ArrayFire: a GPU acceleration platform

Malcolm, James; Yalamanchili, Pavan; McClanahan, Chris; Venugopalakrishnan, Vishwanath; Patel, Krunal; Melonakos, John

doi:10.1117/12.921122

Cited by 41 publications

(15 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While these offer high tunability, they are ill-suited to automatic code generation at runtime (see Section 2). The approach closest to ours is ArrayFire [17] which provides abstract vector operations backed by multiple hardware specific backends (CUDA, OpenCL and C++). ArrayFire even generates code at runtime but only for arithmetic expressions applied using a map operator.…”

Section: Related Workmentioning

confidence: 99%

Voodoo - a vector algebra for portable database performance on modern hardware

Pirk¹,

Moll²,

Zaharia³

et al. 2016

Proc. VLDB Endow.

View full text Add to dashboard Cite

In-memory databases require careful tuning and many engineering tricks to achieve good performance. Such database performance engineering is hard: a plethora of data and hardware-dependent optimization techniques form a design space that is difficult to navigate for a skilled engineer-even more so for a query compiler. To facilitate performanceoriented design exploration and query plan compilation, we present Voodoo, a declarative intermediate algebra that abstracts the detailed architectural properties of the hardware, such as multi-or many-core architectures, caches and SIMD registers, without losing the ability to generate highly tuned code. Because it consists of a collection of declarative, vector-oriented operations, Voodoo is easier to reason about and tune than low-level C and related hardware-focused extensions (Intrinsics, OpenCL, CUDA, etc.). This enables our Voodoo compiler to produce (OpenCL) code that rivals and even outperforms the fastest state-of-the-art in memory databases for both GPUs and CPUs. In addition, Voodoo makes it possible to express techniques as diverse as cacheconscious processing, predication and vectorization (again on both GPUs and CPUs) with just a few lines of code. Central to our approach is a novel idea we termed control vectors, which allows a code generating frontend to expose parallelism to the Voodoo compiler in a abstract manner, enabling portable performance across hardware platforms. We used Voodoo to build an alternative backend for Mon-etDB, a popular open-source in-memory database. Our backend allows MonetDB to perform at the same level as highly tuned in-memory databases, including HyPeR and Ocelot. We also demonstrate Voodoo's usefulness when investigating hardware conscious tuning techniques, assessing their performance on different queries, devices and data.

show abstract

Section: Related Workmentioning

confidence: 99%

Voodoo - a vector algebra for portable database performance on modern hardware

Pirk¹,

Moll²,

Zaharia³

et al. 2016

Proc. VLDB Endow.

View full text Add to dashboard Cite

show abstract

“…Many libraries including GPUSs [21], CUBLAS [22], MAGMA [23], ArrayFire [24] and Maximus [25] have been developed for multi-GPU systems to improve programmability and provide high performance. However, these libraries are limited to a specific programming model or domain such as linear algebra and signal/image processing and thus are not sufficient for handling a diverse set of general workloads.…”

Section: Multi-gpu Programmingmentioning

confidence: 99%

Multi-GPU System Design with Memory Networks

Kim

Lee

Jeong

et al. 2014

2014 47th Annual IEEE/ACM International Symposium on Microarchitecture

View full text Add to dashboard Cite

GPUs are being widely used to accelerate different workloads and multi-GPU systems can provide higher performance with multiple discrete GPUs interconnected together. However, there are two main communication bottlenecks in multi-GPU systems -accessing remote GPU memory and the communication between GPU and the host CPU. Recent advances in multi-GPU programming, including unified virtual addressing and unified memory from NVIDIA, has made programming simpler but the costly remote memory access still makes multi-GPU programming difficult. In order to overcome the communication limitations, we propose to leverage the memory network based on hybrid memory cubes (HMCs) to simplify multi-GPU memory management and improve programmability. In particular, we propose scalable kernel execution (SKE) where multiple GPUs are viewed as a single virtual GPU as a single kernel can be executed across multiple GPUs without modifying the source code. To fully enable the benefits of SKE, we explore alternative memory network designs in a multi-GPU system. We propose a GPU memory network (GMN) to simplify data sharing between the discrete GPUs while a CPU memory network (CMN) is used to simplify data communication between the host CPU and the discrete GPUs. These two types of networks can be combined to create a unified memory network (UMN) where the communication bottleneck in multi-GPU can be significantly minimized as both the CPU and GPU share the memory network. We evaluate alternative network designs and propose a sliced flattened butterfly topology for the memory network that scales better than previously proposed alternative topologies by removing local HMC channels. In addition, we propose an overlay network organization for unified memory network to minimize the latency for CPU access while providing high bandwidth for the GPUs. We evaluate trade-offs between the different memory network organization and show how UMN significantly reduces the communication bottleneck in multi-GPU systems.

show abstract

“…While this method can be efficiently implemented on multi-core processors (see Section 6.7), it is not suited to the Single-Instruction-Multiple-Threads execution model of massively parallel systems 3 . With the recent interest in GPU-based query processing [3,12,14,16,19,23], there is an obvious need for a efficient, massively parallel algorithm to solve the top-k problem. In fact, we found that two of the most mainstream GPU programming frameworks (Tensorflow and Arrayfire) [1,2] have open feature requests to add a top-k operator.…”

Section: Introductionmentioning

confidence: 99%

Efficient Top-K Query Processing on Massively Parallel Hardware

Shanbhag

Pirk

Madden

2018

Proceedings of the 2018 International Conference on Management of Data

View full text Add to dashboard Cite

A common operation in many data analytics workloads is to find the top-k items, i.e., the largest or smallest operations according to some sort order (implemented via LIMIT or ORDER BY expressions in SQL). A naive implementation of top-k is to sort all of the items and then return the first k, but this does much more work than needed. Although efficient implementations for top-k have been explored on traditional multi-core processors, there has been no prior systematic study of top-k implementations on GPUs, despite open requests for such implementations in GPU-based frameworks like TensorFlow 1 and ArrayFire 2 . In this work, we present several top-k algorithms for GPUs, including a new algorithm based on bitonic sort called bitonic top-k. The bitonic top-k algorithm is up to a factor of 15x faster than sort and 4x faster than a variety of other possible implementations for values of k up to 256. We also develop a cost model to predict the performance of several of our algorithms, and show that it accurately predicts actual performance on modern GPUs.

show abstract

ArrayFire: a GPU acceleration platform

Cited by 41 publications

References 3 publications

Voodoo - a vector algebra for portable database performance on modern hardware

Voodoo - a vector algebra for portable database performance on modern hardware

Multi-GPU System Design with Memory Networks

Efficient Top-K Query Processing on Massively Parallel Hardware

Contact Info

Product

Resources

About