Nscale provides direct MI300X, MI250X, MI210, H100, B200, GB200, MI325X, A100, GB300, H200 and V100 GPU infrastructure across London, Oslo and Reykjavik. Its services include inference, fine-tuning, colocation and managed AI infrastructure with Tier III, ISO 27001 and HIPAA coverage.
Nscale is a direct operator with a broad GPU lineup spanning current-generation NVIDIA capacity. Running its own network, it suits teams moving from experimentation into sustained production or those that need range across training, inference and fine-tuning without managing separate provider accounts.
The tradeoff: procurement is configuration-dependent and may require a quote process rather than instant checkout. Best fit for buyers with a clear workload plan.
Best for: Buyers who want a direct relationship with the operator that owns the metal. Not for: Teams prioritising a marketplace's breadth of third-party supply.
Confidence: HighEvidence reviewed: 8 checks
Confirmed
Direct operator classification
Nscale controls GPU infrastructure and sells access directly. Not a marketplace or hyperscaler resale offer.
REVIEWS
Buyer reviews
Moderated buyer feedback. Independent from the viabandwidth assessment above.
Had a good or bad experience with Nscale?
Be the first to review Nscale. Moderated and kept separate from viabandwidth’s own assessment.
Nscale is classified as a direct operator. Buyers who want a direct relationship with the operator that owns the metal. Teams prioritising a marketplace's breadth of third-party supply.
What GPU models does Nscale offer?
Nscale lists GB300, MI250X, H200 and 9 more across its published capacity.
How much does Nscale charge for GPU compute?
Nscale does not publish list pricing. Visit their site or use the contact form on this page for a quote.
Is Nscale's viabandwidth listing verified?
Not yet. This profile is compiled by viabandwidth from public sources. Nscale has not claimed it.
OVERVIEW
Company overview
Operator class
Direct operator
GPU models tracked
GB300, MI250X, H200 and 9 more
Services
Colocation, GPU compute, Managed services and 1 more
CAPACITY
GPU lineup by use case
Tracked GPU SKUs grouped by the workload they fit best.
Training
For multi-node runs where interconnect and cluster availability decide the outcome.
H200
QuoteClaim to publish
VRAM 141 GBClass Hopper
Best for Large-scale training where memory bandwidth decides the outcome.
A100
QuoteClaim to publish
VRAM 80 GBClass SXM
Best for Mature training stacks that value known performance and broad tooling support.
B200
QuoteClaim to publish
VRAM 192 GBClass Blackwell
Best for Frontier-scale training and teams optimizing around the newest platform.
MI300X
QuoteClaim to publish
VRAM 192 GBClass AMD
Best for Training and inference for teams evaluating the AMD ecosystem.
Inference
For predictable production serving, latency targets and steady utilisation.
GB300
QuoteClaim to publish
Best for GB300 workloads.
MI250X
QuoteClaim to publish
Best for MI250X workloads.
MI210
QuoteClaim to publish
Best for MI210 workloads.
L40
QuoteClaim to publish
VRAM 48 GBClass Ada
Best for Production serving for models that fit in 48 GB without SXM pricing.
COMMERCIALS
Pricing
Indicative published rates. Final cluster pricing depends on commit length, storage, networking and capacity.
Pricing on request. Visit Nscale’s site for a quote.
COMPARE
Similar GPU providers
Other direct operator operators buyers compare Nscale against.
Profile pages, buyer guides and model explainers stay open. Vendor packages cover what shows on this profile. Verification stays independent and is never for sale.