HOME
Act Within Days, Not Quarters
A rare convergence of policy, pricing, and AI infrastructure scarcity makes immediate GB300 deployment materially more attractive than waiting. CNEX has access to NVIDIA GB300 AI systems at $5M per system, with pricing locked only until 04/15/2026. 100% bonus depreciation under IRC §168(k), Massachusetts data center tax incentives, and high expected annual EBITDA create a short-duration window for lenders and investors to secure an unusually strong infrastructure return profile.
$5.0M
Entry Price Per System
$10.0M
Annual Revenue Potential
50%
EBITDA Margin Per System
~$1.81M
Year-1 Value Capture
Executive Summary
The Case for Immediate Action
100% bonus depreciation under IRC §168(k) allows CNEX and its capital partners to deduct the full cost of qualifying AI infrastructure — such as a $5M GB300 system — in Year 1 rather than over multiple years. In parallel, Massachusetts offers a 20-year sales-and-use-tax exemption for qualifying data centers on equipment, software, electricity, and certain construction costs, plus meaningful energy-efficiency incentives. Together, these policies front-load cash, compress payback, and materially de-risk AI infrastructure for both lenders and equity investors.
CNEX now has an unusually tight tactical window: a $5M per-GB300 entry price is available only until 04/15/2026, while credible expectations point to 10–30% hardware inflation thereafter. With approximately $10M annual revenue potential and approximately 50% EBITDA margin per system, delaying deployment means higher capex, lost high-margin EBITDA, and lower present-value tax shields. This is not a routine infrastructure timing decision — it is a superior return vintage available only for a narrow window.
Key Drivers
  • 100% IRC §168(k) bonus depreciation in Year 1
  • 20-year MA sales and use tax exemption
  • $5M pricing locked until 04/15/2026
  • $5M annual EBITDA per system
  • $1.56M–$1.81M Year-1 value capture
  • 10–30% hardware inflation expected post-deadline
"This is not a normal infrastructure timing decision. This is a superior return vintage available only for a narrow window."
Core Investment Snapshot
One GB300. One Exceptional Infrastructure Unit.
Each NVIDIA GB300 AI system behaves like a high-yield digital infrastructure asset with unusually strong Year-1 economics. The combination of revenue density, operating leverage, and front-loaded tax value creates a return profile that is materially superior to conventional infrastructure alternatives.
Asset
NVIDIA GB300 AI System — next-generation compute infrastructure purpose-built for AI factory workloads
$5.0M Entry Price
Price locked until 04/15/2026 — subject to 10–30% inflation thereafter
$10-$12M Revenue
Annual revenue potential per deployed GB300 system under current market assumptions
55% EBITDA Margin
~$5.0M annual EBITDA per system — high operating leverage with defensible margin structure
$1.25M–$1.50M
Federal tax shield via IRC §168(k) full Year-1 bonus depreciation on qualifying AI infrastructure
~$312K MA Savings
Avoided Massachusetts sales and use tax under the state's 20-year data center incentive program
$1.81M
Year-1 Value Capture
Combined federal and state incentive benefit in Year 1 alone
36%
Effective Capex Return
Policy returns roughly 30–35% of capex in Year-1 value before energy incentives
<1yr
Payback Horizon
Payback compresses toward sub-1-year under incentive-adjusted economics
Policy Engine
The Tax and Incentive Engine
How policy transforms a $5M system into a materially lower effective economic investment — front-loading cash value and compressing payback for both debt and equity capital providers.
100% Bonus Depreciation — IRC §168(k)
  • Full expensing of qualifying AI infrastructure in Year 1 — no multi-year depreciation schedule
  • $5M system × 25–30% blended tax rate
  • Immediate federal tax savings: $1.25M–$1.50M
  • Policy is currently enacted — not speculative
  • Creates immediate liquidity enhancement for borrowers and equity sponsors
Massachusetts Data Center Incentives
  • 20-year sales and use tax exemption for certified Massachusetts data centers
  • Applies to equipment, software, electricity, and qualifying construction costs
  • On a $5M GB300 system: avoided sales tax of $312K
  • Potential additional energy-efficiency incentives for high-performance cooling and optimized power infrastructure
  • State program is active and available to qualifying CNEX facilities
Effective Economic Basis: The Waterfall
Policy effectively returns roughly 30–35% of capex in Year-1 value and avoided tax, before considering energy-efficiency incentives. The effective economic basis of $3.2M–$3.4M against $5M annual EBITDA creates a return profile that is difficult to replicate in conventional infrastructure asset classes.
Credit Committee Brief
Why Credit Committees Should Care
This is not just tax efficiency. It is credit enhancement. The incentive structure directly improves the borrower's debt service capacity, collateral coverage, and default risk profile — making CNEX GB300 financings materially more attractive to structured lenders than standard infrastructure credit.
