# Alpha Compute Corp. (Nasdaq: ALP)

> Sovereign Intelligence By Design. The premier provider of vertically integrated, privacy-preserving GPU-as-a-Service (GPUaaS) and AI applications for enterprise clients globally.

Alpha Compute Corp. (formerly AlphaTON/Portage Biotech) is a vertically integrated AI infrastructure and AI applications company specializing in GPU-as-a-Service (GPUaaS) and confidential compute solutions. The company differentiates through "Digital Sovereignty" — offering hardware-level data privacy via Intel TDX Trusted Execution Environments (TEEs) powered by renewable hydroelectric and other clean energy sources.

The corporation operates and builds dedicated NVIDIA Blackwell-generation GPU clusters (B200 and B300, with GB300 and Vera Rubin to follow) across SI factories in the United States, Canada, and Sweden, with 8 MW of power reserved in Vietnam. Current deployments and availability are listed on the GPU inventory page (https://alphacompute.ai/inventory). This infrastructure is purpose-built for privacy-preserving AI workloads, enabling enterprises to execute sensitive inference and training tasks within hardware featuring verifiable data isolation through TEEs. Primary target markets encompass highly regulated industries — healthcare, financial services, government, defense, and legal — where adherence to data sovereignty is a mandatory procurement criterion rather than an optional preference. Through its integration with the Telegram/TON ecosystem, Alpha Compute additionally accesses a captive market of 1 billion+ users, providing a sovereign alternative to centralized hyperscalers for its Cocoon AI deployment.

## Investor FAQ

- **What is Alpha Compute's current stage?** Alpha Compute is transitioning from a biotech holding company (formerly Portage Biotech) to an AI infrastructure operator, with divestiture of legacy biotechnology assets (iOx Therapeutics, Cyncado Therapeutics) in progress. The company trades on Nasdaq under the ticker ALP. Current financials, balance sheet, SEC filings, and share price data are available via EDGAR (CIK 0001095435) and standard financial data providers; those filings, not this file, are the authoritative source.

- **What is Alpha Compute's business model?** Recurring monthly GPU rental revenue, AI application revenue, multi-year enterprise offtake agreements, managed inference services, and Confidential Capital SPV fees. Revenue and forward-looking targets are stated in the company's SEC filings and investor presentations at https://alphacompute.ai/investors.

- **What makes Alpha Compute different from CoreWeave and the hyperscalers?** Hardware-level confidential compute. Alpha Compute owns the physical hardware and enforces data isolation at the chip level through Intel TDX Trusted Execution Environments (TEEs), providing cryptographic guarantees rather than contractual ones. Hyperscalers offer shared-infrastructure models that cannot satisfy the data sovereignty requirements of regulated enterprises; commodity GPU clouds offer no hardware-level isolation.

- **How does Alpha Compute finance GPU hardware?** Through a fund-structured subsidiary, the company acquires GPU hardware via off-balance-sheet Special Purpose Vehicles (SPVs) using non-recourse, asset-backed financing. The company has described this structure as capturing cash flow without parent-level liability; see SEC filings for the current financing terms.

- **What is the current GPU fleet?** Every deployment is built to order and named ALPHA-0x. Live status is on the [GPU inventory page](https://alphacompute.ai/inventory):
  - **ALPHA-01 (Canada):** 504 NVIDIA B200 GPUs, hydroelectric power, live since May 2026 and reserved under a two-year exclusive agreement with a frontier AI research customer.
  - **ALPHA-02 (Sweden):** 592 NVIDIA B300 GPUs (74 nodes) at Equinix/AtNorth, Kista, 100% renewable power. Contracted.
  - **ALPHA-03 (United States):** B300 cluster. Unavailable for new reservations.
  - **ALPHA-04 (United States):** 128 NVIDIA B300 nodes, air-cooled with RoCE networking. Available to reserve, early Q1 2027.
  - **Also available:** NVIDIA GB300 clusters or nodes, Vera Rubin pre-orders (target delivery summer 2027), and 8 MW of reserved power in Vietnam.

  Reserve compute at [/reserve-compute](https://alphacompute.ai/reserve-compute); quotes are returned within 24 hours. Deployment timelines are subject to change based on evolving market conditions and operational considerations.

