1 live offerings · 1 tracked regions
Launch Now 1
Loading
Loading workspaces…
engineering · GPU0090
Specifications with their source, supported precisions, interconnect, benchmark coverage and which published models fit NVIDIA A100 80GB — provider does not specify PCIe/SXM.
| Specification | Value | Source | Kovara registry |
|---|---|---|---|
| Memory capacity | 80GB | Kovara registry · verified 2026-09-25 | used |
| Memory type | HBM2E | Kovara registry · verified 2026-09-25 | used |
| Memory bandwidth | — | Not published | — |
| Board power (TDP) | — | Not published | — |
| Form factor | — | Not published | — |
| NVLink | Yes | Kovara registry · verified 2026-09-25 | used |
| MIG partitioning | Yes | Kovara registry · verified 2026-09-25 | used |
No public benchmark runs yet
Nobody has published a hardware run on this accelerator. Measured numbers appear here once a run of a compute repository is published.
AcceleratorsGPU0090
NVIDIA
Core hardware and lifecycle, drawn from Kovara master data.
What the recorded specifications mean in practice, and what this accelerator is used for.
Specs, explained
HBM (High Bandwidth Memory) is stacked in layers and sits right next to the processor on the same package. That short, wide path is what gives accelerators with HBM their memory bandwidth, so they are usually chosen for work limited by how fast data reaches the cores, such as large-model inference and HPC simulation.
VRAM is the working space on the card. Model weights, activations and datasets have to fit in it, or be split across several GPUs. Rule of thumb for the weights alone at 16-bit precision (about 2 GB per billion parameters): up to roughly 40 billion parameters, before activations and cache. This is arithmetic on the recorded VRAM, not a benchmark.
Ampere brought third-generation Tensor Cores, TF32 precision and Multi-Instance GPU (MIG) to its data-center parts. (General context about the architecture, not a statement about this specific card.)
Launched in 2020. Kovara's records list it as a previous-generation part.
NVLink is a direct GPU-to-GPU link. Where it is supported, GPUs in the same server can exchange data faster than over PCIe alone, which matters when one training job is spread across several GPUs.
Multi-Instance GPU (MIG) can divide one physical GPU into isolated slices, each with its own memory and compute, so several smaller workloads can share a card.
These are the uses Kovara has recorded for this part. The notes explain the workload in general terms; they are not performance claims.
Figures are Kovara master data, last verified 25 Sept 2026. Explanations describe the technology in general; they are not benchmarks and say nothing about price or availability.
Explore this accelerator on ComputePublic Kovara observations; not a claim of complete or universal market coverage. Incompatible raw price units are excluded.
Observed GPU-hour range
USD / GPU-hour
Spread $0.71 USD / GPU-hour
Providers represented by current Kovara-tracked offerings.
1 live offerings · 1 tracked regions
Launch Now 1
8 live offerings · 1 tracked regions
In stock 8
1 live offerings · 1 tracked regions
Self-serve 1
4 live offerings · 1 tracked regions
Available 4
2 live offerings · 1 tracked regions
Published configuration; live capacity unconfirmed 2
3 live offerings · 1 tracked regions
Availability unconfirmed 1 · Unconfirmed 2
4 live offerings · 1 tracked regions
Unconfirmed 4
4 live offerings · 1 tracked regions
Self-serve 4
1 live offerings · 1 tracked regions
Spot inventory 1
1 live offerings · 1 tracked regions
Publicly listed 1
1 live offerings · 1 tracked regions
Self-serve 1
1 live offerings · 1 tracked regions
Self-serve 1
1 live offerings · 1 tracked regions
Live marketplace — query required 1
Regions represented by current offerings; this does not describe complete global availability.
2 live offerings · 1 providers
Region profile →1 live offerings · 1 providers
Region profile →29 live offerings · 11 providers
Current configurations, recorded availability, confirmed capacity and the evidence behind each price.
