2026 Data Center Air Filtration

We make air filtration energy efficient.

A nanomaterial coating applied to existing HVAC filters captures particles through molecular attraction — not mechanical sieving. Same efficiency. Fraction of the energy.

99.9%
Capture efficiency
50–80%
Lower energy demand
~1/6
The pressure drop

01 / Problem

High efficiency and low energy have always been a tradeoff — until now.

High-efficiency air filters require denser media, which increases pressure drop and raises fan energy demand — the core physics tradeoff no conventional filter has solved.

HVAC accounts for ~35% of data center energy; air-handling fans are 7–10% of total facility load. As AI workloads scale, this is an acute and growing cost center.

No filter on the market today achieves high capture efficiency without the energy penalty. Data centers are forced to choose between air quality and operating cost.

Energy share — HVAC

35%of facility energy

Fan and air-handling systems alone represent 7–10% of total facility load — a direct operating cost target for any operator running at scale.

Source: Uptime Institute

02 / Technology

Capture by attraction, not by sieving.

A nanomaterial coating applied to existing HVAC filter substrates captures particles through molecular attraction forces (van der Waals) rather than mechanical sieving. The result: the same substrate achieves dramatically higher performance at a fraction of the pressure drop.

99.9%
Capture efficiency at 50–80% lower energy demand
≤100nm
Nanomaterial diameter — the physics threshold where van der Waals becomes dominant
Patent U.S. Patent 12,654,135 B2
01

Existing substrate — the nanomaterial coating is applied to standard filter media. No new filter architecture required; works with existing HVAC infrastructure.

02

Molecular attraction — particles adhere to the nanomaterial coating through van der Waals forces, not by being physically blocked. This eliminates the density-pressure tradeoff.

03

One-sixth the pressure drop — at equivalent capture efficiency to MERV-13. Fans run at lower load, translating directly to measurable energy savings on existing air-handling equipment.

04

Foundational IP — the ≤100nm diameter threshold is a physics-based claim, not an arbitrary specification. Claims are designed to be durable against prior art and design-around attempts.

03 / Market

Data center air filtration — the beachhead.

The global data center air filtration market is expanding at 7.2% CAGR, driven by AI workload growth and tightening energy mandates. We enter through the market segment with the highest energy sensitivity and the clearest ROI story.

$1.93B
Global TAM · 2024
$3.62B
Projected · 2033

7.2% CAGR · North America SAM ~$733M (2024)

Positioning

Higher-performance capture at lower energy cost than today's standard high-efficiency filters — a genuinely new performance point.

Pricing

20% premium over conventional filters — conservative. The market already supports a ~2× premium for V-Bank low-pressure-drop filters.

Energy driver

HVAC is 35% of facility energy; fan and air-handling is 7–10% of total facility load. Operators have direct financial incentive to reduce both.

Expansion

Liquid cooling, semiconductor ultrapure water, industrial HVAC, and produced water — one core technology, multiple markets beyond the beachhead.

04 / Benefits & ROI

$2.5M+ annual savings for a 50 MW data center.

300%+ Year 1 ROI · 2–3 month payback.

At a 20% price premium on filters — a conservative estimate — the energy savings dwarf the incremental cost within a single quarter.

Metric Typical (PUE 1.55) Hyperscale (PUE 1.10)
IT load 50 MW 50 MW
Total facility energy 678.9 GWh/yr 481.8 GWh/yr
Base-case energy savings $2.81M / yr $2.35M / yr
Savings range $1.78M – $4.07M $1.45M – $3.47M
Incremental filter investment $640K / yr $640K / yr
Simple payback ~2.7 months ~3.3 months
Year 1 ROI +339.6% +267.0%

Annual ROI of 300%+ and payback of 2–3 months for a 20% price premium on filters.

05 / Commercialization

Validation at scale.

The science is de-risked. We have coated sheet-size filters and validated the claims with a major filter manufacturer. The next step is scaling to full-size commercial filters.

Proven · Today
Sheet-size proof of concept

Coated ~8.5" × 11" sheet-size filters and validated the performance claims with a major filter manufacturer. The core science holds.

Next · At Scale
Full-size commercial filters

Partner with an equipment manufacturer to coat full-size filters and demonstrate the value proposition on real air-handling equipment used in data centers.

06 / Go-to-Market

Model

Cabomba Inside

Cabomba functions as the internal engine for filtration energy efficiency — embedded inside the products and systems our partners already operate. We supply the performance; our partners deliver it to the end customer. This model is designed to work across filtration formats and market verticals as the technology scales.

Target customers Hyperscale operators (Google, Microsoft, AWS, Meta) · Colocation providers · Data center REITs

08 / Team

Team & key details.

Suman Sinha Ray
Suman Sinha Ray, PhD
CEO · Co-founder
Kumar Natesaiyer
Kumar Natesaiyer, PhD
COO · Co-founder

09 / What's Next

Beyond data centers — one core technology.

The nanomaterial coating platform is not specific to air filtration. The same physics enables multiple high-value filtration markets.

Phase 1 · Now
Data center air

Beachhead market. Highest energy sensitivity, clearest ROI, fastest path to anchor customer.

Phase 2
Liquid cooling / CDU water

As AI density increases, liquid cooling filtration becomes a parallel expansion opportunity.

Beyond
Industrial HVAC

Broad HVAC market with significant energy footprint. Same coating, different substrate specifications.

Market
Semiconductor ultrapure water
Market
Produced water · oil & gas
Market
Pharma & beauty wastewater

Let's talk.

Whether you're an investor or a data center operator interested in a pilot — we'd like to hear from you.

emailtocabomba@gmail.com
Chicago, IL U.S. Patent 12,654,135 B2 IMPEL Accelerator