Direct Lender Benefits
Lower cash taxes improve Year-1 liquidity and operating cash flow
Stronger DSCR from front-loaded tax savings reducing net debt service burden
Reduced effective borrower net investment improves loan-to-value ratios
Faster deleveraging and better collateral coverage throughout the loan term
Earlier EBITDA capture materially reduces default risk in early loan periods
What Improves If the Deal Closes Now
Lower basis against the same revenue stream — stronger collateral coverage from day one
Better early cash sweep potential, accelerating principal reduction
Stronger debt service capacity from combined EBITDA and incentive cash flows
Better-looking underwriting metrics than later vintages with higher capex basis
More attractive risk-adjusted lending profile relative to peer infrastructure credits

Closing now improves both borrower economics and lender protection. Waiting weakens both — higher capex raises the loan amount while lower tax shields reduce cash available for debt service.
Equity Investors
Why Early Equity Wins
For equity investors, the timing advantage is not marginal — it is structural. Earlier deployment means lower cost basis, earlier EBITDA, larger tax shields, and a superior IRR vintage. Each quarter of delay compounds against the investor's return profile.
Immediate Tax Shield
A meaningful portion of the equity check is effectively returned via Year-1 federal and state tax benefits — creating policy-backed downside protection from the outset.
Superior Return Vintage
Early GB300 systems lock in lower cost basis, earlier EBITDA capture, and stronger IRR than later deployments facing higher capex and diminished policy value.
Scarcity Premium
Scarce, tax-advantaged, early-cycle AI capacity should command premium market value relative to later, higher-cost systems — creating embedded asset appreciation potential.
Faster Payback
Combined tax and incentive capture compresses payback toward sub-1-year territory under current operating assumptions — an exceptional profile for infrastructure equity.
Timing Arbitrage
The Real Decision Is Not "Invest or Don't Invest."
It Is "Invest Now or Destroy Value."
The comparison below is not theoretical. Each line item represents measurable economic value that is either captured or forfeited based solely on when the financing closes. The gap widens materially with every passing week.
Waiting is financially irrational: investors risk paying more, earning later, and capturing less tax value — all simultaneously. The window is not wide.
Napkin Math
30-Second Napkin Math
The arithmetic is unambiguous. Strip away the complexity and the case reduces to a straightforward value capture calculation that any credit committee can evaluate in under a minute.
1
$5.0M GB300 Acquired
Pricing locked until 04/15/2026 — system qualifies for full federal and state incentive stack
2
$1.25M–$1.50M
Immediate IRC §168(k) federal tax savings — full Year-1 expensing at 25–30% blended rate
3
+ $312K
Massachusetts sales and use tax savings from 20-year data center exemption program
4
= $1.56M–$1.81M
Total Year-1 value capture — before energy-efficiency incentives or EBITDA contribution

Against $5.0M annual EBITDA, failing to act within the pricing window destroys value before the system is even deployed. The incentive clock and the pricing clock are running simultaneously.
"Not acting in the next 3 days is economically irrational. The numbers are not close."
Return Profile
Cash Flow and IRR Compression vs. Expansion
The incentive-adjusted return profile transforms what is already a compelling infrastructure investment into an exceptional one. The comparison below illustrates the quantified difference between acting now and waiting — measured in real IRR points and payback compression.
Case A — Without Incentives
Capex: $5.0M nominal, no reduction
Payback: 1.2–1.3 years after tax at standard rates
Unlevered IRR: High-teens — solid but not exceptional
Levered IRR: Moderate uplift — limited by higher net basis
Case B — With Bonus Depreciation + MA Incentives
Effective Economic Basis: $3.2M–$3.4M after Year-1 policy value
Payback: Compresses toward 0.8–1.0 years — exceptional for infrastructure
Unlevered IRR: Moves into low- to mid-20s — materially superior vintage
Levered Equity IRR: Significantly higher — early cohort economics compound through leverage
The incentive-adjusted case is not a marginal improvement — it represents a fundamentally different infrastructure return profile. The gap between Case A and Case B is entirely a function of timing: deploy now versus deploy later.
Location Advantage
Why Massachusetts Strengthens the Thesis
Massachusetts is not just a location advantage. It is a capital efficiency advantage. The state's data center incentive framework, combined with its position at the center of the Northeast AI and enterprise technology ecosystem, makes it one of the most defensible jurisdictions for AI infrastructure investment in the United States.
20-Year Tax Exemption
Sales and use tax exemption for certified data centers covering equipment, software, electricity, and qualifying construction — a 20-year runway of capital efficiency.
Energy Efficiency Incentives
Advanced cooling and power optimization systems may qualify for additional Massachusetts energy-efficiency incentives — reducing operating costs and improving system-level economics.
Boston-Cambridge Ecosystem
Proximity to one of the world's premier AI and life sciences innovation clusters — MIT, Harvard, major research hospitals, and a deep enterprise customer base within direct reach.
Enterprise Demand Density
Strong regional demand from financial services, biotech, defense, and academic institutions creates a defensible, high-value customer pipeline for AI compute capacity.
Strategic Narrative
Why This Window Exists Now
The convergence of three independent forces — policy tailwinds, hardware economics, and demand timing — creates a window that is real, measurable, and closing. Understanding why each pillar exists and how they interact is essential to sizing the urgency correctly.