| Provider | Region | Configuration | Availability | Price | Commercial model | Trust / evidence | Action |
|---|---|---|---|---|---|---|---|
| OFR-0390Hyperstack | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXMConfiguration quantity not specified | Availability unconfirmedConfirmed capacity: Not confirmed | $0.95 USD / GPU-hourOriginal: USD 0.95 GPU/hour from | ReservedGPU VM · Starting hourly reservation rate | Confidence AChecked 49 days ago · StaleConfirmed 29 Sept 2026Source: SRC000002 | Inspect |
| OFR-0392Hyperstack | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXMConfiguration quantity not specified | UnconfirmedConfirmed capacity: Not confirmed | $1.08 USD / GPU-hourOriginal: USD 1.08 GPU/hour | SpotGPU VM · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000002 | Inspect |
| OFR-0347Thunder Compute | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Self-serveConfirmed capacity: Not confirmed | $1.09 USD / GPU-hourOriginal: USD 1.09 GPU/hour | On-DemandGPU VM · Per minute | Confidence AChecked 16 days agoConfirmed 24 Sept 2026Source: SRC000042 | Inspect |
| OFR-0756Massed Compute | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM2 GPU configuration · 2 GPUs per node | Self-serveConfirmed capacity: Not confirmed | $1.35 USD / GPU-hourOriginal: USD 2.70 instance-hour | On-DemandGPU VM · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000037 | Inspect |
| OFR-0388Hyperstack | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXMConfiguration quantity not specified | UnconfirmedConfirmed capacity: Not confirmed | $1.35 USD / GPU-hourOriginal: USD 1.35 GPU/hour | On-DemandGPU VM · Per minute | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000002 | Inspect |
| OFR-0758Massed Compute | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPU configuration · 8 GPUs per node | Self-serveConfirmed capacity: Not confirmed | $1.35 USD / GPU-hourOriginal: USD 10.80 instance-hour | On-DemandGPU VM · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000037 | Inspect |
| OFR-0757Massed Compute | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM4 GPU configuration · 4 GPUs per node | Self-serveConfirmed capacity: Not confirmed | $1.35 USD / GPU-hourOriginal: USD 5.40 instance-hour | On-DemandGPU VM · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000037 | Inspect |
| OFR-0755Massed Compute | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Self-serveConfirmed capacity: Not confirmed | $1.35 USD / GPU-hourOriginal: USD 1.35 instance-hour | On-DemandGPU VM · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000037 | Inspect |
| OFR-0796Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM2 GPU configuration · 2 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.39 USD / GPU-hourOriginal: USD 2.78 instance-hour | CommittedGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0800Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPU configuration · 8 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.39 USD / GPU-hourOriginal: USD 11.12 instance-hour | CommittedGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0316Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | In stockConfirmed capacity: Not confirmed | $1.39 USD / GPU-hourOriginal: USD 1.39 instance-hour | CommittedGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0798Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM4 GPU configuration · 4 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.39 USD / GPU-hourOriginal: USD 5.56 instance-hour | CommittedGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0562JarvisLabs | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | UnconfirmedConfirmed capacity: Not confirmed | $1.49 USD / GPU-hourOriginal: USD 1.49 instance-hour | On-DemandGPU VM / template · Per minute | Confidence AChecked 17 days agoConfirmed 23 Sept 2026Source: SRC000023 | Inspect |
| OFR-0563JarvisLabs | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM2 GPU configuration · 2 GPUs per node | UnconfirmedConfirmed capacity: Not confirmed | $1.49 USD / GPU-hourOriginal: USD 2.98 instance-hour | On-DemandGPU VM / template · Per minute | Confidence AChecked 17 days agoConfirmed 23 Sept 2026Source: SRC000023 | Inspect |
| OFR-0564JarvisLabs | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM4 GPU configuration · 4 GPUs per node | UnconfirmedConfirmed capacity: Not confirmed | $1.49 USD / GPU-hourOriginal: USD 5.96 instance-hour | On-DemandGPU VM / template · Per minute | Confidence AChecked 17 days agoConfirmed 23 Sept 2026Source: SRC000023 | Inspect |
| OFR-0565JarvisLabs | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPU configuration · 8 GPUs per node | UnconfirmedConfirmed capacity: Not confirmed | $1.49 USD / GPU-hourOriginal: USD 11.92 instance-hour | On-DemandGPU VM / template · Per minute | Confidence AChecked 17 days agoConfirmed 23 Sept 2026Source: SRC000023 | Inspect |
