GPUBank vs Verda (ex-DataCrunch)
Verda, formerly DataCrunch, is a GPU cloud from Helsinki, founded in 2020, that rents NVIDIA instances of 1 to 8 GPUs, self-service InfiniBand clusters and serverless containers, all in three Finnish data centers. Its H100 SXM costs $3.56 per GPU-hour on demand; GPUBank charges $1.85. Verda offers self-service Blackwell, clusters and a full set of compliance reports; GPUBank bills by the second, keeps a stopped instance at disk cost and takes crypto.
GPUBank
Tallinn, Estonia
H100 SXM
$1.85 / h
- vCPU, RAM, NVMe, IPv4 and traffic included
- Per-second billing after the health check
- Top up in Bitcoin, USDT and 12+ other coins
vCPU, RAM, NVMe, IPv4 and traffic included
Verda (ex-DataCrunch)
Helsinki, Finland · since 2020
H100 SXM
$3.56 / h
- Self-service GB300, B300 and B200, from 1 to 8 GPUs
- Instant InfiniBand clusters with Kubernetes or Slurm pre-installed
- Serverless containers that scale to zero
Verda (ex-DataCrunch)
European cloud
Finland25 points compared25 of their pages readprices last checked on 29 September 2026page updated 2 November 2026
The verdict
Blackwell, clusters and compliance at Verda. Lower prices and per-second billing at GPUBank.
Verda suits teams that want GB300, B300 or B200 without a sales call, self-service InfiniBand clusters, serverless containers and SOC 2, ISO 27001 and C5 reports, with data kept in Finland. For the six GPUs both of us rent, GPUBank costs 39 to 64 % less per GPU-hour on demand. Verda charges prepaid 10-minute increments and keeps billing a shut-down instance until it is deleted; GPUBank bills by the second and bills only the disk once you stop. Verda takes cards and bank transfers; GPUBank takes crypto and asks for no card.
Verda (ex-DataCrunch)GPUBank 01The real cost
The same GPUs, side by side, per hour.
Every GPU model both of us rent, at each provider's own published price for one card, in US dollars, excluding VAT. Prices last checked on 29 September 2026.
| GPU | Verda (ex-DataCrunch) per hour | GPUBank per hour | Difference | 100 h at Verda (ex-DataCrunch) | 100 h at GPUBank |
|---|---|---|---|---|---|
 B200 SXM180 GB |
$6.85 | $4.15 | −39 % | $685.00 | $415.00 |
 H200 SXM141 GB |
$4.64 | $2.49 | −46 % | $464.00 | $249.00 |
 H100 SXM80 GB |
$3.56 | $1.85 | −48 % | $356.00 | $185.00 |
 RTX PRO 600096 GB |
$1.96 | $1.15 | −41 % | $196.00 | $115.00 |
 A100 SXM80 GB |
$1.80 | $0.95 | −47 % | $180.00 | $95.00 |
 L40S48 GB |
$1.54 | $0.55 | −64 % | $154.00 | $55.00 |
A fine-tune, 40 h on an H100 SXM
Verda (ex-DataCrunch)
$142.40
GPUBank
$74.00
Saves $68.40 at GPUBank (−48 %)
An inference box, a month, 720 h, on an L40S
Verda (ex-DataCrunch)
$1,108.80
GPUBank
$396.00
Saves $712.80 at GPUBank (−64 %)
A long-context job, 40 h on an H200
Verda (ex-DataCrunch)
$185.60
GPUBank
$99.60
Saves $86.00 at GPUBank (−46 %)
02Their catalogue
Every GPU Verda (ex-DataCrunch) lists.
12 GPU models listed by Verda (ex-DataCrunch), each at its lowest public price per GPU-hour. 6 of them are also rented at GPUBank (in blue).
GB300288 GB · on demand$9.72- B300268 GB · on demand$8.45
- B200180 GB · on demand$6.85
- H200141 GB · on demand$4.64
- H100 SXM80 GB · on demand$3.56
- A100 SXM 80GB80 GB · on demand$1.80
- A100 40GB40 GB · on demand$1.29
- L40S48 GB · on demand$1.54
V10016 GB · on demand$0.19- RTX PRO 600096 GB · on demand$1.96
RTX 6000 Ada48 GB · on demand$1.16
RTX A600048 GB · on demand$0.64At GPUBank
vCPU, RAM, NVMe and traffic included. All prices
03What the price includes
What appears on the bill besides the GPU.
The same H100 SXM session of 2 h 14 m 03 s, line by line, as each provider describes its billing.
