Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

Sorted by: relevance · author · university · dateNew search

You searched for +publisher:"University of Texas – Austin" +contributor:("Dimakis, Alex"). Showing records 1 – 3 of 3 total matches.

Search Limiters

Last 2 Years | English Only

No search limiters apply to these results.

▼ Search Limiters


University of Texas – Austin

1. Gong, Seong-Lyong. Memory protection techniques for DRAM scaling-induced errors.

Degree: Electrical and Computer Engineering, 2018, University of Texas – Austin

Continued scaling of DRAM technologies induces more faulty DRAM cells than before. These inherent faults increase significantly at sub-20nm technology, and hence traditional remapping schemes such as row/column sparing become very inefficient. Because the inherent faults manifest as single-bit errors, DRAM vendors are proposing to embed single-bit error correctable (SEC) ECC modules inside each DRAM chip, called In-DRAM ECC (IECC). However, IECC can achieve limited reliability improvement due to its weak correction capability. Specifically, at high scaling error rates, multi-bit scaling errors will easily occur in practice and escape from IECC protection. Because of the escaped scaling errors, the overall reliability may be degraded despite the increased overall overheads. For highly reliable systems that apply a strong ECC at the rank level (i.e., across DRAM chips that are accessed simultaneously), for example, Chipkill cannot be guaranteed anymore if the escaped errors occur. In this dissertation, I address this scaling-induced error problem as follows. First, I propose a more sophisticated fault-error model that includes intermittent scaling errors. In general, the effectiveness of proposed solutions strongly relies on the evaluation methodology. Prior related work evaluated their solutions against scaling errors only with a simple model and concluded efficient remapping schemes effectively cope with scaling errors. However, intermittent scaling errors cannot be easily detected and remapped. This implies that rather than the proposed remapping schemes, forward error correction may be the only solution to the scaling error problem. Using the new evaluation model, the proposed solutions to scaling errors can be evaluated in a more comprehensive way than before. Secondly, I propose two alternatives to In-DRAM ECC, Dual Use of On-chip redundancy (DUO) and Why-Pay-More (YPM), for highly reliable systems. DUO achieves higher reliability than In-DRAM ECC-based solutions by transferring on-chip redundancy to the rank level. Then, using the transferred redundancy together with original rank-level redundancy, a stronger rank-level ECC is applied. YPM is the first rank-level-only ECC protection against scaling errors. For this cost-saving design, YPM optimizes the correction capability by exploiting erasure Reed-Solomon (RS) decoding and iterative bit-flipping search. Each alternative is industry-changing in that DUO achieves much higher reliability than current rank-level ECC and YPM does not require In-DRAM ECC at all. Both alternatives are practical in that they require only small changes to DRAM designs. Advisors/Committee Members: Erez, Mattan (advisor), Swartzlander, Earl (committee member), Touba, Nur (committee member), Dimakis, Alex (committee member), Lin, Calvin (committee member), Sullivan, Mike (committee member).

Subjects/Keywords: DRAM; Memory; ECC; Scaling errors

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Gong, S. (2018). Memory protection techniques for DRAM scaling-induced errors. (Thesis). University of Texas – Austin. Retrieved from http://hdl.handle.net/2152/68922

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Gong, Seong-Lyong. “Memory protection techniques for DRAM scaling-induced errors.” 2018. Thesis, University of Texas – Austin. Accessed May 22, 2019. http://hdl.handle.net/2152/68922.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Gong, Seong-Lyong. “Memory protection techniques for DRAM scaling-induced errors.” 2018. Web. 22 May 2019.

Vancouver:

Gong S. Memory protection techniques for DRAM scaling-induced errors. [Internet] [Thesis]. University of Texas – Austin; 2018. [cited 2019 May 22]. Available from: http://hdl.handle.net/2152/68922.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Gong S. Memory protection techniques for DRAM scaling-induced errors. [Thesis]. University of Texas – Austin; 2018. Available from: http://hdl.handle.net/2152/68922

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


University of Texas – Austin

2. -4065-8654. Resource-constrained, scalable learning.

