In-RDBMS Hardware Acceleration of Advanced Analytics

@article{Mahajan2018InRDBMSHA,
  title={In-RDBMS Hardware Acceleration of Advanced Analytics},
  author={Divya Mahajan and Joon Kyung Kim and Jacob Sacks and A. Ardalan and Arun Kumar and Hadi Esmaeilzadeh},
  journal={Proc. VLDB Endow.},
  year={2018},
  volume={11},
  pages={1317-1331}
}
The data revolution is fueled by advances in machine learning, databases, and hardware design. Programmable accelerators are making their way into each of these areas independently. As such, there is a void of solutions that enables hardware acceleration at the intersection of these disjoint fields. This paper sets out to be the initial step towards a unifying solution for in- D atabase A cceleration of Advanced A nalytics (DAnA). Deploying specialized hardware, such as FPGAs, for in… Expand
A Glass Half Full: Using Programmable Hardware Accelerators in Database Analytics
Even though there have been a large number of proposals to accelerate databases using specialized hardware, often the opinion of the community is pessimistic: the performance and energy efficiencyExpand
The Glass Half Full: Using Programmable Hardware Accelerators in Analytics
  • Z. István
  • Computer Science
  • IEEE Data Eng. Bull.
  • 2019
TLDR
This paper highlights several challenges that have traditionally hindered the deployment of hardware acceleration in databases and explains how they have been alleviated or removed altogether by recent research results and the changing hardware landscape. Expand
Specialized Hardware Solutions for In-Database Analytics and Machine Learning
TLDR
This thesis takes an in-memory database management system (DBMS) and focuses on improving online analytical processing (OLAP) and machine learning (ML) workloads within the DBMS by exploring specialization opportunities offered by an FPGA co-processor. Expand
doppioDB 1.0: Machine Learning inside a Relational Engine
TLDR
This first version of doppioDB provides a platform for extending traditional relational processing with customizable hardware to support stochastic gradient descent and decision tree ensembles and shows examples of how they could be included into SQL and embedded as part of conventional components of a relational database engine. Expand
Hardware Acceleration for DBMS Machine Learning Scoring: Is It Worth the Overheads?
TLDR
This paper analyzes the effectiveness of using GPUs and FPGAs to accelerate scoring for random forest, a popular machine learning model, on tabular input data obtained from Microsoft SQL Server and observes that the offloading decision as well as the choice of the optimal hardware backend should depend at least on the model complexity. Expand
High Bandwidth Memory on FPGAs: A Data Analytics Perspective
TLDR
This paper considers three workloads that are often performed in analytics oriented databases and implement them on FPGA showing in which cases they benefit from HBM: range selection, hash join, and stochastic gradient descent for linear model training. Expand
In-memory database acceleration on FPGAs: a survey
TLDR
This paper surveys using FPGAs to accelerate in-memory database systems targeting designs that can operate at the speed of main memory. Expand
doppioDB 2.0: Hardware Techniques for Improved Integration of Machine Learning into Databases
TLDR
The demo explores two complementary alternatives: how to train models directly on compressed/encrypted column-stores using a specialized coordinate descent engine, and how to use a bitwise weaving index for stochastic gradient descent on low precision input data. Expand
Accelerating Raw Data Analysis with the ACCORDA Software and Hardware Architecture
TLDR
ACORDA robustly matches or outperforms prior systems that depend on caching loaded data, while computing on raw, unloaded data, and unlocks aggressive encoding-oriented optimizations that deliver 80% average performance increase over the 7 affected TPC-H queries. Expand
Polystore++: Accelerated Polystore System for Heterogeneous Workloads
TLDR
Polystore++ is envisioned, an architecture to accelerate existing polystore systems using hardware accelerators (e.g., FPGAs, CGRAs, and GPUs) and can achieve high performance at low power by identifying and offloading components of a polystore system that are amenable to acceleration using specialized hardware. Expand
...
1
2
3
...

References

SHOWING 1-10 OF 62 REFERENCES
doppioDB: A Hardware Accelerated Database
TLDR
This work presents doppioDB, a main-memory column store, extended with Hardware User Defined Functions (HUDFs), and evaluates it on an emerging hybrid multicore architecture, the Intel Xeon+FPGA platform, where the CPU and FPGA have cache-coherent access to the same memory, such that the hardware operators can directly access the database tables. Expand
Centaur: A Framework for Hybrid CPU-FPGA Databases
TLDR
Centaur is presented, a framework running on theFPGA that allows the dynamic allocation of FPGA operator plans to query plans, pipelining these operators among themselves when needed, and the hybrid execution of operator pipelinesrunning on the CPU and the FPGa. Expand
Towards a unified architecture for in-RDBMS analytics
TLDR
This work proposes a unified architecture for in-database analytics that requires changes to only a few dozen lines of code to integrate a new statistical technique, and demonstrates the feasibility of this architecture by integrating several popular analytics techniques into two commercial and one open-source RDBMS. Expand
Scale-Out Acceleration for Machine Learning
TLDR
This work offers CoSMIC, a full computing stack constituting language, compiler, system software, template architecture, and circuit generators, that enable programmable acceleration of learning at scale that takes an effective and vital step towards enabling scale-out acceleration for machine learning. Expand
LINQits: big data on little clients
TLDR
LINQits is a flexible hardware template that can be mapped onto programmable logic or ASICs in a heterogeneous system-on-chip for a mobile device or server and accelerates a domain-specific query language called LINQ, which benefits extensively from hardware acceleration. Expand
TABLA: A unified template-based framework for accelerating statistical machine learning
TLDR
TABLA provides a template-based framework that generates accelerators for a class of machine learning algorithms and rigorously compares the benefits of FPGA acceleration to multi-core CPUs and many-core GPUs using real hardware measurements. Expand
HippogriffDB: Balancing I/O and GPU Bandwidth in Big Data Analytics
TLDR
HippogriffDB is proposed, an efficient, scalable GPU-accelerated OLAP system that adopts a query-over-block execution model that provides scalability using a stream-based approach and improves kernel efficiency with the operator fusion and double buffering mechanism. Expand
An Architecture for Compiling UDF-centric Workflows
TLDR
A novel architecture for automatically compiling workflows of UDFs is described and several optimizations that consider properties of the data,UDFs, and hardware together in order to generate different code on a case-by-case basis are proposed. Expand
In-datacenter performance analysis of a tensor processing unit
  • N. Jouppi, C. Young, +73 authors D. Yoon
  • Computer Science
  • 2017 ACM/IEEE 44th Annual International Symposium on Computer Architecture (ISCA)
  • 2017
TLDR
This paper evaluates a custom ASIC-called a Tensor Processing Unit (TPU)-deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN) and compares it to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the samedatacenters. Expand
Data Processing on FPGAs
TLDR
The results indicate that efficient usage of FPGAs involves non-trivial aspects such as having the right computation model (an asynchronous sorting network in this case); a careful implementation that balances all the design constraints in an FPGA; and the proper integration strategy to link theFPGA to the rest of the system. Expand
...
1
2
3
4
5
...