What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Contabo GPU VPS is a single-card, full-GPU virtual server built on an NVIDIA RTX PRO 6000 Blackwell Server Edition with 96 GB of VRAM, priced from a headline €999, $1,199, or £969 per month excluding VAT. It suits a buyer who needs one large-memory CUDA GPU for inference, fine-tuning, rendering, or simulation on a flat monthly bill, and who is comfortable administering an unmanaged Ubuntu server. It does not suit anyone who needs multi-GPU scaling, multi-node training, or proof of performance from independent benchmarks, because none has been published for this product yet.
This review is based on Contabo’s own product page and launch announcement, dated September 21, 2026, plus a small number of third-party and user sources. We did not run the server, benchmark it, or test support. Statements below are labeled as coming from Contabo, from independent context, or from individual users.
As an Amazon Associate I earn from qualifying purchases.
What Contabo is selling
Contabo describes GPU VPS as a Performance VPS paired with one dedicated NVIDIA GPU, attached through PCI passthrough. The advertised configuration, taken from the Contabo GPU VPS product page and the launch announcement, is:
| Component | Advertised value (Contabo) |
|---|---|
| GPU | NVIDIA RTX PRO 6000 Blackwell Server Edition, one card |
| GPU memory | 96 GB GDDR7 ECC |
| CUDA cores | 24,064 |
| FP32 compute | 120 TFLOPS |
| GPU memory bandwidth | 1,597 GB/s |
| vCPUs | 18 |
| System RAM | 96 GB |
| Storage | 900 GB NVMe |
| Preinstalled image | Ubuntu 24.04 LTS, with CUDA and nvidia-container-toolkit |
| Regions at launch | Europe and US Central |
These are vendor specifications. They describe what the product is supposed to provide; we have not measured them. Confirm them on the live product page at checkout, because configuration can change after launch.
#1 Best Overall
- [ Maximum AI Compute Power ] Dominate complex workloads with the ASUS ESC8000A-E13. This 4U rack server is a powerhouse engineered for mass-scale AI, machine learning, and deep training. Featuring support for dual AMD EPYC 9005/9004 processors and up to eight dual-slot GPUs, it delivers the raw computational muscle required to train LLMs and run complex simulations effortlessly. Accelerate your data science pipeline and transform raw data into actionable intelligence faster than ever.
- [ Advanced Thermal Efficiency ] High performance demands elite cooling. The ESC8000A-E13 features a cutting-edge aerodynamic design with independent CPU and GPU airflow tunnels. Equipped with redundant hot-swap fans and optimized for liquid cooling integrations, this 4U server ensures maximum uptime under heavy, sustained workloads. Keep your data center running cool, quiet, and highly efficient while preventing thermal throttling during mission-critical enterprise operations.
- [ Scale with Flexible Storage ] Future-proof your infrastructure with unmatched storage and expansion flexibility. This offers comprehensive front-panel drive bays supporting Gen5 NVMe, SAS, or SATA drives alongside multiple PCIe 5.0 slots. Designed as a high-density 4U server capable of housing eight dual-slot GPUs: NVD H200, RTX PRO 6000 Blackwell, RTX PRO 4500 Blackwell or AMD Instinct MI350P PCIe Card, each supporting up to 600 watts.
- [ Enterprise-Grade Reliability ] Minimize downtime and secure your ecosystem with server-grade redundancy. The ESC8000A-E13 is built for 24/7 continuous operation, boasting 2+2 redundant (3200W total) 80 PLUS Titanium power supplies and integrated ASUS ASMB11-iKVM for comprehensive out-of-band management. Ideal for cloud service providers, rendering farms, and large enterprise infrastructure, it combines robust physical hardware with smart remote monitoring to safeguard your digital assets.
- [Reliability Guaranteed] Shop with total peace of mind knowing that every new computer component we sell is backed by our EPC 3-year warranty. Whether you are investing in high-speed DDR5 RAM or a powerhouse GPU, we protect your build against defects and performance failures. We stand firmly behind the quality of our hardware, ensuring that your setup remains fast, stable, and secure for years to come.
Pricing and what the discounts mean
The launch announcement lists monthly headline prices excluding VAT and stated discounts for committed terms. The table below calculates those discounts from the headline figures. The calculated values are our arithmetic, not separate Contabo quotes, and the checkout total is what counts.
| Currency | Headline per month (excl. VAT) | 12-month term, 15% stated discount | 24-month term, 20% stated discount |
|---|---|---|---|
| EUR | €999 | €849.15 | €799.20 |
| USD | $1,199 | $1,019.15 | $959.20 |
| GBP | £969 | £823.65 | £775.20 |
Three pricing details matter before you compare this with other providers:
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →- VAT is extra for many buyers. The headline is excluding VAT, so the amount you pay depends on your country and tax status. We have not calculated VAT for any country.
- Commitment changes the monthly cost. The 12-month and 24-month discounts lower the monthly rate but lock you into a longer term. A shorter test run at the headline rate costs more per month.
- Contabo’s comparison is its own. The product page compares this GPU VPS with hourly-equivalent configurations from Akamai, AWS, and Google Cloud. Contabo says prices are calculated for a 31-day month and were collected on September 15, 2026. That is a vendor-selected pricing claim, not a controlled test of cost per unit of performance, and it does not show that the configurations are equivalent in GPU model, region, storage, or egress terms.
