Performance Benchmarks¶

Measured performance of the HCLS AI Factory on NVIDIA DGX Spark ($4,699). All timings represent end-to-end wall clock time under default configurations.

Summary¶

Metric	Value
End-to-end time (DNA to Drug Candidates)	< 5 hours
Traditional approach	6-18 months
Reduction	~99%
Hardware cost	$4,699 (single workstation)
Traditional infrastructure cost	$50K-500K+ (cluster + licenses)
Intelligence agents	11
Milvus collections	139
Total vectors	~47,691
Services	21

Hardware: NVIDIA DGX Spark¶

Component	Specification
GPU	NVIDIA GB10 Grace Blackwell
Memory	128 GB unified LPDDR5x
CPU	20 ARM cores (Grace)
Interconnect	NVLink-C2C (GPU-CPU)
Storage	NVMe SSD
Power	Desktop form factor
Price	$4,699

Engine 1: GPU Genomics (FASTQ to VCF)¶

Pipeline: NVIDIA Parabricks 4.6 with BWA-MEM2 + DeepVariant

Step	Time	GPU Utilization
Alignment (BWA-MEM2)	20-45 min	85-95%
Sorting + deduplication	included	--
Indexing (samtools)	2-5 min	CPU
Variant calling (DeepVariant)	10-35 min	85-95%
Total	120-240 min	85-95%

Metric	Value
Input	~200 GB paired-end FASTQ (HG002 WGS)
Output	~11.7 million variant calls (VCF)
Accuracy	>99% concordance (DeepVariant)
Speedup vs. CPU	10-50x
CPU baseline	24-48 hours

Engine 2: Evidence RAG (VCF to Target Hypothesis)¶

Pipeline: Milvus + BGE-small-en-v1.5 + Claude

Vector Database¶

Collection	Records	Embedding Time
ClinVar variants	~2.7M records	~45 min (one-time)
AlphaMissense predictions	71M records (sampled)	~30 min (one-time)
Clinker knowledge base	201 genes, 150+ diseases	~5 min (one-time)
Total annotated variants	3.56M	--

Query Performance¶

Operation	Latency
Vector embedding (BGE-small-en-v1.5)	< 50 ms
Milvus similarity search (top-10)	< 100 ms
Claude evidence synthesis	2-5 sec
End-to-end query	< 5 sec

Metric	Value
Embedding model	BGE-small-en-v1.5 (384 dimensions)
LLM	Claude (Anthropic)
Therapeutic areas covered	13
Target genes	201
Druggability rate	85%

Engine 3: Drug Discovery (Target to Molecules)¶

Pipeline: BioNeMo MolMIM + DiffDock + RDKit

Step	Time	Mode
Structure retrieval (RCSB PDB)	< 5 sec	API
Structure preparation	< 30 sec	CPU
Molecule generation (MolMIM)	10-60 sec	Cloud NIM
3D conformer generation (RDKit)	< 30 sec	CPU
Molecular docking (DiffDock)	2-8 min	Cloud NIM
Scoring and ranking (QED + Lipinski)	< 10 sec	CPU
Report generation	< 30 sec	CPU
Total	8-16 min	--

Metric	Value
Candidate molecules generated	10-100 per run
Docking poses per molecule	10
Drug-likeness filter	Lipinski Rule of 5 + QED
Seed compound (demo)	CB-5083 (VCP inhibitor)
PDB structures used (demo)	5FTK, 8OOI, 9DIL, 7K56

NIM Execution Modes¶

Mode	MolMIM	DiffDock	Best For
Cloud	health.api.nvidia.com	health.api.nvidia.com	DGX Spark (ARM64), no local GPU containers needed
Local	localhost:8001	localhost:8002	x86 workstations with dedicated GPU
Mock	Simulated output	Simulated output	Testing, CI/CD, demos without API keys

Intelligence Agent Benchmarks¶

1. Precision Biomarker Agent (Port 8502)¶

Metric	Value
Collections	11 (10 owned + 1 shared)
Biomarker interpretation	< 3 sec
Biological age estimation (PhenoAge/GrimAge)	< 2 sec
Pharmacogenomic profiling	< 3 sec
Disease trajectory detection	< 5 sec
PDF + FHIR R4 export	< 5 sec