Pillar I — Policy Tailwinds
Federal expensing under IRC §168(k) and Massachusetts state incentives are front-loading investment value in ways that are unlikely to persist indefinitely. The current policy stack is enacted — not speculative — but policy environments evolve. Investors who act now capture the full stack.
Pillar II — Hardware Inflation and Scarcity
AI infrastructure costs are rising, not falling, in the near term. GB300 systems at $5M represent a pricing point that reflects current supply conditions. As demand accelerates and supply chains tighten, credible expectations point to 10–30% capex inflation within 6–12 months — before accounting for any demand premium.
Pillar III — Demand Timing
Each month of delay forfeits high-margin EBITDA and strategic customer capture. Enterprise AI customers are making infrastructure commitments now. Early systems capture premium customers and lock in revenue streams that later entrants will compete to access at higher cost.
AI infrastructure today resembles the best early vintages in renewable energy, fiber, and hyperscale digital infrastructure — when early movers captured structurally superior economics that late entrants never recovered.
Platform Advantage
Why CNEX Is Positioned to Capture This Opportunity
CNEX is positioned to convert this narrow policy-and-pricing window into an investable, financeable, and scalable AI infrastructure platform opportunity. The combination of asset access, incentive qualification, and institutional-grade economics creates a compelling proposition for structured capital.
AI Factory Platform Model
CNEX operates as an AI factory platform — not a single-asset play. The model is designed for repeatability, scalability, and institutional monetization across multiple GB300 deployments.
Access to Premium GB300 Systems
CNEX has secured access to NVIDIA GB300 AI systems at the current $5M price point — a position that is both time-sensitive and not universally available to new entrants.
Massachusetts Infrastructure Advantage
Massachusetts-based infrastructure positioning enables CNEX to capture the full state incentive stack while serving the high-density demand market of the Boston-Cambridge innovation corridor.
Lender and Investor-Ready Economics
The unit economics of each GB300 deployment are structured to meet institutional credit and equity return thresholds — with DSCR, LTV, and IRR profiles that compare favorably to peer infrastructure assets.
Final Investment Thesis
Bottom Line
A $5M GB300 acquired now is not merely a hardware purchase. It is a tax-advantaged, high-yield digital infrastructure asset with immediate policy support, strong EBITDA potential, and superior timing economics. The investment case is not built on speculation — it is built on enacted policy, locked pricing, and measurable demand.
Lower Capex Basis
$5.0M entry price locked until 04/15/2026 — avoid 10–30% hardware inflation that begins immediately after the pricing window closes.
Earlier EBITDA Capture
$5.0M annual EBITDA per system begins from deployment — every quarter of delay is a quarter of high-margin cash flow permanently forfeited.
Larger and Earlier Tax Shields
$1.56M–$1.81M in combined Year-1 federal and state incentive value — front-loaded, policy-backed, and available only to systems deployed within the current window.
Stronger DSCR and Better Return Vintage
For both debt and equity, the combination of lower basis, earlier cash flow, and larger tax shields creates a structurally superior return profile relative to any later deployment cohort.
This is a rare convergence of policy, pricing, and technology cycle. The most attractive returns belong to the earliest systems deployed — not the later ones. The window is measured in days.
Institutional References
Selected References
The investment thesis set forth in this presentation is grounded in enacted federal tax law, active state incentive programs, and CNEX internal operating assumptions. The following reference categories underpin key claims throughout the analysis.
1
IRC §168(k) — Bonus Depreciation
Internal Revenue Code Section 168(k) governing bonus depreciation and full expensing of qualifying property. Consult current IRS guidance for applicable tax years and eligibility requirements.
2
IRS Guidance — Qualified Property
IRS publications and revenue procedures governing the definition of qualified property, placed-in-service requirements, and applicable depreciation methodologies for AI and computing infrastructure.
3
Massachusetts Data Center Incentive Program
Massachusetts General Laws and Department of Revenue guidance governing the data center sales and use tax exemption program, including certification requirements and eligible expenditure categories.
4
Massachusetts Sales and Use Tax Guidance
Massachusetts Department of Revenue administrative guidance on sales and use tax treatment of data center equipment, software, electricity, and qualifying construction and infrastructure costs.
5
Massachusetts Energy Efficiency Resources
Mass Save program resources and Massachusetts Department of Energy Resources guidance on energy-efficiency incentives applicable to data center cooling, power, and infrastructure optimization.
6
CNEX Internal Assumptions
Revenue projections, EBITDA margin assumptions, and operating model parameters are based on CNEX internal management estimates and are subject to revision. Past performance of comparable assets does not guarantee future results.

Important Disclaimer: This presentation is for informational purposes only and does not constitute an offer to sell or a solicitation to buy any security or financial instrument. All financial projections and tax benefit estimates are based on current assumptions and applicable law as of the date of preparation. Investors should conduct their own due diligence and consult qualified legal, tax, and financial advisors before making any investment decision. Tax treatment depends on individual circumstances and may be subject to change.
©2026 CambridgeNexus, Inc. · [email protected] · GB300 NVL72 · AIFaaS · New England AI Infrastructure