Lecture

New Technologies in Concurrent Algorithms: RDMA & Persistent Memory

Graph Chat

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

Description

This lecture by the instructor introduces two new technologies, RDMA and Persistent Memory, discussing their features, benefits, and challenges. The first part covers RDMA, focusing on Remote Direct Memory Access, permissions, failures, and consensus modeling. The second part delves into Persistent Memory, explaining its characteristics, stability, and challenges in data structures. Various concepts such as preventing equivocations, non-equivocating broadcast, and log-free data structures are explored, along with the importance of durable linearizability. The lecture concludes with a discussion on the lower bound for persistent memory and efficient data structures. Throughout the lecture, the instructor provides insights into the impact of RDMA on agreement and microsecond consensus for applications.

Official source

Related concepts (32)

Persistent data structure

In computing, a persistent data structure or not ephemeral data structure is a data structure that always preserves the previous version of itself when it is modified. Such data structures are effectively immutable, as their operations do not (visibly) update the structure in-place, but instead always yield a new updated structure. The term was introduced in Driscoll, Sarnak, Sleator, and Tarjans' 1986 article. A data structure is partially persistent if all versions can be accessed but only the newest version can be modified.

Data structure

In computer science, a data structure is a data organization, management, and storage format that is usually chosen for efficient access to data. More precisely, a data structure is a collection of data values, the relationships among them, and the functions or operations that can be applied to the data, i.e., it is an algebraic structure about data. Data structures serve as the basis for abstract data types (ADT). The ADT defines the logical form of the data type. The data structure implements the physical form of the data type.

Array (data structure)

In computer science, an array is a data structure consisting of a collection of elements (values or variables), of same memory size, each identified by at least one array index or key. An array is stored such that the position of each element can be computed from its index tuple by a mathematical formula. The simplest type of data structure is a linear array, also called one-dimensional array. For example, an array of ten 32-bit (4-byte) integer variables, with indices 0 through 9, may be stored as ten words at memory addresses 2000, 2004, 2008, .

Purely functional data structure

In computer science, a purely functional data structure is a data structure that can be directly implemented in a purely functional language. The main difference between an arbitrary data structure and a purely functional one is that the latter is (strongly) immutable. This restriction ensures the data structure possesses the advantages of immutable objects: (full) persistency, quick copy of objects, and thread safety. Efficient purely functional data structures may require the use of lazy evaluation and memoization.

Search data structure

In computer science, a search data structure is any data structure that allows the efficient retrieval of specific items from a set of items, such as a specific record from a database. The simplest, most general, and least efficient search structure is merely an unordered sequential list of all the items. Locating the desired item in such a list, by the linear search method, inevitably requires a number of operations proportional to the number n of items, in the worst case as well as in the average case.

Related lectures (64)

Memory Virtualization: Address Space Abstraction and Management

Covers memory virtualization, focusing on address space abstraction and the management of memory resources by the operating system.

Dynamic Memory Allocation: Understanding Pointers in C

Provides an overview of dynamic memory allocation in C programming, focusing on pointers and memory management techniques.

Cache Memory

Explores cache memory design, hits, misses, and eviction policies in computer systems, emphasizing spatial and temporal locality.

Python Lists: Manipulation and Comprehension

Covers Python list manipulation and comprehension, emphasizing memory representation and mutability.

Memory Management: Address Space Abstraction