Protocol

DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1)

Wall-clock time, on-demand cost, speedup and cost saving for DeepVariant on CPU machines and 2, 4 or 8 GPUs.

Overview

Wall-clock time, on-demand cost, speedup and cost saving for DeepVariant on CPU machines and 2, 4 or 8 GPUs.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

11 recorded evaluations, 42 metric rows. A comparison chart has not yet been validated for these results. The table retains the individual findings and their sources.

View coverage and remaining gaps across all benchmarks

Results

Results are available, but no reviewed comparison panel is linked in this release.

All evaluations

11 evaluations · 42 results. Different protocols are not a single leaderboard.

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Source checking is not independent reproduction. Release 2026-10-09-8cc1db47c7f9.

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Baseline coverage

Reference methods help show what a model adds beyond simple controls. We track a null control and a conventional method for each protocol.

0 of 2 active baseline roles have published Rewire measurements in this release. Measurements on a selected protocol do not establish coverage of an entire suite.

No execution recipe linked to this protocol. Recipe availability does not establish a completed evaluation.

External evaluations: 11

Literature evidence is not a Rewire measurement. Executed but unpublished runs and private review status are not included.

Null control

Proposed control: requires review

Select a task-valid null control after reviewing inputs and metric

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Conventional reference

Proposed control: requires review

Select an upstream conventional reference after reviewing the full protocol

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Protocol coverage CSV (gzip) · Model evaluation matrix (gzip) · Source table (gzip) · Release and checksums (gzip)

Coverage is derived from release 2026-10-09-8cc1db47c7f9. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.

Run instructions

No runnable recipe has been reviewed for this protocol. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

Strengths, limitations and unresolved questions

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-09-8cc1db47c7f9
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

All evidence (JSONL) (gzip)How to read the evidence

Sources and history

Release 2026-10-09-8cc1db47c7f9 · Record review: source checked

1 source records and release history

Accelerating genomic workflows using NVIDIA Parabricks · Original source · BMC Bioinformatics 24:221, published 2023-05-31; PMC10230726 full-text XML

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: model-execution-20261009-protocol-oconnell2023-deepvariant

areas: dna-genomes
contexts: clinical_research
protocol: Germline DeepVariant workflow timed end to end on each machine; fold acceleration and % cost-savings are relative to the CPU machine on the same cloud platform.
version: Table 1 rows for Variant-caller 'DeepVariant'
metric: runtime
unit: minute
limitations: One recorded run per configuration (Methods: 'we recorded the time of our final workflow run'); no repeats printed. DGX runs were repeated at least three times and only the final run is shown (Methods 'DGX configuration').; Costs are on-demand cloud prices at the time of the study and exclude storage and licences; DGX rows have no cost.; Authors are from Deloitte Consulting, an NVIDIA, AWS and Google alliance partner (Competing interests).; AWS GPU model is unresolved: the Table 1 footnote says the AWS GPU rows are the p3 family with V100 GPUs, but Results 'GPU performance across cloud platforms' paragraph 2 attributes the AWS savings printed in Table 1 (for example 63% for HaplotypeCaller with 4 GPUs) to the p4 machine with A100 GPUs. The printed AWS GPU costs equal runtime times 12.24 or 31.22 USD per hour, not the 32.8 USD per hour p4d price in the text.; Germline pipelines run from FASTQ to unfiltered VCF; accuracy is not reported.
source locator: Table 1, rows 1-11 (Variant-caller 'DeepVariant')

Related records

Suggest a correction

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	29.9 compute-cost us-dollar · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 1 (AWS, C6i.8xlarge, DeepVariant), column 'Cost ($)'
Configuration: CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	1320 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 1 (AWS, C6i.8xlarge, DeepVariant), column 'Time (min)'
Configuration: CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	22 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on CPU tools on AWS c6i.8xlarge (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 1 (AWS, C6i.8xlarge, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	29.6 compute-cost us-dollar · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 2 (AWS, GPU2, DeepVariant), column 'Cost ($)'
Configuration: Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	0.83% cost-saving percent · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 2 (AWS, GPU2, DeepVariant), column '% cost-savings'
Configuration: Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	145 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 2 (AWS, GPU2, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	2.42 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 2 (AWS, GPU2, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	9.07 speedup unitless · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 2 (AWS, GPU2, DeepVariant), column 'Fold acceleration'
Configuration: Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	19.8 compute-cost us-dollar · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 3 (AWS, GPU4, DeepVariant), column 'Cost ($)'
Configuration: Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	33.7% cost-saving percent · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 3 (AWS, GPU4, DeepVariant), column '% cost-savings'
Configuration: Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	97.1 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 3 (AWS, GPU4, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	1.62 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 3 (AWS, GPU4, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	13.6 speedup unitless · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 3 (AWS, GPU4, DeepVariant), column 'Fold acceleration'
Configuration: Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	21.9 compute-cost us-dollar · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 4 (AWS, GPU8, DeepVariant), column 'Cost ($)'
Configuration: Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	26.5% cost-saving percent · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 4 (AWS, GPU8, DeepVariant), column '% cost-savings'
Configuration: Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	42.2 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 4 (AWS, GPU8, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	0.7 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 4 (AWS, GPU8, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	31.2 speedup unitless · higher Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 GPUs on AWS (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 4 (AWS, GPU8, DeepVariant), column 'Fold acceleration'
Configuration: Parabricks 3.7.0-1, 2 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	87.9 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 9 (DGX, GPU2, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 2 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	1.47 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 9 (DGX, GPU2, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 4 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	49.1 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 10 (DGX, GPU4, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 4 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	0.82 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 4 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 10 (DGX, GPU4, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 8 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	27.1 runtime minute · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 11 (DGX, GPU8, DeepVariant), column 'Time (min)'
Configuration: Parabricks 3.7.0-1, 8 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	0.45 runtime hour · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 8 A100 GPUs on NVIDIA DGX A100 (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 11 (DGX, GPU8, DeepVariant), column 'Time (h)'
Configuration: Parabricks 3.7.0-1, 2 A100 GPUs on GCP a2-highgpu (O'Connell et al. 2023)	Protocol: DeepVariant execution on CPU and GPU cloud and DGX machines (O'Connell et al. 2023 Table 1) Dataset: HG002 (GIAB) WGS FASTQ down-sampled to 30x, precisionFDA Truth Challenge V2	19.4 compute-cost us-dollar · lower Uncertainty: Not reported by the source: Single recorded run per cell; no repeats or intervals printed Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DeepVariant on Parabricks 3.7.0-1, 2 A100 GPUs on GCP a2-highgpu (O'Connell et al. 2023) model-execution-20261009-protocol-oconnell2023-deepvariant Aggregation: Not reported Accelerating genomic workflows using NVIDIA Parabricks · Table 1, row 6 (GCP, GPU2, DeepVariant), column 'Cost ($)'