Neuralwatt
Neuralwatt provides an AI power optimization platform that measures and optimizes GPU energy consumption at the workload level for AI data centers and enterprises running inference, through hosted cloud services and on-premise software priced on actual GPU energy consumed (kWh).
- Company typePrivate
- Founded2024
- HeadquartersSeattle, United States
- Headcount1–10
- GTM typeB2B
- OfferingSoftware
What Neuralwatt does
Neuralwatt is an AI infrastructure software company that measures and optimizes the energy consumption of GPU workloads at the request level. Founded in 2024 in Seattle by Chad Gibson (CEO) and Scott Chamberlin (CTO) — veterans of power optimization at Microsoft and power-management chip design at Intel — the company targets AI data center operators and enterprises running large GPU fleets that are constrained by power envelope, cooling, or sustainability mandates rather than by silicon availability. Its product portfolio comprises Neuralwatt Cloud (hosted AI inference with an OpenAI-compatible API), Neuralwatt Optimize (continuous on-premise GPU power-tuning software with adaptive frequency scaling and optional Q-learning reinforcement learning), Neuralwatt Deploy (automated GPU server provisioning and fleet orchestration via a WebSocket control plane), and supplementary modules including Neuralwatt Allowance (per-API-key budget controls), Flex Tier (discounted time-flexible capacity), and per-request carbon emissions tracking built on a live integration with Electricity Maps.
The technical stack rests on direct GPU hardware telemetry via NVIDIA NVML/SMI, per-request energy attribution weighted by token-pool ratios and bounded by per-server caps, and real-time grid carbon intensity lookup. Validated outcomes from joint work with Crusoe Cloud on NVIDIA HGX H100 clusters include 33% higher inference throughput, 33% better server density (8 GPUs in a 6-GPU power envelope), and over 40% reduction in idle GPU power draw, while combined testing with ZutaCore on B200 hardware demonstrated 18-19% lower server power consumption and 3x improvement in power stability.
Neuralwatt monetizes through a mix of usage-based and subscription revenue. The default pricing model is energy-based at $5/kWh on Neuralwatt Cloud (with token-based pricing retained for migration compatibility), tiered subscriptions (Basic, Standard, Pro, Enterprise) that price approximately 40% below PAYG and bundle energy allowances and concurrent slots, an enterprise tier with custom multi-year agreements for on-premise Optimize/Deploy, and a $1 free trial that seeds PLG adoption. Go-to-market combines self-serve developer adoption (portal.neuralwatt.com, OpenAI-compatible API, open-source integrations with Claude Code, OpenCode, simonw's llm CLI, and a Discord community) with enterprise field sales targeting 100MW+ GPU operators via dedicated sales and partnerships channels. Strategic investors include Microsoft, NVIDIA, Remarkable Ventures, Avesta Fund, and Climate Collective, alongside commercial partners Crusoe Cloud, Parasail, GreenPT, ZutaCore, Hugging Face, and Electricity Maps.
Neuralwatt firmographics
Firmographics- Name
- Neuralwatt
- Legal name
- Neuralwatt Inc.
- Website
- https://neuralwatt.com
- Company type
- Private
- Founded year
- 2024
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- Neuralwatt provides an AI power optimization platform that measures and optimizes GPU energy consumption at the workload level for AI data centers and enterprises running inference, through hosted cloud services and on-premise software priced on actual GPU energy consumed (kWh).
- Ownership category
- akta.pro rank
Neuralwatt industry classification
Industry- Product category
- AI Infrastructure Power Optimization
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Custom Computer Programming Services (541511)
- SIC
- Services-Prepackaged Software (7372), Services-Computer Programming, Data Processing, Etc. (7370)
- akta.pro primary industry
- AI Datacenter Power, Cooling & Racks (liquid cooling, power delivery) (HDAAAAAH)
- akta.pro secondary industries
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK), Model Hosting, Serving & Inference Platforms (HDAAACAB), AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)
Keywords
Where Neuralwatt is headquartered
LocationHeadquarters
- HQ city
- Seattle
- HQ country
- United States
- HQ region
- North America
Offices3 records
Markets served
Neuralwatt business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Infrastructure, Personnel, Marketing or Sales, Operations
Revenue model
- Hosted AI Inference (Neuralwatt Cloud): Self-hosted AI cloud providing power-efficient inference with energy-based pricing. Revenue from usage-based billing on GPU energy consumption (kWh) and optional token-based billing for compatibility.
- Neuralwatt Optimize (On-Premise): Continuous GPU power optimization software for organizations running their own GPU infrastructure. Generates revenue from software licensing and professional services for enterprise deployments.
- Neuralwatt Deploy: Automated GPU server provisioning solution. Revenue from deployment services and ongoing software subscriptions for fleet orchestration capabilities.
