Free tools Windows power users keep installed
One-click scans. No signup required.
AI neoclouds make money by selling access to specialized computing capacity—especially GPUs—along with the networking, storage, software and support needed to run AI workloads. Their economics depend on turning costly infrastructure into billable service: long-term contracts can make demand and financing more predictable, but they do not guarantee that equipment will be delivered, kept busy or operated profitably.
What an AI neocloud sells
An AI neocloud is a cloud provider focused on specialized, high-performance computing for workloads such as AI training and inference. Customers are not simply renting a chip. They buy access to GPU systems and the surrounding infrastructure that makes those systems useful: high-speed networking, storage, orchestration software and technical support.
CoreWeave describes its platform as an integrated infrastructure and software stack for training, inference, model development and related workloads. Its disclosures provide a concrete example of one provider’s business model, not a template or financial average for the entire neocloud sector.
Where the revenue comes from
Committed capacity contracts
A customer can commit to using a specified amount of capacity over a contract term. Some agreements are take-or-pay: the customer agrees to pay for contracted capacity even if it does not use all of it. That can give the provider a clearer view of future revenue than relying entirely on customers buying compute hour by hour.
Recommended Free Tools
#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.
Committed contracts represented the following shares of CoreWeave’s revenue, according to its 2025 Form 10-K:
| Year | Share of CoreWeave revenue from committed contracts |
|---|---|
| 2023 | 88% |
| 2024 | 96% |
| 2025 | Over 98% |
These figures describe CoreWeave’s revenue mix in those years; they are not industry-wide percentages. The same filing reported that its committed contracts had a weighted-average duration of approximately five years as of December 31, 2025. Across active contracts at that date, customer prepayment averaged 15% to 25% of total contract value. Prepayments can help fund deployment, but they are not the same thing as profit: the provider still has to build and operate capacity and deliver the contracted service.
On-demand usage
Cloud providers can also sell capacity according to customer consumption. Usage-based service can suit workloads whose demand changes over time, but revenue may be less predictable than under a long-term commitment. CoreWeave warns in its 2025 Form 10-K that a shift away from take-or-pay contracts toward pay-as-you-go models could affect cash-flow predictability and margins. That is a company-specific risk disclosure, not proof that all providers use the same contract mix.
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
Why utilization matters
Utilization is the share of installed, available GPU capacity that is productively used and billed over time. GPUs, servers, facilities, power arrangements and financing can cost money whether a system is busy or idle. When more of the available capacity generates billable work, those largely fixed costs can be spread across more customer revenue. When capacity sits unused, costs continue while fewer GPU-hours earn revenue.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Utilization alone does not determine profitability. The result also depends on achieved prices, workload mix, power and hosting costs, financing, depreciation, networking, maintenance and whether the equipment is actually ready to serve customers. A GPU that has been ordered, a data center with power contracted, or a project listed in a backlog is not necessarily an installed GPU available for customer workloads.
The official CoreWeave and Core Scientific materials cited here do not provide a comparable provider-wide utilization rate measured as billable GPU-hours divided by available GPU-hours. CoreWeave’s reported active power and contracted power are infrastructure-capacity measures, not GPU utilization. Its backlog is also not a utilization measure.
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.
What contracts and backlog do—and do not—tell you
Long-term commitments can make expected demand more visible and can support financing tied to particular assets. But a contract is not cash already earned, proof that capacity is ready, or a guarantee of profit. The provider must secure equipment and facilities, bring power and systems online, and deliver service under the agreement.
CoreWeave reported a revenue backlog of $104 billion as of June 30, 2026, excluding more than $25 billion in net new customer commitments added in early Q3. The company said its backlog measure includes remaining performance obligations plus other amounts estimated to be recognized under committed contracts. It is subject to delivery and service-availability requirements; it should not be read as cash on hand or realized profit.