Verda (ex-DataCrunch)
2 h 14 m 03 s on H100 SXM
- GPU · $3.56 / h
- $7.95
- vCPU and RAM
- Included
- Storage
- $0.20 per GiB-month (NVMe block)
- Egress
- Not published
- Public IPv4
- Not published
- Billing granularity
- Prepaid 10-minute increments
- Minimum charge
- One 10-minute increment, unused part refunded
GPUBank
2 h 14 m 03 s on H100 SXM
- GPU · $1.85 / h
- $4.13
- vCPU and RAM
- Included
- Storage
- Local NVMe included; persistent disk $0.03 per GB-month
- Egress
- Included
- Public IPv4
- Included
- Billing granularity
- Per second, from the health check
- Minimum charge
- None
While stopped
What keeps running on the bill.
Verda (ex-DataCrunch)
Billed while shut down, until deleted.
GPUBank
Only the persistent disk, $0.03 per GB-month. Compute, IPv4 and scratch cost nothing.
04Where the GPUs run
1 country against 4, and where they are.
Countries where Verda (ex-DataCrunch) says its GPUs run, from its own region list or documentation, against GPUBank's four regions. A pin marks a country, not the exact site.
Finland
Paris
Frankfurt
Ashburn
Singapore
Verda (ex-DataCrunch)1
- Finland
GPUBank4
Paris
Frankfurt
Ashburn
Singapore
Verda (ex-DataCrunch) · 1 countryGPUBank · 4 regions
05Paying and signing up
How each one takes your money.
Verda (ex-DataCrunch)
Payments and sign-up
Payment methods
Cards
Bank transfer
No cryptoMinimum top-up Not published for cards
To start
- 1Email and password
- 2Card
Billing modelPrepaid credits
GPUBank
Payments and sign-up
Payment methods
A prepaid USD balance, topped up in crypto. Every GPU-second is drawn from it.
Minimum top-up $25
To start
- 1Email and password
- 2Top up in crypto
- 3Launch a GPU
Billing modelPrepaid balance
06Trust and support
Certificates, uptime and who answers.
| Verda (ex-DataCrunch) | GPUBank | |
|---|---|---|
| SOC 2 | Type II | Not published |
| ISO 27001 | Yes | Not published |
| HIPAA | Not published | No |
| GDPR DPA | DPA available | Published DPA, applies automatically |
| Availability target | No numeric SLA | 99.9 % monthly, per region |
| Status page | status.verda.com | Public status page |
| Support | Chat, email, Discord, forum | Tickets from a $25 balance |
07Their site, our site
What you see before you pay.
Both sites as a visitor sees them, without an account: the Verda (ex-DataCrunch) pricing page captured on 29 September 2026, and GPUBank's live board. Cookie banners and chat bubbles are closed; nothing else is edited.
- 1Per-GPU hourly and spot prices for every GPU are public without an account, with a USD and EUR toggle.
- 2Reserved pricing is shown only as a discount table, from 2 % for 1 month to 25 % for 2 years.
- 3The H100 page headline says from $1.78/h, which is the spot price, not on demand.
- 1Live stock per region and every price, without an account.
- 2Per-hour and per-second prices, with vCPU, RAM, NVMe, IPv4 and traffic in them.
- 3A machine-readable price feed at /prices.json.
08The spec sheet
25 points, read off both sites.
GPUBank aheadVerda (ex-DataCrunch) aheadEven
| Criterion | GPUBank | Verda (ex-DataCrunch) |
|---|---|---|
| Price and catalogue | ||
| H100 SXM, per GPU-hour | $1.85, all-in | $3.56 on demand, $1.78 spot |
| Cheapest GPU-hourAn older 16 GB card. GPUBank's cheapest is the RTX 4090 at $0.23. | $0.23, RTX 4090 | $0.19, Tesla V100 16GBAhead |
| GPU models listed | 10 | 12Ahead |
| What the GPU price includesVerda block volumes, also used as OS drives, cost $0.20 per GiB-month. IPv4 and egress terms are not published. | vCPU, RAM, NVMe, IPv4, traffic | vCPU and RAM; disk billed apart |
| Billing | ||
| Billing granularity | Per second | Prepaid 10-minute increments |
| Minimum chargeThe unused part of the increment is refunded in the next billing period on early termination. | None | One 10-minute increment, unused part refunded |
| Stopped instance | Disk only | Billed while shut down, until deleted |
| Persistent storage | $0.03 per GB-month | $0.20 per GiB-month (NVMe block) |
| Egress traffic | Included | Not published |
| Public IPv4 | Included | Not published |
| Billing model | Prepaid balance | Prepaid credits |
| Payment and sign-up | ||
| Crypto payments | Bitcoin, USDT, Ethereum and 12+ more | Not documented |
| Card payments | No | Credit and debit cards, SEPA transferAhead |