Is the GPU dedicated or shared?
Contabo says the full GPU is assigned to your VPS rather than divided into vGPU slices. In practice, that means the card’s 96 GB of memory is available to your workload and not partitioned with other tenants. Because the GPU is attached by PCI passthrough, the hypervisor does not mediate it in the way a sliced vGPU would. The benefit is that one workload gets the whole card; the cost is that you get one card, not a pool.
Rank #2
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
After provisioning, check that the card and its memory are visible from inside the server:
nvidia-smi --query-gpu=name,memory.total --format=csvYou should see the RTX PRO 6000 Blackwell card with roughly 96 GB of total memory. If the command fails, the driver is not loaded, and you should open a support ticket before loading any workload.
Can it run a 70B model?
It depends on the precision and the context length you need, and the answer is arithmetic rather than a tested result. A 70-billion-parameter model stored at 16-bit precision needs about 140 GB for weights alone, which exceeds 96 GB, so it will not fit unquantized. Quantized to 4-bit, the weights take roughly 35 to 40 GB, which leaves room for the KV cache and a reasonable context window. Real headroom depends on the inference engine, batch size, and model architecture. We have not run any 70B model on this server, and you should load your own model before planning production capacity around it.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #3
- AI-Optimized: Designed to support up to 4 GPUs, it is perfect for handling intensive AI and machine learning tasks, ensuring high performance and scalability for advanced computational needs.
- Intelligent Storage: Equipped with 8 hot-swappable 3.5" SATA/SAS drives (12Gbps), featuring SGPIO and temperature control, it ensures efficient data management and reliable storage performance.
- Robust Cooling: The system includes 3x 12038 hot-swap PWM fans and 2x 8038 rear fans, providing advanced thermal management to maintain optimal temperatures and ensure stable operation under heavy workloads.
- Rack-Ready: Comes with a pre-installed rail kit, allowing for quick and easy installation in standard 19-inch server racks, making it ideal for data center environments and enterprise setups.
- Versatile Connectivity: Offers USB 3.0 and the latest USB 3.2 Type-C ports, ensuring high-speed data transfer and compatibility with a wide range of peripherals and devices for enhanced connectivity options.
Which workloads Contabo targets, and what it excludes
Contabo lists AI inference and fine-tuning, image and video generation, single-node scientific simulation, batch rendering, and video encoding as intended uses. Its announcement says the product is not intended for multi-GPU or multi-node training clusters. If your workload needs several GPUs communicating at high speed, this product is outside Contabo’s own scope.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What independent evidence exists
We found no independent benchmark of this GPU VPS in the sources we could check as of October 9, 2026. Two items are often mistaken for evidence and are not:
- A Contabo blog post reports that Hosttest gave a favorable review to Cloud VPS 10 in March 2026. That review covered a different, non-GPU product, and the post is published by Contabo itself.
- General VPS reviews, such as the VPS Verdict Contabo review, test standard VPS plans. They describe performance caveats for Contabo’s CPU-based products, but they do not measure the GPU instance, and they should not be read as a proxy for GPU throughput.
Until an independent test of the GPU configuration appears, performance claims rest on Contabo’s specifications. Treat them as the advertised ceiling, not a measured result.
Rank #4
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
What users report
Recent user discussion is mixed, and it is anecdotal. One Reddit user reports that a production server on Contabo was unavailable for more than three days and criticizes both the status communication and the support responses, in a thread about that outage. A broader Reddit discussion of Contabo reviews contains contrasting accounts, including CPU variability and connectivity complaints alongside reports of trouble-free use lasting years.
These accounts concern Contabo VPS in general, not the GPU product. They are not a representative sample and do not measure incident rates or uptime. The GPU VPS is also recent enough that little user history exists for it. Use these accounts to decide which risks to plan for, not to estimate how often they happen.
Quick Recap
Who should buy and who should not
Contabo GPU VPS fits a buyer who:
- needs one card with a large amount of memory for a single CUDA workload;
- wants flat monthly billing and accepts a term commitment for the lower rate;
- can use a Europe or US Central location;
- is comfortable managing Ubuntu 24.04 LTS, the CUDA stack, and backups without managed services.
It is a weaker fit if you need:
- more than one GPU, or multi-node training;
- a region other than Europe or US Central;
- a non-Ubuntu system image;
- independently benchmarked performance before you commit;
- managed operations or a service-level guarantee that covers the GPU product.
Before you buy
- Open the live product page and confirm the GPU, memory, image, and region you need are offered and in stock.
- Calculate your total with VAT for your country and the term you plan to commit to, not only the headline price.
- Check the Contabo product documentation for the GPU VPS specifications, image details, add-ons, and limitations.
- Ask Contabo support whether the GPU VPS is covered by the support tier you are buying, and what response times apply.
- Confirm the outbound traffic policy for your workload. Contabo describes unlimited traffic subject to a fair-use policy, so unlimited does not mean every traffic pattern is unrestricted.
- Run a short test with your own model or job before committing to a 12- or 24-month term, and keep independent backups of anything you cannot rebuild.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