The infrastructure bill behind each GPU-hour
To sell usable capacity, an operator must obtain GPU servers, secure powered data-center space, and install the cooling, networking and storage needed for dense AI workloads. It also has to finance the build-out and maintain the systems. CoreWeave says it funds infrastructure primarily through asset-level debt supported by take-or-pay contracts, alongside corporate debt and equity. That links the contract model to capital spending: committed customer demand can help support financing, while the equipment and facilities create significant costs that must be recovered over time.
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.
For CoreWeave, active power reached 1.5 GW and total contracted power was approximately 3.7 GW as of June 30, 2026, according to its August 11, 2026 second-quarter release. The distinction matters: contracted power is not necessarily energized, customer-ready capacity, and neither figure reveals how intensively GPUs are being used.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How a hosting partner can earn revenue
A neocloud may rely on an outside data-center operator rather than own every facility it uses. In that arrangement, the host can earn fees for providing infrastructure while the neocloud sells computing services to customers. The two companies’ revenues come from different activities.
Core Scientific’s March 2, 2026 earnings presentation described a CoreWeave take-or-pay agreement covering approximately 590 MW of leased customer power across five sites. Core Scientific estimated more than $10 billion of potential revenue over the contracts’ terms and approximately $850 million in average annual revenue. Under the summarized arrangement, CoreWeave pays for capital expenditures, power and utilities; some construction costs are funded by Core Scientific and credited against hosting payments subject to specified limits. These are estimates and terms for that disclosed relationship, not a sector-wide hosting margin or a measure of CoreWeave’s cloud-service revenue.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesBest Value
What CoreWeave’s reported results show
Revenue growth and a large backlog do not by themselves establish profitability. CoreWeave’s results illustrate why it is important to distinguish sales, GAAP earnings and adjusted measures.
| Period | Revenue | GAAP operating result | GAAP net result | Adjusted EBITDA |
|---|---|---|---|---|
| Full year 2025 | $5.1 billion | Not stated in the cited 2025 Form 10-K summary | $1.2 billion net loss | Not stated in the cited 2025 Form 10-K summary |
| Q2 2026 | $2.575 billion | $49 million operating loss | $626 million net loss | $1.510 billion |
For Q2 2026, CoreWeave reported that revenue had risen from $1.212 billion in Q2 2025. The adjusted EBITDA figure is a non-GAAP measure; the company says such measures are supplemental and not substitutes for GAAP results. It should not be used to describe the quarter as simply profitable while omitting the operating and net losses. For 2025, the company reported infrastructure investment and depreciation and amortization among the factors behind rising costs.
The tension in the model is capital recovery: sales and commitments can grow quickly, but the provider must still absorb the cost of infrastructure, financing and delivery. In its 2025 Form 10-K, CoreWeave also says its business and pricing models have not been fully proven and it has limited operating history with its current models. Those disclosures are reasons to treat forecasts and contract totals carefully, not evidence that every neocloud has the same financial profile.
How to assess a neocloud’s business model
When comparing providers, look for disclosures that keep these questions distinct:
- Contract mix: What share of revenue comes from take-or-pay or other committed contracts versus on-demand use? What are the contract terms, prepayments, termination provisions and customer concentration?
- Capacity readiness: How much power is secured, how much is energized, and how much GPU capacity is installed and available to customers? These are different stages.
- Utilization and pricing: Does the provider disclose comparable billable GPU-hours and achieved prices? If not, do not infer utilization from revenue, backlog or power capacity.
- Capital and ownership: Who funds and owns the GPUs, facilities and power infrastructure? Consider debt, leases, customer advances and partner financing.
- Operating economics: Separate cost of revenue, depreciation, interest, operating cash flow and GAAP profit or loss from adjusted metrics.
- Service beyond hardware: Consider software, networking, storage, reliability, workload support and technical assistance as well as access to GPUs.
The cited company materials do not provide a like-for-like scorecard across multiple neoclouds, so they cannot establish a sector leader or support a reliable provider ranking.
Quick Recap
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.