| Minimum depositBank-invoice top-ups in USD or from non-SEPA banks require at least 1,000 EUR or USD. | $25 | Not published for cards |
| Platform | ||
| API and CLI | REST API, CLI, Python SDK | REST API, CLI, Python SDK |
| Terraform | Official provider | Official provider, OpenTofu too |
| Serverless GPUs | No | Serverless containers, scale to zeroAhead |
| Managed Kubernetes | No | Kubernetes or Slurm on Instant ClustersAhead |
| Ready-made templatesJupyterLab images and startup scripts are documented. | 25 | Not published |
| Interruptible or spotGPUBank interruptible instances are 40 % off. | Yes, 40 % off | Yes, about 50 % off on-demandAhead |
| GPUs per instance | Up to 8 | Up to 8, clusters of 16 to 128Ahead |
| Multi-node interconnect | Not offered | NVLink; InfiniBand on clustersAhead |
| Network and regions | ||
| Regions | 4 | 3 data centers, all in Helsinki |
| Countries | France, Germany, US, Singapore | Finland |
| EU regions | Paris, Frankfurt | Yes, Finland only |
| Trust and support | ||
| Availability targetThe site promises sensible SLAs and quotes historical uptime over 99.9 %. | 99.9 % monthly | No numeric SLA |
| Compliance reports | None published | SOC 2 Type II, ISO 27001, C5Ahead |
| Public status page | Yes | status.verda.com |
| Support | Tickets from a $25 balance | Chat, email, Discord, forumAhead |
09The right fit
Different needs. A clear choice.
What Verda (ex-DataCrunch) does better
- Self-service GB300, B300 and B200, from 1 to 8 GPUs
- Instant InfiniBand clusters with Kubernetes or Slurm pre-installed
- Serverless containers that scale to zero
- SOC 2 Type II, ISO 27001, 27017, 27018, 27701 and C5 listed
- Spot at about half the on-demand price, H100 SXM at $1.78
Choose GPUBank if
- H100 SXM at $1.85 per GPU-hour on demand, against $3.56
- Per-second billing instead of prepaid 10-minute increments
- Stop an instance and pay only for its disk, at $0.03 per GB-month
- Four regions in France, Germany, the US and Singapore
- Pay in Bitcoin, USDT and 12+ other coins, no card
Moving from Verda (ex-DataCrunch) takes an afternoon.
Verda block volumes can be cloned and attached to new instances, and its CLI syncs to S3-compatible storage with resumable transfers. There is no documented volume export format, so copy files, not disks.
- 1Launch the same GPU on GPUBank with a persistent disk in Paris or Frankfurt.
- 2Copy your data with rsync over SSH, or through S3-compatible storage with the Verda CLI.
- 3Run the same Docker image, then delete Verda instances and any volumes you no longer need.
# 1. a GPU and a disk on GPUBank gpubank launch --gpu h100-sxm --region eu-paris --disk data # 2. copy your data from the old machine rsync -azP old-host:/workspace/ root@<new-ip>:/workspace/ # 3. run the same image or template ssh root@<new-ip> 'cd /workspace && ./start.sh'
A few things you might be wondering.
Yes, on demand. Verda lists the H100 SXM at $3.56 per GPU-hour, and $1.78 on spot. GPUBank charges $1.85 per GPU-hour on demand, with vCPU, RAM, NVMe, IPv4 and traffic included, and 40 % less on interruptible instances.
Yes. Verda docs say a shut-down instance keeps being charged, and only deletion stops compute billing. On GPUBank a stopped instance costs only its persistent disk, at $0.03 per GB-month.
Pay-as-you-go instances are charged in prepaid 10-minute increments, and the unused part is refunded on early termination. Long-term contracts are paid up front, with discounts of 2 % for 1 month to 25 % for 2 years. GPUBank bills by the second from a prepaid balance.
No crypto option is documented. Verda lists credit or debit cards, SEPA bank transfer and bank-invoice top-ups. GPUBank is paid from a USD balance topped up in Bitcoin, USDT, Ethereum and more than a dozen other coins.
Verda. It rents GB300, B300 and B200 on demand and runs self-service Instant Clusters on InfiniBand with Kubernetes or Slurm, all in Helsinki. GPUBank offers the B200 SXM at $4.15 per GPU-hour, against $6.85 at Verda, up to 8 GPUs per instance and no clusters.
Without. Every price on this page excludes VAT and sales tax, for Verda (ex-DataCrunch) and for GPUBank, so the two columns compare like for like. Tax is added at checkout according to where the buyer is based.
Sources
Every figure about Verda (ex-DataCrunch) was read on its own pages or official documentation. Prices last checked on 29 September 2026.