Degree: Electrical and Computer Engineering, 2015, University of Texas – Austin

Our unprecedented capacity for data generation and acquisition often reaches the limits of our data storage capabilities. Situations when data are generated faster or at a greater volume than can be stored demand a streaming approach. Memory is an even more valuable resource. Algorithms that use more memory than necessary can pose bottlenecks when processing high-dimensional data and the need for memory-efficient algorithms is especially stressed in the streaming setting. Finally, network along with storage, emerge as the critical bottlenecks in the context of distributed computation. These computational constraints spell out a demand for efficient tools that guarantee a solution in the face of limited resources, even when the data is very noisy or highly incomplete. For the first part of this dissertation, we present our work on streaming, memory-limited Principal Component Analysis (PCA). Therein, we give the first convergence guarantees for an algorithm that solves PCA in the single-pass streaming setting. Then, we discuss the distinct challenges that arise when the received samples are overwhelmingly incomplete and present an algorithm and analysis that deals with this issue. Finally, we give a set of extensive experiment results that showcase the practical merits of our algorithm over the state of the art. The need for heavy network communication arises as the bottleneck when dealing with cluster computation. In that paradigm, a set of worker nodes are connected over the network to produce a cluster with improved computational and storage capacities. This comes with an increased need for communication across the network. In the last part of this work, we consider the problem of PageRank on graph engines. Therein, we make changes to GraphLab, a state-of-the-art platform for distributed graph computation, in a way that leads to a 7x-10x speedup for certain PageRank approximation tasks. Accompanying analysis supports the behaviour we see in our experiments. Advisors/Committee Members: Vishwanath, Sriram (advisor), Caramanis, Constantine (advisor), Dimakis, Alex (committee member), Sanghavi, Sujay (committee member), Ravikumar, Pradeep (committee member).

Subjects/Keywords: Resource contraints; Limited memory; Storage; Network; Principle component analysis; PageRank; Graph engines

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

-4065-8654. (2015). Resource-constrained, scalable learning. (Thesis). University of Texas – Austin. Retrieved from http://hdl.handle.net/2152/32226

Note: this citation may be lacking information needed for this citation format:
Author name may be incomplete
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

-4065-8654. “Resource-constrained, scalable learning.” 2015. Thesis, University of Texas – Austin. Accessed May 22, 2019. http://hdl.handle.net/2152/32226.

Note: this citation may be lacking information needed for this citation format:
Author name may be incomplete
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

-4065-8654. “Resource-constrained, scalable learning.” 2015. Web. 22 May 2019.

Note: this citation may be lacking information needed for this citation format:
Author name may be incomplete

Vancouver:

-4065-8654. Resource-constrained, scalable learning. [Internet] [Thesis]. University of Texas – Austin; 2015. [cited 2019 May 22]. Available from: http://hdl.handle.net/2152/32226.

Note: this citation may be lacking information needed for this citation format:
Author name may be incomplete
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

-4065-8654. Resource-constrained, scalable learning. [Thesis]. University of Texas – Austin; 2015. Available from: http://hdl.handle.net/2152/32226

Note: this citation may be lacking information needed for this citation format:
Author name may be incomplete
Not specified: Masters Thesis or Doctoral Dissertation


University of Texas – Austin

3. Shah, Virag. Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics.

Degree: Electrical and Computer Engineering, 2015, University of Texas – Austin

We consider a centralized content delivery infrastructure where a large number of storage-intensive files are replicated across several collocated servers. To achieve scalable delays in file downloads under stochastic loads, we allow multiple servers to work together as a pooled resource to meet individual download requests. In such systems basic questions include: How and where to replicate files? How significant are the gains of resource pooling over policies which use single server per request? What are the tradeoffs among conflicting metrics such as delays, reliability and recovery costs, and power? How robust is performance to heterogeneity and choice of fairness criterion? In this thesis we provide a simple performance model for large systems towards addressing these basic questions. For large systems where the overall system load is proportional to the number of servers, we establish scaling laws among delays, system load, number of file replicas, demand heterogeneity, power, and network capacity. Advisors/Committee Members: de Veciana, Gustavo (advisor), Baccelli, Francois (committee member), Dimakis, Alex (committee member), Hasenbein, John (committee member), Shakkottai , Sanjay (committee member).

Subjects/Keywords: Content delivery infrastructure; Performance models; Queueing theory; Delays; Robustness; Scaling laws

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Shah, V. (2015). Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics. (Thesis). University of Texas – Austin. Retrieved from http://hdl.handle.net/2152/31419

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Shah, Virag. “Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics.” 2015. Thesis, University of Texas – Austin. Accessed May 22, 2019. http://hdl.handle.net/2152/31419.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Shah, Virag. “Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics.” 2015. Web. 22 May 2019.

Vancouver:

Shah V. Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics. [Internet] [Thesis]. University of Texas – Austin; 2015. [cited 2019 May 22]. Available from: http://hdl.handle.net/2152/31419.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Shah V. Centralized content delivery infrastructure exploiting resource pools : performance models and asymptotics. [Thesis]. University of Texas – Austin; 2015. Available from: http://hdl.handle.net/2152/31419

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

.