- Subscription Plans: Tiered subscription plans (Basic, Standard, Pro, Enterprise) with monthly or annual billing. Includes energy allowances, concurrent slots, and rate limit tiers.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | Pay-As-You-Go (PAYG) - Energy-based pricing at $5/kWh |
| Subscription | Monthly | Subscription plans with included energy allowances |
| Freemium | Pay-as-you-go | Trial - Free access with $1 starting credit |
| Usage-based | Pay-as-you-go | Flex Tier - Discounted pricing for time-flexible workloads |
| Subscription | Multi-year contract | Enterprise - Custom agreements with guaranteed capacity |
Go-to-market motion3 records
Distribution channels4 records
Marketing channels7 records
Neuralwatt product offering
Product offeringCore offering
Neuralwatt sells an AI power optimization platform composed of Neuralwatt Cloud (hosted AI inference priced by measured GPU energy at $5/kWh), Neuralwatt Optimize (continuous GPU power optimization software that adapts frequency, batches workloads, and tunes GPU power in real time), and Neuralwatt Deploy (on-premise GPU server provisioning and fleet orchestration for customer data centers). The offering targets organizations operating large-scale GPU infrastructure for AI inference, providing GPU-level energy telemetry, per-request carbon emissions tracking integrated with Electricity Maps, and adaptive power tuning validated to deliver 33% more compute from existing power envelopes on NVIDIA datacenter GPUs.
Product overview
Neuralwatt offers an AI power optimization platform consisting of Neuralwatt Cloud (hosted inference with energy-based pricing), Neuralwatt Deploy (on-premise software for customer's GPU infrastructure), and Neuralwatt Optimize (continuous GPU power optimization). The Core Platform adds fleet orchestration capabilities via agents and orchestrator. Key platform features include Neuralwatt Allowance (per-key budget controls), Flex Tier (discounted flexible capacity tier), and Carbon Emissions Tracking (real-time per-request carbon accounting). Energy-based pricing is the default model, with token pricing available for migration compatibility.
Differentiator
Problem solved
Functional benefit
Brands
- Neuralwatt Cloud: Hosted AI inference platform with energy-based pricing, OpenAI-compatible API, and real-time energy metrics.
- Neuralwatt Deploy
- Neuralwatt Optimize
Products and services
- Neuralwatt Cloud Hosted AI inference service providing self-serve access to frontier AI models (text and code) through an OpenAI-compatible REST API, with industry-first energy-based pricing at $5/kWh and per-request energy and carbon metrics. Targets teams running AI workloads who need transparent, usage-aligned inference billing and is available via a self-serve portal with a $1 free trial credit.
- Neuralwatt Optimize Continuous GPU power optimization software for existing GPU infrastructure that performs adaptive frequency scaling, dynamic load management, and grid-aware demand response, with optional Q-learning reinforcement learning. Compatible with NVIDIA Ada, Hopper, and Blackwell datacenter GPUs and designed for organizations operating their own GPU fleets who need to reclaim stranded power.
- Neuralwatt Deploy On-premise deployment product that brings energy-optimized inference to customer-owned GPU infrastructure, providing automated GPU server provisioning, zero-downtime operations, and a full observability stack deployable via Docker or Kubernetes. Targets enterprises that cannot or choose not to use hosted inference and need private, controlled deployment.
Quantifiable outcome
- 33% more compute from existing power envelopes without new hardware
- +6 more outcomes
Companies that use Neuralwatt
Customer profileNamed customers3 records
Segments4 records
Ideal customer profiles3 records
Neuralwatt technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration4 records
AI capability5 records
Feature11 records
Neuralwatt partnerships and signals
Strategic signalPartnerships
Seven partnerships are on record, tiered core and moderate.
- ParasailcoreStrategic partnership where Neuralwatt's energy intelligence runs in Parasail's GPU fleet, bringing real-time thermal awareness and smarter workload placement. Parasail's compute helps Neuralwatt iterate faster. Both companies share conviction that gains come from optimizing existing infrastructure.
- ZutaCorecoreJoint testing partnership with ZutaCore (waterless direct-to-chip liquid cooling) proving AI data centers can operate at 45C warm-water temperatures on NVIDIA B200 hardware. Nearly 500 paired experiments showed zero thermal throttling, no performance penalty, and triple thermal buffer with combined solution.
- GreenPTcoreStrategic partnership to set a new global standard for sustainable AI inference. GreenPT brings renewable infrastructure and carbon measurement expertise; Neuralwatt provides GPU-level energy optimization and energy-based pricing. Joint commitment to bring more sustainable AI products to market.
- Hugging FacemoderateNeuralwatt (via Scott Chamberlin) contributed to AI Energy Score v2 refresh and expansion to support reasoning-model energy testing. Helped streamline benchmarking workflow and expand evaluation suite. Supported creation of AI Energy Benchmarks open-source package.
- Crusoe CloudcoreCrusoe Cloud provided NVIDIA HGX H100 GPU clusters in Iceland (powered by 100% clean geothermal and hydro energy) for Neuralwatt to validate power optimization software. Results: 33% increase in throughput, 33% better density, 40%+ idle power reduction.
- Microsoft for StartupsmoderateMicrosoft for Startups is listed among Neuralwatt's backers. This program likely provides go-to-market support, credits, and resources for AI startups.
