Skip to content
Log in Start now
Netslim
Log in Create account

Enterprise infrastructure

Your platform.
Your hardware. Our management.

A microservices platform, your own AI cluster or your complete technical infrastructure: designed, built and managed 24/7 by Netslim, on your own hardware in ISO 27001 data centers in the EU. From € 100,000.00 one-time and € 10,000.00 per month.

View the projectsTalk first
  • You own your environment, no lock-in
  • Kick-off within two working days
  • Dedicated engineer, 24/7 on-call service
Three projects

Choose the scale. The design determines the exact price.

All amounts are excl. VAT and "from". You order the project with a down payment of 30% of the setup. This starts the design phase. We only order hardware once you approve the design with the exact price. The monthly amount starts at delivery.

01Platform and microservices SaaS, platforms, custom webshops, organizations with their own software team 6 to 12 servers € 100,000.00 € 10,000.00 Details

Your own Kubernetes platform for your applications: microservices, APIs, data and CI/CD. For SaaS companies, webshops with their own software and organizations that want to move away from one large application.

What's included

  • clusterKubernetes on your own dedicated hardware, 3 control plane + 3 or more worker nodes, in two data centers
  • servicesMicroservices architecture: API gateway, service mesh, message bus (Kafka or NATS), caching
  • dataPostgreSQL with replication and point-in-time recovery, object storage (S3-compatible)
  • ci/cdGitLab or GitHub Actions, container registry, environments for test, acceptance and production
  • inzichtMetrics, logs and traces (Prometheus, Loki, Tempo, Grafana), alerting to your team
  • veiligSecrets management, network policy, WAF and DDoS protection, pentest on delivery
  • sla99.9% availability, response within 1 hours on an outage, 24/7 incident support
Due now: deposit for design and build phase € 30,000.00 30% of the setup from € 100,000.00 · excl. VAT · by bank transfer or at checkout Remainder of the setup in instalments after approval of the design · then from € 10,000.00 per month after delivery
Would you rather talk to an engineer first?
02AI compute and inference AI applications with many users, companies that do not want their data in a US cloud 16 or more GPUs € 250,000.00 € 25,000.00 Details

Your own GPU cluster with NVIDIA RTX PRO 6000 Blackwell Server Edition: language models, image and speech run on your own hardware in the EU, with the software around it.

What's included

  • gpu2 or more systems, each with 8x NVIDIA RTX PRO 6000 Blackwell Server Edition (96 GB per card)
  • netwerkNVIDIA ConnectX network cards, 400 Gbit/s between the systems, 10 Gbit/s uplink
  • opslagNVMe storage for models and datasets, object storage for archive
  • softwareKubernetes with GPU operator, vLLM or TensorRT-LLM, your own API with rate limiting and keys
  • modellenOpen-weight models (Llama, Mistral, Qwen, GLM, Whisper) set up and tested
  • inzichtGPU utilization, tokens per second, cost per request, in one dashboard
  • sla99.9% availability, spare GPUs, 24/7 incident support
Due now: deposit for design and build phase € 75,000.00 30% of the setup from € 250,000.00 · excl. VAT · by bank transfer or at checkout Remainder of the setup in instalments after approval of the design · then from € 25,000.00 per month after delivery
Would you rather talk to an engineer first?
03Complete technical infrastructure Scale-ups, healthcare and education, industry, organizations with 100 to 5,000 employees two data centers € 500,000.00 € 40,000.00 Details

Everything: your own racks in two data centers, network, storage, compute, platform and management. For organizations that outsource their entire IT foundation and still want to stay the owner.

What's included

  • racksOwn racks in ISO 27001 data centers in Germany and Finland, 24/7 administrator on site
  • netwerkRedundant uplinks, own IP space and BGP, firewall cluster, DDoS protection, VPN and zero-trust access
  • computeVirtualization (Proxmox or VMware) and Kubernetes, GPU systems as needed
  • opslagStorage cluster (Ceph) with snapshots, backups at a second location, kept for 30 days
  • herstelFailover to the second data center, recovery time under 1 hours, tested every year
  • werkplekIdentity (SSO), email and office via Google Workspace or Microsoft 365, device management
  • sla99.99% availability, 24/7 management, dedicated engineer, monthly review
Due now: deposit for design and build phase € 150,000.00 30% of the setup from € 500,000.00 · excl. VAT · by bank transfer or at checkout Remainder of the setup in instalments after approval of the design · then from € 40,000.00 per month after delivery
Would you rather talk to an engineer first?
AI compute

NVIDIA RTX PRO 6000 Blackwell Server Edition.
Up to eight per system.

The AI project builds on systems with eight RTX PRO 6000 Blackwell Server Edition cards (96 GB GDDR7 ECC per card), NVIDIA ConnectX networking and NVMe storage. Open-weight models on your own hardware in the EU: your data does not leave your environment.

  • 96 GBGDDR7 ECC per card
  • 5th genTensor cores · FP4
  • 4th genRT cores
  • 1 to 8cards per system
View AI project Enterprise: own cluster
How it works

From deposit to delivery.

  1. 01Kick-offWithin two working days of the down payment, an engineer calls you: goals, current situation, requirements and planning.2 working days
  2. 02DesignArchitecture, hardware list, security, migration plan and the exact price. You approve it. Only then do we order.2 to 4 weeks
  3. 03Build and migrationHardware in the rack, platform on top, your applications moved over, testing together. Your old environment keeps running until everything checks out.6 to 12 weeks
  4. 04Delivery and managementDocumentation, handover to your team, 24/7 management with a dedicated engineer. The monthly fee starts from here.ongoing
Server with eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs
System builders

For each project, we choose the builder that fits.

For enterprise projects, we buy the systems from the manufacturers that NVIDIA certifies for RTX PRO servers, and from Dell, HPE, Lenovo, Supermicro and Cisco for compute, storage and network. You own the hardware. We manage it in our rack.

System builders and hardware we choose from

  • NVIDIA
  • AMD
  • Intel
  • Dell Technologies
  • Supermicro
  • HPE
  • Lenovo
  • Cisco
  • ASUS
  • GIGABYTE
  • MSI
  • ASRock Rack
  • Advantech
  • Eviden
  • Wiwynn
  • Wistron
  • Inventec
  • Foxconn
  • Compal
  • Pegatron

Frequently asked questions

Why a down payment and not a quote?

A quote without a design is a gamble. The down payment starts a real design phase: an engineer looks at your environment and delivers a plan with the exact price. If you don't approve it, it stops there and you keep the design. The down payment covers that work.

Who owns the hardware?

You do. We buy, build and manage. The servers sit in our rack in your name. If you ever want to leave, you take them with you or we sell them for you.

What is included in the monthly fee?

Rack space, power, network, 24/7 management and fault service, monitoring, backups, updates, a dedicated engineer and monthly review. Not included: new hardware for expansion. That goes into the design or a separate order.

Can you demonstrate ISO 27001 or NEN 7510?

The data centers are ISO/IEC 27001 certified. For healthcare and government, we supply the documentation for NEN 7510 and the GDPR. A pentest is part of every delivery.

How fast will it be ready?

Expect three to five months from down payment to delivery, depending on hardware delivery times (GPU systems sometimes take longer) and on your own migration.

Can it be smaller?

Yes. For one or a few machines, our dedicated and GPU servers are there: per month, without a design phase. Enterprise is for when you need a platform, a cluster or your whole infrastructure.

Talk it over first?

An engineer calls you back and thinks along with you, before you order anything.

Plan a call Contact us