Compare AI products

Compare research use, compute requirements, licence conditions and evidence alongside product specifications.

Popular Benchmark Comparisons:

Showing up to 60 matches. Search to find more products.

Open Product Finder · View Benchmark Leaderboard ↗

Compare the same version, region and billing basis. Missing information does not mean a feature is absent. Verify prices and availability with the source.
Attribute
Research GPUs & Servers

NVIDIA DGX H100

NVIDIA
Research GPUs & Servers

NVIDIA H100

NVIDIA
Choose a product
CategoryResearch GPUs & ServersResearch GPUs & Servers
CompanyNVIDIANVIDIA
PricePrice not recorded — check supplierPrice not recorded — check supplier
StatusStatus not confirmedStatus not confirmed
AvailabilityCurrent availability and regional access not confirmed — check provider.Current availability and regional access not confirmed — check provider.
Last checked11/09/202611/09/2026
FeaturesIntegrated Hopper AI systemHopper data-center accelerator
Entry typeHardware / instrumentHardware / instrument
Access modeSupplier quotationSupplier quotation
AI roleCompute infrastructureCompute infrastructure
Input dataNot recordedNot recorded
Output dataNot recordedNot recorded
Licence conditionsHardware purchase does not provide rights to models, biological data or proprietary discovery software.Code and model parameters have separate terms. Parameter access must be obtained from the provider; commercial eligibility is not established here.
Commercial eligibilityNot independently confirmedNot independently confirmed
Compute requirementsInstitutional cluster planning: networking, storage, power, cooling and software operations must be sized separately.H100 80 GB; CPU/RAM and sequence databases are also required. Larger inputs can require configuration changes.
ValidationNot recordedNot recorded
TypeIntegrated Hopper AI systemHopper data-center accelerator
Intended useNot recordedNot recorded
CompatibilityNot recordedNot recorded
ManufacturerNVIDIANVIDIA
GPUs8 NVIDIA H100 GPUsNot recorded
Memory640 GB total GPU memorySXM: 80 GB; NVL: 94 GB
Format8U rackmountSXM module or NVL dual-slot PCIe card
CPUTwo Intel Xeon 8480C; 56 cores eachNot recorded
System memory2 TBNot recorded
WeightUp to 130.45 kgNot recorded
Operating temperature5–30 °CNot recorded
Biological applicationUsed in Recursion BioHive-2 for AI drug discoveryUsed in Recursion BioHive-2 for AI drug discovery
Research workflowPlan compute for AI drug discoveryPredict biomolecular structures
Evidence levelDocumented deploymentProvider-tested configuration
Integration evidencehttps://www.recursion.com/partnershttps://github.com/google-deepmind/alphafold3/blob/main/docs/performance.md
Laboratory handoffEditorial next step: define the experimental feedback loop with laboratory scientists. No connection to a catalog instrument has been verified.Editorial next step: select an appropriate structural or functional validation assay with the research team. No direct instrument integration has been verified.
BandwidthNot recordedSXM: 3.35 TB/s; NVL: 3.9 TB/s
PowerNot recordedSXM: up to 700 W; NVL: 350–400 W, configurable