GenAx: A Genome Sequencing Accelerator

Fujiki, Daichi; Subramaniyan, Aran; Zhang, Tianjun; Zeng, Yu; Das, Reetuparna; Blaauw, David; Narayanasamy, Satish

doi:10.1109/isca.2018.00017

Cited by 80 publications

(65 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GPU power is collected by the Nvidia visual profiler. To estimate the performance of short read alignment, we selected the ASIC designs including RaceLogic [22] and GenAx [12], and ReRAM content address memory (CAM)-based PIM architecture including RADAR [25], Bio-CAM [24] and R-CAM [21]. To evaluate the performance of long read alignment, we compared FindeR against a recent ASIC chip Darwin [11].…”

Section: Schemesmentioning

confidence: 99%

“…To highlight the performance of FindR on short read alignment, we compared it against prior hardware accelerators [12], [21], [22], [24], [25] focusing on aligning short reads. These accelerators can boost the performance of either seeding [24], [25] or seed extension [12], [21], [22] for short reads. In contrast, only our FindeR is able to process both short read seeding by FM-Index searches and seed extension by FM-Index-based k-mismatch searches.…”

Section: Fm-index For Short Read Alignment 1) Performancementioning

confidence: 99%

“…To evaluate the seed quality, we implemented the GenAx hash-table-based SMEM seeding to generate seeds, and used BWA-MEM to produce alignment mappings on our CPU baseline. Because GenAx uses k = 12 [12], the SMEM length of GenAx is often ≤ 24. FindeR creates SMEMs with an average length of 48.…”

Section: ) Seed Qualitymentioning

confidence: 99%

“…Because of the high time complexity of APM algorithms for seed extensions, Hash The throughput and quality of long read seeding. recent work presents ASIC- [11], [12], [22], FPGA- [23] and PIM-based [21], [24], [25] accelerators to enhance the APM performance during seed extensions. Particularly, several works [21], [24], [25] take advantage of power-hungry and low-density ReRAM-based content address memories (CAMs) to implement the SW algorithm to align short reads.…”

Section: Fm-index For Long Read Alignmentmentioning

confidence: 99%

See 3 more Smart Citations

FindeR: Accelerating FM-Index-Based Exact Pattern Matching in Genomic Sequences through ReRAM Technology

Zokaee

Zhang

Jiang

2019

2019 28th International Conference on Parallel Architectures and Compilation Techniques (PACT)

View full text Add to dashboard Cite

Genomics is the critical key to enabling precision medicine, ensuring global food security and enforcing wildlife conservation. The massive genomic data produced by various genome sequencing technologies presents a significant challenge for genome analysis. Because of errors from sequencing machines and genetic variations, approximate pattern matching (APM) is a must for practical genome analysis. Recent work proposes FPGA, ASIC and even process-in-memory-based accelerators to boost the APM throughput by accelerating dynamic-programmingbased algorithms (e.g., Smith-Waterman). However, existing accelerators lack the efficient hardware acceleration for the exact pattern matching (EPM) that is an even more critical and essential function widely used in almost every step of genome analysis including assembly, alignment, annotation and compression.State-of-the-art genome analysis adopts the FM-Index that augments the space-efficient BWT with additional data structures permitting fast EPM operations. But the FM-Index is notorious for poor spatial locality and massive random memory accesses. In this paper, we propose a ReRAM-based process-in-memory architecture, FindeR, to enhance the FM-Index EPM search throughput in genomic sequences. We build a reliable and energyefficient Hamming distance unit to accelerate the computing kernel of FM-Index search using commodity ReRAM chips without introducing extra CMOS logic. We further architect a full-fledged FM-Index search pipeline and improve its search throughput by lightweight scheduling on the NVDIMM. We also create a system library for programmers to invoke FindeR to perform EPMs in genome analysis. Compared to state-of-the-art accelerators, FindeR improves the FM-Index search throughput by 83% ∼ 30K× and throughput per Watt by 3.5× ∼ 42.5K×.

show abstract

Section: Schemesmentioning

confidence: 99%

Section: Fm-index For Short Read Alignment 1) Performancementioning

confidence: 99%

Section: ) Seed Qualitymentioning

confidence: 99%

Section: Fm-index For Long Read Alignmentmentioning

confidence: 99%

See 2 more Smart Citations

FindeR: Accelerating FM-Index-Based Exact Pattern Matching in Genomic Sequences through ReRAM Technology

Zokaee

Zhang

Jiang

2019

2019 28th International Conference on Parallel Architectures and Compilation Techniques (PACT)

View full text Add to dashboard Cite

show abstract

“…Genome sequencing [8,21,34,35,37] is a cornerstone for enabling personalized medicine, global food security, and virus surveillance. The emerging nanopore genome sequencing technology [15] is revolutionizing the genome research, industry and market due to its ability to generate ultra-long DNA fragments, aka long reads, as well as provide portability.…”

Section: Introductionmentioning

confidence: 99%

Helix

Lou

Janga

Jiang

2020

Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques

View full text Add to dashboard Cite

Nanopore genome sequencing is the key to enabling personalized medicine, global food security, and virus surveillance. The state-ofthe-art base-callers adopt deep neural networks (DNNs) to translate electrical signals generated by nanopore sequencers to digital DNA symbols. A DNN-based base-caller consumes 44.5% of total execution time of a nanopore sequencing pipeline. However, it is difficult to quantize a base-caller and build a power-efficient processing-in-memory (PIM) to run the quantized base-caller. Although conventional network quantization techniques reduce the computing overhead of a base-caller by replacing floating-point multiply-accumulations by cheaper fixed-point operations, it significantly increases the number of systematic errors that cannot be corrected by read votes. The power density of prior nonvolatile memory (NVM)-based PIMs has already exceeded memory thermal tolerance even with active heat sinks, because their power efficiency is severely limited by analog-to-digital converters (ADC). Finally, Connectionist Temporal Classification (CTC) decoding and read voting cost 53.7% of total execution time in a quantized base-caller, and thus became its new bottleneck. In this paper, we propose a novel algorithm/architecture codesigned PIM, Helix, to power-efficiently and accurately accelerate nanopore base-calling. From algorithm perspective, we present systematic error aware training to minimize the number of systematic errors in a quantized base-caller. From architecture perspective, we propose a low-power SOT-MRAM-based ADC array to process analog-to-digital conversion operations and improve power efficiency of prior DNN PIMs. Moreover, we revised a traditional NVM-based dot-product engine to accelerate CTC decoding operations, and create a SOT-MRAM binary comparator array to process read voting. Compared to state-of-the-art PIMs, Helix improves base-calling throughput by 6×, throughput per Watt by 11.9× and per 2 by 7.5× without degrading base-calling accuracy. CCS CONCEPTS • Hardware → Spintronics and magnetic technologies; • Applied computing → Computational genomics.

show abstract