GenieHD: Efficient DNA Pattern Matching Accelerator Using Hyperdimensional Computing

Kim, Yeseong; Imani, Mohsen; Moshiri, Niema; Rosing, Tajana

doi:10.23919/date48585.2020.9116397

Cited by 71 publications

(41 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are 6 MAJ LUTs in the first stage of Figure 3(a), hence the number of inputs for the subsequent adder-tree reduces to 6 . From…”

Section: Proposed Method: Shear 31 Approximate Encodingmentioning

confidence: 99%

“…This phase is commonly referred to in literature as "associative search". Despite the simplicity of this "learning" scheme, HD computing has been successfully applied to a number of practical problems in the literature ranging from optimizing the performance of web-browsers [5], to DNA sequence alignment [6], bio-signal processing [7], robotics [8,9], and privacy preserving learning [10,11]. The primary appeal of HD computing lies in its amenability to implementation in modern hardware accelerators.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

SHEAR er

Khaleghi

Salamat

Thomas

et al. 2020

Proceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design

Self Cite

View full text Add to dashboard Cite

Hyperdimensional computing (HD) is an emerging paradigm for machine learning based on the evidence that the brain computes on highdimensional, distributed, representations of data. The main operation of HD is encoding, which transfers the input data to hyperspace by mapping each input feature to a hypervector, followed by a bundling procedure that adds up the hypervectors to realize the encoding hypervector. The operations of HD are simple and highly parallelizable, but the large number of operations hampers the efficiency of HD in embedded domain. In this paper, we propose SHEAR , an algorithmhardware co-optimization to improve the performance and energy consumption of HD computing. We gain insight from a prudent scheme of approximating the hypervectors that, thanks to error resiliency of HD, has minimal impact on accuracy while provides high prospect for hardware optimization. Unlike previous works that generate the encoding hypervectors in full precision and then and then perform ex-post quantization, we compute the encoding hypervectors in an approximate manner that saves resources yet affords high accuracy. We also propose a novel FPGA architecture that achieves striking performance through massive parallelism with low power consumption. Moreover, we develop a software framework that enables training HD models by emulating the proposed approximate encodings. The FPGA implementation of SHEAR achieves an average throughput boost of 104,904× (15.7×) and energy savings of up to 56,044× (301×) compared to state-of-the-art encoding methods implemented on Raspberry Pi 3 (GeForce GTX 1080 Ti) using practical machine learning datasets.

show abstract

“…There are 6 MAJ LUTs in the first stage of Figure 3(a), hence the number of inputs for the subsequent adder-tree reduces to 6 . From…”

Section: Proposed Method: Shear 31 Approximate Encodingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

SHEAR er

Khaleghi

Salamat

Thomas

et al. 2020

Proceedings of the ACM/IEEE International Symposium on Low Power Electronics and Design

Self Cite

View full text Add to dashboard Cite

show abstract

“…As a practical example, in bioinformatics it is common to search for regions of high similarity between a "reference" and "query" genome. Work in (Imani et al, 2018) and (Kim et al, 2020) explored the use HD computing to accelerate this process by encoding short segments of DNA and estimating similarity on the HD representations.…”

Section: Comparing Structuresmentioning

confidence: 99%

A Theoretical Perspective on Hyperdimensional Computing

ThomasAnthony¹,

DasguptaSanjoy²,

RosingTajana³

2021

jair

Self Cite

View full text Add to dashboard Cite

Hyperdimensional (HD) computing is a set of neurally inspired methods for obtaining highdimensional, low-precision, distributed representations of data. These representations can be combined with simple, neurally plausible algorithms to effect a variety of information processing tasks. HD computing has recently garnered significant interest from the computer hardware community as an energy-efficient, low-latency, and noise-robust tool for solving learning problems. In this review, we present a unified treatment of the theoretical foundations of HD computing with a focus on the suitability of representations for learning.

show abstract

“…Thanks to much researches undergone on DNA sequences pattern matching have been actively undergone [12], [13], it is revealed that particular genes are related to a feature of human personality. C15orf17, message passing interface (MPI), secretory carrier membrane protein 2 (SCAMP2), protein kinase (ULK3), Cytochrome c oxidase subunit 5a (COX5A) are associated with neuroticism [14], Fyn interacts with Nmethyl-D-aspartate (NDMA) receptors and inositol-1,4,5trisphosphate (IP3)-gated channels to regulate calcium influx and intracellular release in the post-synaptic density.…”

Section: B Dna Matchingmentioning

confidence: 99%

Virtual Simulation for Entertainment using Genetic Information

Kim¹,

Lim²,

You³

et al. 2020

IJACSA

View full text Add to dashboard Cite

The genetic information has been researched to predict the disease and to discover the clue of biological causes, especially in the medical field. However, based on the reliability of genetic information, it also gives a powerful realistic experience with VR devices. In this paper, we developed a dating simulation game that users can meet celebrities. The human personality of the celebrity, the favorability feedback depends on each choice of the conversation and a proper choice creation is based on the genetic information of the user and a celebrity. Besides, a method for utilizing genetic DNA information for virtual simulations is proposed. In this study, we have established that DNA information is related to human relationships both in the real and virtual worlds. Also, we concluded that the DNA information contributes to the development of interpersonal and mental health. However, a person's personality or preference for a specific situation or object, etc. are not determined only through genes. Therefore, more quantitative and multifaceted studies will need to be conducted on the effect of genes on personal preferences. We only experimented with virtual characters in virtual reality. It would be meaningful to proceed with the human experiment not only with a virtual character in the virtual environment, but also another human in the virtual environment. Eventually, the result of the experiment between virtual characters and humans should be compared.

show abstract

GenieHD: Efficient DNA Pattern Matching Accelerator Using Hyperdimensional Computing

Cited by 71 publications

References 18 publications

SHEAR er

SHEAR er

A Theoretical Perspective on Hyperdimensional Computing

Virtual Simulation for Entertainment using Genetic Information

Contact Info

Product

Resources

About