2. Precision Oncology Agent (Port 8503)¶

Metric	Value
Collections	11 (10 owned + 1 shared)
Case creation	< 2 sec
MTB packet generation	10-30 sec
Trial matching	< 5 sec
Therapy ranking (CIViC/OncoKB)	< 5 sec
FHIR R4 bundle export	< 2 sec
Test suite	516 tests, < 1 sec

3. CAR-T Intelligence Agent (Port 8504)¶

Metric	Value
Collections	11 (10 owned + 1 shared)
Total vectors	6,266+
Query latency (evidence retrieval)	< 3 sec
Comparative analysis (e.g., 4-1BB vs CD28)	< 8 sec
Deep research mode	10-30 sec
PDF export	< 5 sec
Test suite	241 tests, < 1 sec

4. Imaging Intelligence Agent (Port 8505)¶

Metric	Value
Collections	10
NIM services	4 (VISTA-3D, MAISI, VILA-M3, Llama-3)
Workflow demo execution	5-15 sec per modality
Evidence query	< 5 sec
Comparative analysis	< 8 sec
FHIR R4 DiagnosticReport export	< 2 sec
Test suite	539 tests, ~3 sec

5. Precision Autoimmune Agent (Port 8506)¶

Metric	Value
Collections	10
Autoimmune conditions covered	13
Diagnostic engine evaluation	< 3 sec
Disease timeline construction	< 5 sec
Cross-modal genomic enrichment	< 3 sec
PDF clinical report export	< 5 sec
Evidence query	< 5 sec

6. Pharmacogenomics (PGx) Agent (Port 8507)¶

Metric	Value
Collections	15
Pharmacogenes	25
Drugs covered	100+
CPIC dosing algorithms	9
HLA associations	15
Star allele to phenotype conversion	< 1 sec
Phenoconversion (inhibitor adjustment)	< 1 sec
HLA adverse reaction screening	< 1 sec
Full PGx pipeline (genotype to dosing)	< 3 sec
FHIR R4 PGx report export	< 2 sec
Test suite	1,001+ tests

7. Cardiology Intelligence Agent (Port 8527)¶

Metric	Value
Collections	13 (12 owned + 1 shared)
Risk calculators	6 (ASCVD, HEART, CHA2DS2-VASc, HAS-BLED, MAGGIC, EuroSCORE II)
ASCVD 10-year risk calculation	14.6% example in < 1 sec
HEART score calculation	< 1 sec
CHA2DS2-VASc stroke risk	< 1 sec
GDMT optimization (4-pillar HFrEF)	< 3 sec
Clinical workflows	8 (CAD, HF, valvular, arrhythmia, cardiac MRI, stress test, prevention, cardio-oncology)
Cross-modal genomic triggers	< 3 sec (cardiomyopathies, channelopathies, FH)
Evidence query	< 5 sec
FHIR R4 export	< 2 sec
Test suite	1,927 tests, all passing

8. Neurology Intelligence Agent (Port 8528)¶

Metric	Value
Collections	14 (13 owned + 1 shared)
Clinical scales	10 (NIHSS, GCS, MoCA, MDS-UPDRS, EDSS, mRS, HIT-6, ALSFRS-R, ASPECTS, Hoehn & Yahr)
NIHSS stroke severity calculation	< 1 sec
GCS assessment	< 1 sec
MoCA cognitive screening	< 1 sec
MDS-UPDRS motor examination	< 1 sec
EDSS disability scoring (MS)	< 1 sec
ASPECTS early CT scoring	< 1 sec
Sub-domains covered	8 (cerebrovascular, degenerative, epilepsy, neuro-oncology, MS, movement, headache, neuromuscular)
Clinical workflows	8 (acute stroke, dementia eval, epilepsy focus, brain tumor, MS monitoring, Parkinson's, headache, neuromuscular)
Evidence query	< 5 sec
FHIR R4 export	< 2 sec

9. Rare Disease Diagnostic Agent (Port 8526)¶

Metric	Value
Collections	14 (13 owned + 1 shared)
Rare diseases covered	88
ACMG variant classification criteria	23
HPO phenotype matching	< 2 sec
Phenotype-to-gene ranking	< 3 sec
WES/WGS variant interpretation	< 5 sec
Gene therapy eligibility assessment	< 3 sec
Newborn screening evaluation	< 2 sec
Clinical workflows	10 (phenotype-driven, WES/WGS, metabolic screening, dysmorphology, neurogenetic, cardiac genetics, connective tissue, inborn errors, gene therapy, undiagnosed disease)
Evidence query	< 5 sec
FHIR R4 export	< 2 sec

