HashGraph—Scalable Hash Tables Using a Sparse Graph Data Structure

  title={HashGraph—Scalable Hash Tables Using a Sparse Graph Data Structure},
  author={Oded Green},
  journal={ACM Transactions on Parallel Computing (TOPC)},
  pages={1 - 17}
  • Oded Green
  • Published 5 July 2019
  • Computer Science
  • ACM Transactions on Parallel Computing (TOPC)
In this article, we introduce HashGraph, a new scalable approach for building hash tables that uses concepts taken from sparse graph representations—hence, the name HashGraph. HashGraph introduces a new way to deal with hash-collisions that does not use “open-addressing” or “separate-chaining,” yet it has the benefits of both these approaches. HashGraph currently works for static inputs. Recent progress with dynamic graph data structures suggests that HashGraph might be extendable to dynamic… 

Figures from this paper

Scalable Hash Table for NUMA Systems

This work presents a multi-GPU hash table implementation that can process keys at a throughput comparable to that of distributed hash tables, and shares many of the same features of the single GPU algorithm.

WarpCore: A Library for fast Hash Tables on GPUs

The fast memory interface of modern GPUs together with a parallel hashing scheme tailored to improve global memory access patterns are exploited, to design WarpCore – a versatile library of hash table data structures that can be used for accelerating a real world bioinformatics application with speedups of over two orders ofmagnitude against state-of-the-art CPU-based solutions.

Generating Permutations Using Hash Tables

A new method for creating random permutations that is scalable, efficient, and simple is described and it is shown that the operation of generating a random permutation shares traits with building a hashtable.

Billion degree of freedom granular dynamics simulation on commodity hardware via heterogeneous data-type representation

We discuss modeling, algorithmic, and software aspects that allow a simulation tool called Chrono::Granular to run billion-degree-of-freedom dynamics problems on commodity hardware, i.e., a

Blockchain Consensus Mechanism Based on Improved Distributed Consistency and Hash Entropy

  • Jue Ma
  • Computer Science
    Sci. Program.
  • 2021
A novel and effective consensus algorithm is designed that can reduce the waste of computing resources, increase the block generation speed, and ensure the fairness of nodes participating in the competition, which is an effective solution to ensure the stable operation of the blockchain system.

Blockchain Consensus Mechanism Based on Improved Distributed Consistency and Hash Entropy

A novel and effective consensus algorithm is designed that constructs a more random additive constant through the generation matrix of the error correction code and uses the value of the hash entropy to prove that the constructed hash function can meet the requirements of high throughput and fast consensus in performance.

REFLOW Portable Crypto Functions Project deliverable 2

This work implements and evaluates a Reflow smart contract for Zenroom and presents an application related to multiple anonymous signatures by authenticated parties and their non-interactive verification.

Towards Lightweight Authorisation of IoT-Oriented Smart-Farms using a Self-Healing Consensus Mechanism

This paper proposes a novel consensus mechanism towards access control in networks populated by constrained devices, and proposes a hybrid design to produce a robust solution that suffers from neither the vulnerabilities of centralisation or data aggregation of distribution.

Reflow multi-party signatures-Secure BLS multi-signatures Zero-knowledge proof

  • 2021



Real-time parallel hashing on the GPU

An efficient data-parallel algorithm for building large hash tables of millions of elements in real-time, which considers a classical sparse perfect hashing approach, and cuckoo hashing, which packs elements densely by allowing an element to be stored in one of multiple possible locations.

Stadium Hashing: Scalable and Flexible Hashing on GPUs

Stash is presented with collaborative lanes (clStash) that enhances GPU's SIMD resource utilization for batched insertions during hash table creation and can be up to 2 and 3 times faster than GPU Cuckoo hashing for in-core and out-of-core tables respectively.

Bin-Hash Indexing: A Parallel Method for Fast Query Processing

The Bin-Hash index offers extremely high levels of concurrency, and is therefore well-suited for the emerging commodity of parallel processors, such as multi-cores, cell processors, and general purpose graphics processing units (GPU).

A Dynamic Hash Table for the GPU

A warp-cooperative work sharing strategy that reduces branch divergence and provides an efficient alternative to the traditional way of per-thread (or per-warp) work assignment and processing is proposed, which builds a dynamic non-blocking concurrent linked list, the slab list, that supports asynchronous, concurrent updates as well as search queries.

Concurrent hash tables: fast and general?(!)

This work explains how to lift limitations in a provably scalable way and demonstrates that dynamic growing has a performance overhead comparable to the same generalization in sequential hash tables.

Sort vs. Hash Revisited: Fast Join Implementation on Modern Multi-Core CPUs

This paper re-examines two popular join algorithms to determine if the latest computer architecture trends shift the tide that has favored hash join for many years and offers multicore implementations of hash join and sort-merge join which consistently outperform all previously reported results.

Design and evaluation of main memory hash join algorithms for multi-core CPUs

A very simple hash join algorithm is very competitive to the other more complex methods, and improves dramatically as the skew in the input data increases, and it quickly starts to outperform all other algorithms.

Optimizing High Performance Distributed Memory Parallel Hash Tables for DNA k-mer Counting

  • Tony PanSanchit MisraS. Aluru
  • Computer Science
    SC18: International Conference for High Performance Computing, Networking, Storage and Analysis
  • 2018
This work presents two optimized distributed parallel hash table techniques that utilize cache friendly algorithms for local hashing, overlapped communication and computation to hide communication costs, and vectorized hash functions that are specialized for fc-mer and other short key indices.

Distributed Join Algorithms on Thousands of Cores

This paper explains how to use MPI to implement joins, shows the impact and advantages of RDMA, discusses the importance of network scheduling, and study the relative performance of sorting vs. hashing.

Multi-Core, Main-Memory Joins: Sort vs. Hash Revisited

The experiments show that, contrary to claims, radix-hash join is still clearly superior, and sort-merge approaches to performance of radix only when very large amounts of data are involved.