- 01verda.com/pricingGPU instance, spot, serverless, instant cluster and storage prices, reserved discount table
- 02verda.com/gpu-instances3-decimal per-GPU prices, sizes, B300 CC and B200 CC prices
- 03verda.com/serverless-containersServerless billing intervals and spot at ~50%
- 04verda.com/instant-clustersCluster node ranges and per-GPU prices
- 05verda.com/h100H100 configurations and prices
- 06verda.com/gb300GB300 tray/rack configurations
- 07verda.com/Compliance claims, uptime claim, funding
- 08verda.com/companyFounding year, funding, revenue run rate
- 09verda.com/contactOffices, support channels, 99.95% uptime claim
- 10docs.verda.com/welcome-to-verda/pricing-and-billing/10-minute prepaid billing, top-up methods, bank invoice minimum, balance-zero deletion
- 11docs.verda.com/cpu-and-gpu-instances/set-up-a-gpu-instance/Billing increments, spot eviction, Jupyter images
- 12docs.verda.com/cpu-and-gpu-instances/shutdown-hibernate-and-delete/Stopped instance billing
- 13docs.verda.com/welcome-to-verda/locations-and-sustainability/FIN-01/02/03 data centers
- 14docs.verda.com/clusters/instant-clusters/Cluster sizes, InfiniBand, FIN-03
- 15docs.verda.com/containers/overview/Per-minute serverless billing, Python SDK
- 16docs.verda.com/infrastructure-as-code/terraform/Official Terraform provider
- 17docs.verda.com/cli/CLI features, object storage, 96-hour trash
- 18docs.verda.com/resources/services-overview/Bare metal, spot, contract credit limits
What we could not verify (8)
- Egress and public IPv4 pricing: not published on pricing or billing docs
- Numeric SLA and credit schedule: only 'sensible SLAs and fair compensation' stated
- Minimum card top-up amount: not published
- Spot prices for B300 CC and B200 CC: not shown
- Object storage price: product exists in CLI docs, no price published
- EUR prices not recorded (USD view captured)
- Reserved per-GPU prices are computed from the published discount table, not published as figures
- Instant cluster maximum size: docs say 128 GPUs, pricing page says up to 144 GPUs
In short
GPUBank vs Verda (ex-DataCrunch), in facts.
The essentials of this comparison in one table, for quick readers and AI assistants. Prices last checked on 29 September 2026.
Read as Markdown- H100 SXM, per GPU-hour
- Verda (ex-DataCrunch): $3.56 on demand, $1.78 spot
GPUBank: $1.85, all-in - Cheapest GPU-hour
- Verda (ex-DataCrunch): $0.19, Tesla V100 16GB
GPUBank: $0.23, RTX 4090 - Billing granularity
- Verda (ex-DataCrunch): Prepaid 10-minute increments
GPUBank: Per second - Stopped instance
- Verda (ex-DataCrunch): Billed while shut down, until deleted
GPUBank: Disk only - Persistent storage
- Verda (ex-DataCrunch): $0.20 per GiB-month (NVMe block)
GPUBank: $0.03 per GB-month - Egress traffic
- Verda (ex-DataCrunch): Not published
GPUBank: Included - Crypto payments
- Verda (ex-DataCrunch): Not documented
GPUBank: Bitcoin, USDT, Ethereum and 12+ more - Card payments
- Verda (ex-DataCrunch): Credit and debit cards, SEPA transfer
GPUBank: No - Minimum deposit
- Verda (ex-DataCrunch): Not published for cards
GPUBank: $25 - GPU models listed
- Verda (ex-DataCrunch): 12
GPUBank: 10 - Regions
- Verda (ex-DataCrunch): 3 data centers, all in Helsinki
GPUBank: 4 - Serverless GPUs
- Verda (ex-DataCrunch): Serverless containers, scale to zero
GPUBank: No - Availability target
- Verda (ex-DataCrunch): No numeric SLA
GPUBank: 99.9 % monthly - Compliance reports
- Verda (ex-DataCrunch): SOC 2 Type II, ISO 27001, C5
GPUBank: None published - Support
- Verda (ex-DataCrunch): Chat, email, Discord, forum
GPUBank: Tickets from a $25 balance - Headquarters
- Verda (ex-DataCrunch): Helsinki, Finland
GPUBank: Tallinn, Estonia
How this page was made
Every price and policy of Verda (ex-DataCrunch) was read on its own website, documentation or public API on 29 September 2026, with the address of each page listed above. A value we could not read on a primary source is marked "Not published" rather than estimated. GPUBank figures come from this site and its terms. Prices change: check both providers before you buy. Verda (ex-DataCrunch), its logo and its website belong to their owner and are shown only to identify the service compared. GPUBank is not affiliated with Verda (ex-DataCrunch).
