Minesh Patel awarded the DSN-2022 William C. Carter Award

We’re very honoured to announce that Minesh Patel has been awarded the DSN-2022 William C. Carter Award for his PhD Dissertation titled “Enabling Effective Error Mitigation in Memory Chips That Use On-Die Error-Correcting Codes“.  Congratulations Minesh!

About the award:
The William C. Carter Award recognizes an individual who has made a significant contribution to the field of dependable and secure computing throughout his or her PhD dissertation, and commemorates the late William (Bill) C. Carter, a key figure in the formation and development of the field of dependable computing. The award is jointly sponsored by the IEEE TC on Dependable Computing and Fault Tolerance (TCFT) and IFIP Working Group 10.4 on Dependable Computing and Fault Tolerance (WG 10.4), and presented annually at the DSN Conference.  Minesh received the award at DSN 2022 in Baltimore this week.
The selection committee awarded Minesh’s PhD thesis with the following recognition:
The committee considered that the thesis constitutes a ground-breaking work in the understanding and handling of state-of-the-art error correction techniques opaquely implemented inside the newest DRAM chips. These chips are widespread and, as they keep scaling down, they become more susceptible to disturbance errors. As a reaction, the DRAM industry is integrating Error Correction Codes (ECC) inside the DRAM devices, on the die, in a confidential and proprietary way, masking these errors from the systems using them and, thus, preventing a system-level error-processing optimization. For instance, large datacenters cannot use anymore the previously seen DRAM error rates to train predictive models of hardware failures, something they need to do. Dr. Patel’s work addresses this challenging problem through rigorous experimental analyses of more than 300 DRAM chips. This enables him not only to produce new statistical error models and tools, but also to build new system-level techniques to improve the chips error tolerance. This work has been published in top venues in computer architecture and dependable systems; two of these publications won a best paper award, including one at DSN 2019. It is also important to note that Dr. Minesh’s work has already accumulated close to 1000 citations and that his h-index is already at 16, highlighting his influence within these communities. Last but not least, Dr. Patel’s work has already had some industrial impact: several industrial research labs and major DRAM manufacturers have provided strongly positive feedbacks under various forms.

The thesis is well written and easy to read, even for non-experts in the field. The committee was impressed by the breadth and the depth of the numerous contributions made, as well as by the thoroughness and soundness of the experimental validations carried out. For these reasons, the committee decided to award the 2022 Carter Award to Dr. Minesh Patel.


Thesis background and moving forward:

Minesh started his PhD thesis at CMU under the supervision of Onur Mutlu, and continued his PhD research at ETH Zurich with Onur, following Onur’s move from CMU to ETH.  During his PhD work, Minesh carried out two internships at Microsoft, in Fort Collins, CO, and Redmond, WA, and three internships at Apple in Cupertino, CA.

Minesh defended his PhD thesis last October and is currently exploring options for what comes next.  His graduate work focuses primarily on improvements to memory systems and, going forward, he is interested in broader topics pertaining to cutting-edge computing systems and broadly interested in architecture/systems topics.  Read more on his personal webpage: https://www.mineshp.com.

Minesh Patel, April 2022 (defended 01 October 2021)
Thesis title: “Enabling Effective Error Mitigation in Modern Memory Chips that Use On-Die Error-Correcting Codes”
[Slides (pptx) (pdf)]
[Thesis arXiv (pdf)]
[SAFARI Live Seminar Video]

Acceptance Speech, William C. Carter PhD Dissertation Award in Dependability at DSN 2022 (Baltimore, MD), 28 June 2022.
[Acceptance Speech Video]
[Slides (pptx)(pdf]

Thesis works: 

Minesh Patel, Taha Shahroodi, Aditya Manglik, A. Giray Yaglikci, Ataberk Olgun, Haocong Luo, Onur Mutlu,
“A Case for Transparent Reliability in DRAM Systems”,
[arXiv version]

Minesh Patel, Geraldo F. de Oliveira Jr., and Onur Mutlu,
“HARP: Practically and Effectively Identifying Uncorrectable Errors in Memory Chips That Use On-Die Error-Correcting Codes”
Proceedings of the 54th International Symposium on Microarchitecture (MICRO), Virtual, October 2021.
[Slides (pptx) (pdf)]
[Short Talk Slides (pptx) (pdf)]
[Lightning Talk Slides (pptx) (pdf)]
[Talk Video (20 minutes)]
[Lightning Talk Video (1.5 minutes)]
[HARP Source Code (Officially Artifact Evaluated with All Badges)]
[arXiv version]

Minesh Patel, Jeremie S. Kim, Taha Shahroodi, Hasan Hassan, and Onur Mutlu,
“Bit-Exact ECC Recovery (BEER): Determining DRAM On-Die ECC Functions by Exploiting DRAM Data Retention Characteristics”
Proceedings of the 53rd International Symposium on Microarchitecture (MICRO), Virtual, October 2020.
[Slides (pptx) (pdf)]
[Short Talk Slides (pptx) (pdf)]
[Lightning Talk Slides (pptx) (pdf)]
[Lecture Slides (pptx) (pdf)]
[Talk Video (15 minutes)]
[Short Talk Video (5.5 minutes)]
[Lightning Talk Video (1.5 minutes)]
[Lecture Video (52.5 minutes)]
[BEER Source Code]
Best paper award [read Minesh’s interview below on his Best Paper Award.  This interview originally appeared in our January 2021 Newsletter]

Minesh Patel, Jeremie S. Kim, Hasan Hassan, and Onur Mutlu,
“Understanding and Modeling On-Die Error Correction in Modern DRAM: An Experimental Study Using Real Devices”
Proceedings of the 49th Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN), Portland, OR, USA, June 2019.
[Slides (pptx) (pdf)]
[Talk Video (26 minutes)]
[Full Talk Lecture (29 minutes)]
Source Code for EINSim, the Error Inference Simulator]
Best paper award

Minesh Patel, Jeremie S. Kim, and Onur Mutlu,
“The Reach Profiler (REAPER): Enabling the Mitigation of DRAM Retention Failures via Profiling at Aggressive Conditions”
Proceedings of the 44th International Symposium on Computer Architecture (ISCA), Toronto, Canada, June 2017.
[Slides (pptx) (pdf)]
[Lightning Session Slides (pptx) (pdf)]

We interviewed Minesh on his Best Paper Award at MICRO’20 for his work on BEER.  This interview originally appeared in our January 2021 Newsletter.

You recently won the Best Paper Award at MICRO, congratulations! Can you tell us about the significance of this paper?

Minesh: This paper addresses the larger problem that hidden proprietary features implemented by DRAM manufacturers impede end-users from bringing out the best of DRAM technology. We believe BEER takes an important step towards bridging the gap between industry and end-users, starting by focusing on a key example of such features: on-die ECC. Our work discusses how and why on-die ECC limits third-party DRAM consumers and then introduces techniques that the consumers can use to overcome these limitations. We have released our tools in an open-source project and look forward to having the community use and extend them.

What were the biggest challenges for you during the writing and review process?

Minesh: I would say that the biggest challenge we faced when writing this paper was to clearly articulate the problem of on-die ECC limiting third-party users. This includes both (i) describing how and why this limitation arises and (ii) providing concrete examples that the reader can relate to. We spent considerable effort in crafting these arguments such that both we and the reader have a clear understanding of the problem we tackle, our goal in this work, and the final value of our contributions.

Posted in Awards, Papers, PhD Defense.