Untitled

Shirazi, N.; Benyamin, D.; Luk, Wayne; Cheung, Peter Y. K.; Guo, Shaori

doi:10.1023/a:1008163222529

Cited by 6 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…FPGAs are amenable to partial reconfiguration, where only some of their internal architecture is changed. Such partial reconfiguration can reduce the time required for reprogramming and speed up reconfiguration [66,67]. An FPGA design like ours, with its closeness of different neutral networks, can serve to minimize the number of changes necessary to compute a new function.…”

Section: Discussionmentioning

confidence: 99%

The evolvability of programmable hardware

Raman

Wagner

2010

J. R. Soc. Interface.

View full text Add to dashboard Cite

In biological systems, individual phenotypes are typically adopted by multiple genotypes. Examples include protein structure phenotypes, where each structure can be adopted by a myriad individual amino acid sequence genotypes. These genotypes form vast connected 'neutral networks' in genotype space. The size of such neutral networks endows biological systems not only with robustness to genetic change, but also with the ability to evolve a vast number of novel phenotypes that occur near any one neutral network. Whether technological systems can be designed to have similar properties is poorly understood. Here we ask this question for a class of programmable electronic circuits that compute digital logic functions. The functional flexibility of such circuits is important in many applications, including applications of evolutionary principles to circuit design. The functions they compute are at the heart of all digital computation. We explore a vast space of 10 45 logic circuits ('genotypes') and 10 19 logic functions ('phenotypes'). We demonstrate that circuits that compute the same logic function are connected in large neutral networks that span circuit space. Their robustness or fault-tolerance varies very widely. The vicinity of each neutral network contains circuits with a broad range of novel functions. Two circuits computing different functions can usually be converted into one another via few changes in their architecture. These observations show that properties important for the evolvability of biological systems exist in a commercially important class of electronic circuitry. They also point to generic ways to generate fault-tolerant, adaptable and evolvable electronic circuitry.

show abstract

Section: Discussionmentioning

confidence: 99%

The evolvability of programmable hardware

Raman

Wagner

2010

J. R. Soc. Interface.

View full text Add to dashboard Cite

show abstract

“…The Bloom filters are used to store and query the K-mers in the genome assembly process. It has also been used in other applications such as the search engines [18], networking sites for spotting malicious URLs and most visited sites [19] and in online shopping sites to collect the most or recently viewed products [20]. Bloom filters use hash functions for mapping the input elements into the memory.…”

Section: Introductionmentioning

confidence: 99%

A Cache Efficient One Hashing Blocked Bloom Filter (OHBB) for Random Strings and the K-mer Strings in DNA Sequence

Prakasam

Manoharan

2022

Symmetry

View full text Add to dashboard Cite

Bloom filters are widely used in genome assembly, IoT applications and several network applications such as symmetric encryption algorithms, and blockchain applications owing to their advantages of fast querying, despite some false positives in querying the input elements. There are many research works carried out to improve both the insertion and querying speed or reduce the false-positive or reduce the storage requirements separately. However, the optimization of all the aforementioned parameters is quite challenging with the existing reported systems. This work proposes to simultaneously improve the insertion and querying speeds by introducing a Cache-efficient One-Hashing Blocked Bloom filter. The proposed method aims to reduce the number of memory accesses required for querying elements into one by splitting the memory into blocks where the block size is equal to the cache line size of the memory. In the proposed filter, each block has further been split into partitions where the size of each partition is the prime number. For insertion and query, one hash value is required, which yields different values when modulo divided with prime numbers. The speed is accelerated using simple hash functions where the hash function is called only once. The proposed method has been implemented and validated using random strings and symmetric K-mer datasets used in the gene assembly. The simulation results show that the proposed filter outperforms the Standard Bloom Filter in terms of the insertion and querying speed.

show abstract

“…The performance of the central processing unit (CPU) is not growing sufficiently quickly to handle the rapidly increasing amount of data, leading to demands for new processing methods to speed up database systems. The field-programmable gate array (FPGA) is a good choice to migrate the computational pressure of CPUs by offloading certain data-intensive and time-consuming operations in databases [1][2][3][4]. In relational databases, the join operator is intensively used to combine tuples from two or more relations given a common key, which requires a significant amount of computing resources.…”

Section: Introductionmentioning

confidence: 99%

Bloom Filter-Based Parallel Architecture for Accelerating Equi-Join Operation on FPGA

Xue

LIU

et al. 2021

Electronics

View full text Add to dashboard Cite

As one of the most important operations in relational databases, the join is data-intensive and time-consuming. Thus, offloading this operation using field-programmable gate arrays (FPGAs) has attracted much interest and has been broadly researched in recent years. However, the available SRAM-based join architectures are often resource-intensive, power-consuming, or low-throughput. Besides, a lower match rate does not lead to a shorter operation time. To address these issues, a Bloom filter (BF)-based parallel join architecture is presented in this paper. This architecture first leverages the BF to discard the tuples that are not in the join result and classifies the remaining tuples into different channels. Second, a binary search tree is used to reduce the number of comparisons. The proposed method was implemented on a Xilinx FPGA, and the experimental results show that under a match rate of 50%, our architecture achieved a high join throughput of 145.8 million tuples per second and a maximum acceleration factor of 2.3 compared to the existing SRAM-based join architectures.

show abstract

Untitled

Cited by 6 publications

References 9 publications

The evolvability of programmable hardware

The evolvability of programmable hardware

A Cache Efficient One Hashing Blocked Bloom Filter (OHBB) for Random Strings and the K-mer Strings in DNA Sequence

Bloom Filter-Based Parallel Architecture for Accelerating Equi-Join Operation on FPGA

Contact Info

Product

Resources

About