WIPP opened, and especially for the past five years. The numbers from 2006 were record setting, at over 10,000 m{sup 3} and more than 1100 shipments. While these numbers share information on the volume received at WIPP, they do not fully portray the actual waste volume of any shipment. This paper provides an expanded view of the differences in how volume values are tracked and reported. (authors)« less
  • Surface scientists argue about the fundamental nature of Schottky barriers, or more precisely what determines the location of the Fermi level at semiconductor surfaces and interfaces. Electrical and materials engineers worry about how to make Schottky barrier diodes and gates to field effect transistors and the control of barrier heights. There is some interesting middle ground in which the location of the surface and interface Fermi level can, for example, determine semiconductor doping characteristics during crystal growth. The authors will discuss several interesting and well known examples of doping characteristics which are still somewhat mysterious. Specifically, they address the followingmore » question: (1) why is Ge doped GaAs p type when grown from Ga melts but n type when grown from Au melts (2) why is low resistivity p type ZnSe, AlAs, and AlGaInP hard to make, and more importantly, how can the problem be fixed. In addition they describe how this concept relates to the electronic properties to a new type of GaAs, GaAs:As, which contains a high density of As precipitates.« less
  • In this paper, we present two schemes for concurrency control and recovery in distributed main-memory databases. In the client-server scheme, clients ship log records to the server, which applies the updates to its database copy. In the shared disk scheme, each site broadcasts its updates to other sites. The above enable our schemes to support concurrent updates to the same page at different sites. Both schemes support an explicit multi-level recovery abstraction for high concurrency, reduced disk I/O by writing only redo log records to disk during normal processing, and use of per-transaction redo and undo logs to reduce contention.more » Further, we use a fuzzy checkpointing scheme that writes only dirty pages to disk, yet minimally interferes with normal processing, not requiring updaters to even acquire a latch before updating a page.« less
  • Due to their low access latency, high read speed, and power-efficient operation, flash memory storage devices are rapidly emerging as an attractive alternative to traditional magnetic storage devices. However, tests show that the most efficient indexing methods are not able to take advantage of the flash memory storage devices. In this paper, we present a set of multi-level bitmap indexes that can effectively take advantage of flash storage devices. These indexing methods use coarsely binned indexes to answer queries approximately, and then use finely binned indexes to refine the answers. Our new methods read significantly lower volumes of data atmore » the expense of an increased disk access count, thus taking full advantage of the improved read speed and low access latency of flash devices. To demonstrate the advantage of these new indexes, we measure their performance on a number of storage systems using a standard data warehousing benchmark called the Set Query Benchmark. We observe that multi-level strategies on flash drives are up to 3 times faster than traditional indexing strategies on magnetic disk drives.« less
  • Modern architectures are characterized by deeper levels of memory hierarchy, often explicitly addressable. Optimizing applications for such architectures requires careful management of the data movement across all these levels. In this paper, we focus on the problem of mapping tensor contractions to memory hierarchies with more than two levels, specifically addressing placement of memory allocation and data movement statements, choice of loop fusions, and tile size selection. Existing algorithms to find an integrated solution to this problem even for two-level memory hierarchies has been shown to be expensive. We improve upon this work by focusing on the first-order cost components,more » simplifying the analysis required and reducing the number of candidates to be evaluated. We have evaluated our framework on a cluster of GPUs. Using five tensor contraction expressions, we show that fusion at multiple levels improves performance, and our framework is effective in determining protable transformations.« less