(The talk will be given in English)
Speaker: Dr. Amit Golander
|
011 hall, Electrical Engineering-Kitot Building
|
Monday, November 25th, 2024
12:00 - 13:00
Persistent-Memory based Storage Systems
Abstract
Persistent Memory is an emerging technology that will offer substantial advantages compared to current flash and block-based storage devices. With PM, applications should run faster, consume fewer compute resources and potentially be made simpler as well. Yet, new software is required to allow such PM benefits. For most of the last decade, I was the chief architect for PM-based software at Plexistor, NetApp and Huawei Cloud, all pioneers in leveraging PM and coming from different product/business constraints.
In this talk, I will explore the many opportunities such a radically new storage-class memory PM device opens, discuss the architectural tradeoffs and exemplify with two PM-based systems that I designed:
1) the first local feature-rich PM-based file system with optional remote mirroring (Plexistor & NetApp, 2014-2020) and
2) the first remote random-access PM pool with optional near-memory processing (Huawei Cloud 2020-2024)
Short Bio
Dr. Amit Golander was until last month the Cloud Storage CTO and Cloud lab director for Huawei R&D center in Israel (2020-2024). He received his B.Sc. in EE & CS in 2001 and his Ph.D. in Computer Architecture in 2008, both from Tel- Aviv University.
Amit co-founded Plexistor (acquired by NetApp, 2014-2017-2020), where he was responsible for developing the world’s first feature-rich PM-based file system and surrounding MAX Data product. Before that, Amit was a technology leader in: Tonian (acquired by PrimaryData in 2013), IBM Next-Gen. Systems Research (2008-2011) and IBM System & Technology Group (1999-2008).
Amit composed 59 patents and 26 peer-reviewed papers & posters. He taught Storage and Computer Architecture courses and co-advised 3 M.Sc. students at TAU EE. His research interests include hardware and software for data center infrastructure and in particular the potential of emerging rack-scale computing to revolutionize heterogeneous compute, memory disaggregation, storage systems and photonic fabrics.
השתתפות בסמינר תיתן קרדיט שמיעה = עפ"י רישום שם מלא + מספר ת.ז. בטופס הנוכחות שיועבר באולם במהלך הסמינר