- Electricity MapscoreElectricity Maps provides real-time carbon intensity data for Neuralwatt's carbon emissions tracking. Integration delivers live data for grids worldwide, including direct measurements from grid operators and power production data by generation type.
Scale indicators7 records
Recent moves6 records
Expansion highlights6 records
Neuralwatt competitors and assessment
Company assessmentDirect peers
- GreenPT: Sustainable AI inference provider partnering with Neuralwatt to combine renewable infrastructure and GPU-level energy optimization. Direct peer with shared sustainability positioning and AI inference focus.
- Lambda Labs: GPU cloud and on-premise cluster provider serving AI/ML workloads with NVIDIA hardware. Comparable GPU-as-a-service offering and developer-focused go-to-market.
- Fireworks AI: AI inference cloud platform offering hosted frontier and fine-tuned models with OpenAI-compatible APIs and enterprise deployment. Comparable hosted inference model targeting production AI workloads.
- Modal Labs: Developer-focused serverless GPU compute platform for AI workloads with API-driven inference. Comparable cloud-hosted inference and per-request compute model targeting developers and AI startups.
- RunPod: GPU cloud provider offering on-demand and serverless AI inference with per-second billing. Comparable hosted inference platform targeting individual developers and small teams.
- Together AI: AI cloud platform offering hosted inference for open and frontier models with an OpenAI-compatible API. Closely comparable to Neuralwatt Cloud's hosted inference product and developer adoption model.
- CoreWeave: Large-scale GPU cloud provider offering NVIDIA-accelerated compute for AI training and inference. Comparable business model of hosted GPU-as-a-service targeting enterprise AI workloads.
- Nebius: AI-focused cloud infrastructure provider offering NVIDIA GPU compute for training and inference. Comparable full-stack GPU cloud targeting AI enterprises and researchers.
- Voltage Park: GPU cloud provider operating large-scale NVIDIA H100 fleets for AI training and inference. Comparable infrastructure-as-a-service model with NVIDIA-accelerated compute focus.
- Crusoe Cloud: AI-focused GPU cloud provider operating NVIDIA HGX H100 clusters powered by renewable energy in Iceland. Direct customer and validation partner of Neuralwatt; offers overlapping hosted inference and sustainability positioning.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
Neuralwatt social profiles
Digital presenceNeuralwatt financial estimates
Financial estimateRevenue estimate
Valuation estimate
Neuralwatt leadership team
Management profileNumber of profiles
Profiles4 records
Neuralwatt funding detail
Funding detailFunding overview
Funding rounds1 record
Investors3 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Neuralwatt M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Neuralwatt
What does Neuralwatt do?
Neuralwatt sells an AI power optimization platform composed of Neuralwatt Cloud (hosted AI inference priced by measured GPU energy at $5/kWh), Neuralwatt Optimize (continuous GPU power optimization software that adapts frequency, batches workloads, and tunes GPU power in real time), and Neuralwatt Deploy (on-premise GPU server provisioning and fleet orchestration for customer data centers). The offering targets organizations operating large-scale GPU infrastructure for AI inference, providing GPU-level energy telemetry, per-request carbon emissions tracking integrated with Electricity Maps, and adaptive power tuning validated to deliver 33% more compute from existing power envelopes on NVIDIA datacenter GPUs.
Is Neuralwatt a public or private company?
Neuralwatt is a private company. It is classified as venture growth investor backed and is currently operating.
When was Neuralwatt founded?
Neuralwatt was founded in 2024. It employs 1 to 10 people.
Where is Neuralwatt based?
Neuralwatt is headquartered in Seattle, United States, in the North America region.
How does Neuralwatt make money?
Four revenue lines are on record. Hosted AI Inference (Neuralwatt Cloud) is the primary driver. The others are neuralwatt Optimize (On-Premise), neuralwatt Deploy and subscription Plans.
Who are Neuralwatt's main competitors?
Direct peers on record are GreenPT, Lambda Labs, Fireworks AI, Modal Labs, RunPod, Together AI, CoreWeave, Nebius, Voltage Park and Crusoe Cloud.
Does Neuralwatt have an API?
Yes. Neuralwatt provides an OpenAI-compatible REST API at https://api.neuralwatt.com/v1 for AI inference. Supports chat completions, models listing, usage/energy metrics, quota, and billing endpoints. Authentication via API keys (Bearer token). Includes real-time energy metrics on every request. Also offers a WebSocket control plane for agent orchestration and a preview Core Platform API for GPU fleet management. Developer documentation is at portal.neuralwatt.com/docs.
What industry is Neuralwatt in?
Neuralwatt's product category is AI Infrastructure Power Optimization. Its primary akta.pro industry code is HDAAAAAH, AI Datacenter Power, Cooling & Racks (liquid cooling, power delivery), with a secondary code of HDAAAAAK, AI Compute Cloud & GPU-as-a-Service. Its NAICS code is 5182 and its SIC code is 7372.