| OFR-0317Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | In stockConfirmed capacity: Not confirmed | $1.79 USD / GPU-hourOriginal: USD 1.79 instance-hour | On-DemandGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0797Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM2 GPU configuration · 2 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.79 USD / GPU-hourOriginal: USD 3.58 instance-hour | On-DemandGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0799Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM4 GPU configuration · 4 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.79 USD / GPU-hourOriginal: USD 7.16 instance-hour | On-DemandGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0801Civo | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPU configuration · 8 GPUs per node | In stockConfirmed capacity: Not confirmed | $1.79 USD / GPU-hourOriginal: USD 14.32 instance-hour | On-DemandGPU compute · Hourly | Confidence AChecked todayConfirmed 29 Sept 2026Source: SRC000058 | Inspect |
| OFR-0325E2E Networks | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXMConfiguration quantity not specified | Self-serveConfirmed capacity: Not confirmed | $2.10 USD / GPU-hourOriginal: USD 2.10 GPU/hour | On-DemandGPU instance · Per-minute billing | Confidence AChecked 16 days agoConfirmed 29 Sept 2026Source: SRC000040 | Inspect |
| OFR-0214Runpod | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Self-serveConfirmed capacity: Not confirmed | $2.72 USD / GPU-hourOriginal: USD 2.72 GPU/hour | ServerlessServerless GPU worker · Per second / hourly equivalent | Confidence AChecked 49 days ago · StaleConfirmed 29 Sept 2026Source: SRC000022 | Inspect |
| OFR-0691Vast.ai | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Live marketplace — query requiredConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: Not available | On-DemandMarketplace GPU instance · Per-second / dynamic | Confidence BChecked 55 days ago · StaleNot confirmedSource: SRC000052 | Inspect |
| OFR-0639Novita AI | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Spot inventoryConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: Not available | SpotGPU instance · Live marketplace pricing | Confidence AChecked 55 days ago · StaleConfirmed 23 Sept 2026Source: SRC000050 | Inspect |
| OFR-0519Gcore | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPU configuration · 8 GPUs per node | AvailableConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: EUR 10.08 instance-hour | On-DemandInference GPU compute · Per minute | Confidence AChecked 4 days agoConfirmed 23 Sept 2026Source: SRC000015 | Inspect |
| OFR-0518Gcore | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM4 GPU configuration · 4 GPUs per node | AvailableConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: EUR 5.04 instance-hour | On-DemandInference GPU compute · Per minute | Confidence AChecked 4 days agoConfirmed 23 Sept 2026Source: SRC000015 | Inspect |
| OFR-0517Gcore | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM2 GPU configuration · 2 GPUs per node | AvailableConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: EUR 2.52 instance-hour | On-DemandInference GPU compute · Per minute | Confidence AChecked 4 days agoConfirmed 23 Sept 2026Source: SRC000015 | Inspect |
| OFR-0516Gcore | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | AvailableConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: EUR 1.26 instance-hour | On-DemandInference GPU compute · Per minute | Confidence AChecked 4 days agoConfirmed 23 Sept 2026Source: SRC000015 | Inspect |
| OFR-0473OVHcloud | Unspecified, Multiple / Unspecified | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Publicly listedConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: EUR 3.00 instance-hour | On-DemandManaged AI GPU resource — product scope unconfirmed · Hourly | Confidence AChecked 16 days agoConfirmed 24 Sept 2026Source: SRC-OVHCLOUD-CATALOG-API-20261031 | Inspect |
| OFR-0360Ace Cloud | Noida, India | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU configuration · 1 GPU per node | Launch NowConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: INR 90000.00 per month | On-DemandGPU cloud VM · Monthly | Confidence AChecked 54 days ago · StaleConfirmed 24 Sept 2026Source: SRC000144 | Inspect |
| OFR-0929Highreso GPUSOROBAN | Ishikawa, Japan | NVIDIA A100 80GB — provider does not specify PCIe/SXM1 GPU per node | Published configuration; live capacity unconfirmedConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: JPY 398.00 instance-hour | On-DemandGPU cloud instance · Hourly or monthly (¥243,631/month) | Confidence BChecked 12 days agoNot confirmedSource: SRC-APAC-20261105-HIGHRESO-COMPUTE | Inspect |
| OFR-0931Highreso GPUSOROBAN | Ishikawa, Japan | NVIDIA A100 80GB — provider does not specify PCIe/SXM8 GPUs per node | Published configuration; live capacity unconfirmedConfirmed capacity: Not confirmed | Comparable price unavailableOriginal: JPY 1985940.00 server-month | On-DemandGPU server · Monthly only (no hourly rate published) | Confidence BChecked 12 days agoNot confirmedSource: SRC-APAC-20261105-HIGHRESO-COMPUTE | Inspect |
Recorded hardware attributes. Not tracked is distinct from No.
Public evidence and confidence signals supporting Kovara-tracked coverage.