10. Clinical Trial Intelligence Agent (Port 8521)¶

Metric	Value
Collections	14 (13 owned + 1 shared)
Patient-trial matching	< 5 sec
Eligibility criteria analysis	< 3 sec
Protocol design assistance	10-30 sec
Site selection ranking	< 5 sec
Safety signal detection	< 5 sec
Adaptive design evaluation	< 5 sec
Regulatory document intelligence	< 5 sec
Competitive intelligence query	< 5 sec
Clinical workflows	11 (protocol design, patient matching, site selection, eligibility optimization, adaptive design, safety signal, regulatory docs, competitive intel, diversity assessment, decentralized planning, general)
FHIR R4 / DOCX export	< 5 sec

11. Single-Cell Intelligence Agent (Port 8525)¶

Metric	Value
Collections	12 (11 owned + 1 shared)
Cell types annotated	57
Cell type identification query	< 3 sec
TME profiling (immune phenotype)	< 5 sec
Spatial niche analysis	< 5 sec
Drug response prediction	< 5 sec
Trajectory analysis	< 5 sec
Ligand-receptor interaction query	< 5 sec
CAR-T target validation	< 5 sec
Biomarker discovery query	< 5 sec
Clinical workflows	10 (cell type annotation, TME profiling, drug response, subclonal architecture, spatial niche, trajectory analysis, ligand-receptor, biomarker discovery, CAR-T target, treatment monitoring)
Evidence query	< 5 sec

Combined Test Suite¶

Agent	Tests	Time
Precision Biomarker	--	--
Precision Oncology	516	0.40 sec
CAR-T Intelligence	241	0.18 sec
Imaging Intelligence	539	3.20 sec
Precision Autoimmune	--	--
Pharmacogenomics	1,001+	--
Cardiology Intelligence	1,927	--
Neurology Intelligence	--	--
Rare Disease Diagnostic	--	--
Clinical Trial Intelligence	--	--
Single-Cell Intelligence	--	--
Total (measured)	4,224+	--

Infrastructure¶

Service Startup¶

Component	Cold Start	Warm Restart
Milvus	30-60 sec	10-15 sec
Landing page	< 5 sec	< 2 sec
RAG Chat UI	5-10 sec	< 5 sec
Drug Discovery UI	5-10 sec	< 5 sec
Agent UIs (11 agents)	5-10 sec each	< 5 sec
Full platform (21 services)	2-3 min	< 1 min

Resource Usage (Idle)¶

Resource	Usage
CPU	< 5% (20 ARM cores)
Memory	~8-12 GB (128 GB available)
GPU Memory	< 2 GB (128 GB unified)
Disk (platform + data)	~400-500 GB

Resource Usage (Peak -- Genomics Pipeline Running)¶

Resource	Usage
CPU	40-60%
Memory	30-50 GB
GPU	85-95% utilization
Disk I/O	High (FASTQ read + BAM write)

Scalability¶

Dimension	Current	Potential
Samples per day	3-6 (sequential)	10-20 (with pipeline parallelism)
Vector database	~47,691 vectors across 139 collections	100M+ (Milvus scales horizontally)
Knowledge base	201 genes, 13 areas	Expandable with additional collections
Concurrent users	5-10 (single workstation)	50+ (with load balancing)
Intelligence agents	11	Additional agents via plugin architecture

Methodology¶

All benchmarks measured on NVIDIA DGX Spark ($4,699) with Ubuntu 22.04 LTS
Timings are wall clock measurements averaged over 3 runs
GPU utilization measured via nvidia-smi and DCGM Exporter
Query latencies measured end-to-end including network overhead
Risk calculator and clinical scale timings measured as pure computation (no network)
Test suite timings measured via pytest with default configuration
"Traditional approach" estimates based on published literature for manual genomics + drug discovery workflows at academic medical centers
Agent benchmarks represent typical single-query performance under normal load

Clinical Decision Support Disclaimer

The HCLS AI Factory platform and its components are clinical decision support research tools. It is not FDA-cleared and is not intended as a standalone diagnostic device. All recommendations should be reviewed by qualified healthcare professionals. Apache 2.0 License.