Applied Digital To Merge Cloud Unit With EKSO, Creating AI Compute Platform

Applied Digital has entered a non-binding term sheet to spin out its cloud business and merge it with Ekso Bionics to form ChronoScale, a standalone GPU-accelerated compute platform targeting next-generation AI workloads; Applied Digital Cloud reported roughly $75.2 million in trailing twelve-month revenue as of August 31, 2025. Upon closing (expected in H1 2026 and subject to due diligence, regulatory approvals and shareholder votes) Applied Digital would own about 97% of the combined company while EKSO intends to continue its exoskeleton operations during a strategic review and may pursue sale options; EKSO shares jumped ~41% in overnight trading on the announcement.

Analysis

Market structure: The carve‑out and combination effectively creates a pure‑play, GPU‑accelerated infrastructure provider (ChronoScale) that directly benefits Applied Digital (APLD) as ~97% owner and potentially NVIDIA (NVDA) via incremental H100 demand; EKSO (EKSO) benefits only if minority holders capture value from a sale of its legacy exoskeleton business. Expect pricing power for bespoke, high‑density GPU racks in 2026 if supply of H100/H200 remains constrained — that supports >20% premium pricing versus commodity cloud for latency/control‑sensitive customers in the first 12–24 months. Incumbent hyperscalers (AWS/GOOG/MSFT) are mildly threatened at the margin but retain scale advantages, so market share gains for ChronoScale are likely single‑digit percentage points annually, not a wholesale disruption.

Risk assessment: Key tail risks are deal collapse, regulatory pushback, accelerated NVDA supply easing (which would compress gross margins), or heavy capex overruns; each could erase >50% of current paper gains. Timeline: immediate (days) — EKSO volatility spike and potential mean reversion; short (weeks–months) — due diligence reveals liabilities or dilution in S‑4; long (quarters–years) — execution risk on rolling out rack scale and securing long‑term power contracts. Hidden dependencies include NVDA roadmap/timing, colo power contracts, and customer stickiness; loss of a few anchor customers or concessions on GPU pricing are asymmetric negatives.

Trade implications: Direct plays — prefer APLD exposure (majority owner) over EKSO equity pop: allocate 2–3% notional to APLD stock or LEAP calls to capture upside to ChronoScale over 12–18 months. Short/hedge EKSO post‑pop via a 0.5–1% short or buy Jan 2027 $6 puts to protect against a >30% retracement when S‑4 details dilution; close on S‑4 filing or H1 2026 outcome. Options strategies — buy APLD Jan 2027 40–60% delta LEAP calls (1–2% of book) and sell short‑dated EKSO calls or buy puts to exploit elevated near‑term IV.

Contrarian angles: The market is likely overpricing EKSO minority participation — Applied Digital’s 97% stake means most economic upside accrues to APLD, not existing EKSO holders; the 41% overnight move is likely >realizable fundamental value. Historical parallels (specialized rack spinoffs) show value only if companies secure multi‑year reserved contracts; without PBAs/term sheets for anchor customers, revenue growth can stall and valuation compress. Unintended consequence: hyperscalers could cut GPU spot pricing to defend share, pressuring margin for niche providers — a 10–20% price concession scenario would materially reduce ChronoScale IRR projections.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in APLD within 10 trading days to capture majority exposure to ChronoScale; target +30–50% upside over 12 months, set a hard stop‑loss at −25% from entry and take profits at +50–70% or upon S‑4 filing showing <15% dilution.
• Open a 0.5–1% notional short position in EKSO (or buy Jan 2027 $6 puts sized to 1% of portfolio) to fade the 41% pop; cover if EKSO trades above $12 or if S‑4 shows EKSO minority economics materially aligned (>=20% pro‑rata stake in newco).
• Buy APLD Jan 2027 LEAP calls (40–60% delta) sized 1–2% of portfolio as a convex, medium‑term upside play; trim half at +100% and unwind if Form S‑4 discloses >15% adverse dilution or loss of anchor customer contracts within 90 days of filing.
• Require S‑4 review trigger: if SEC filings disclose material capex commitments >$200M, GPU supply shortfall, or customer concentration >40% within 30–60 days, reduce APLD exposure by 50% and shift proceeds into larger, diversified infra names (e.g., Equinix/CORE REIT equivalents) to de‑risk execution/capital intensity.