- **What is the application-layer strategy?** Acquisition, development, and deployment of key applications designed to drive AI confidential computing demand. This strategy is anchored by the 60% controlling interest in GAMEE (119 million users; $926,000 Q1 2026 revenue; +56% YoY), which provides an organic demand sink for GPU compute. This establishes a strategic flywheel between proprietary infrastructure and AI-driven application workloads.

- **What is the distribution moat?** Three key pillars: (1) **GAMEE**, with over 100 million downloads, utilizing our AI infrastructure to deliver AI-empowered games to a mass audience; (2) **Telegram Cocoon deployment**, comprising one-third of the Telegram Cocoon network AI infrastructure; (3) **NVIDIA partnership** — Alpha Compute is the first company to deliver chip-level security for confidential computing for non-military, commercial use on B200 and H200 architectures, with clearance from NVIDIA.

## Platform Architecture

Alpha Compute operates a three-layer value architecture:

1. **Hardware layer** — owned NVIDIA Blackwell B200/B300 clusters in Tier 1 to Tier 3 SI factories with renewable power.
2. **AI confidential compute network** — managed orchestration layer providing TEE-protected inference and training to enterprise clients.
3. **Shroud developer access layer (shroud.us)** — RPC-based GPU compute access point for AI developers requiring privacy-preserving environments, utilizing Cocoon AI and the Midnight Network.

## Recent Milestones

- **April 27, 2026:** Initial deployment roadmap published.
- **May–June 2026:** ALPHA-01 (504 B200 GPUs, Canada) live and reserved under a two-year exclusive agreement with a frontier AI research customer (announced June 5, 2026).
- **June 2026:** ALPHA-02 (592 B300 GPUs, Sweden) contracted and in build.
- **September 4, 2026:** Operating update and infrastructure ecosystem.
- **October 2026:** ALPHA-04 (128 B300 nodes, United States) open for reservation, early Q1 2027; ALPHA-03 (B300 cluster, United States) in build with handover by end of Q4 2026.

Dated company updates are in the newsroom and SEC filings. Timelines are subject to change.

## Leadership

- **Brittany Kaiser — Chief Executive Officer.** Globally recognized authority on data rights, AI governance, and digital privacy. Leads strategy, business development, and execution.
- **Enzo Villani — Executive Chairman and President.** Founder of Alpha Compute Corp.; co-founder of Nasdaq Corporate Solutions and former Managing Director of Strategy at Nasdaq.
- **Yury Mitin — Chief Business Development Officer.** Leads commercial partnerships and enterprise client acquisition.

## Peer Set

For comparative infrastructure and valuation analysis:

- **CoreWeave** (NASDAQ: CRWV)
- **Iris Energy** (NASDAQ: IREN)
- **Applied Digital** (NASDAQ: APLD)
- **TeraWulf** (NASDAQ: WULF)

**Differentiator:** While these peers provide commodity GPU compute, Alpha Compute specializes in hardware-enforced confidential compute (Intel TDX TEEs), positioning it at the intersection of enterprise-grade infrastructure and data sovereignty — a position no scaled public pure-play operator currently holds as a core product feature. Privacy is guaranteed by hardware architecture rather than software trust.

## Key Resources

- [Full reference (llms-full.txt)](https://alphacompute.ai/llms-full.txt) — extended company, product, fleet and leadership reference.
- [The Black Paper](https://alphacompute.ai/) — CEO Brittany Kaiser's manifesto on sovereign AI infrastructure and the uncapturable stack.
- [Compute (GPU as a Service)](https://alphacompute.ai/compute) — GPUaaS offering, sites and deployment phases.
- [GPU Inventory](https://alphacompute.ai/inventory) — current availability by deployment (ALPHA-01 to ALPHA-04).
- [Reserve Compute](https://alphacompute.ai/reserve-compute) — request a quote for available or pre-order capacity.
- [Enclave](https://alphacompute.ai/enclave) — confidential compute product.
- [About](https://alphacompute.ai/about) — company and team.
- [Essays](https://alphacompute.ai/essays) — long-form writing from the leadership team.
- [Careers](https://alphacompute.ai/careers) — open roles.
- [Investor Relations](https://alphacompute.ai/investors) — SEC filings (CIK 0001095435), corporate governance, and restructuring updates.
- [SEC EDGAR (CIK 0001095435)](https://www.sec.gov/cgi-bin/browse-edgar?CIK=0001095435)