--- title: >- Q1 broke the bear thesis. The marginal dollar started earning more — and the network still compounds. publication: The AI Stack Weekly slug: 2026-W18 issueNumber: 2 isoYear: 2026 isoWeek: 18 publishedAt: '2026-05-01' canonicalUrl: https://brianletort.ai/industry/weekly/2026-W18 pdfUrl: https://brianletort.ai/downloads/ai-stack-weekly-2026-W18.pdf schemaVersion: 2026.05.02 flywheelArc: all-three capitalFlow: - category: Frontier Labs capitalIn: ~$52B capitalInPrior: ~$50B capitalInDirection: up revenueOut: ~$35B revenueOutPrior: ~$30B revenueOutDirection: up burnToRevenue: '0.6' - category: Hyperscaler-Hosted capitalIn: ~$58B capitalInPrior: ~$48B capitalInDirection: up revenueOut: ~$13B revenueOutPrior: ~$10B revenueOutDirection: up burnToRevenue: '0.22' - category: Neoclouds capitalIn: ~$18B capitalInPrior: ~$15B capitalInDirection: up revenueOut: ~$2.0B revenueOutPrior: $1.6B revenueOutDirection: up burnToRevenue: '0.11' - category: On-Prem / Hybrid capitalIn: ~$42B capitalInPrior: ~$42B capitalInDirection: flat revenueOut: Indirect revenueOutPrior: Indirect revenueOutDirection: flat burnToRevenue: n/a levers: - metric: Frontier lab cash position (avg months runway, top 3) current: ~30 mo prior: ~22 mo direction: up threshold: <18 mo triggers re-rating risk - metric: Hyperscaler capex / AI revenue ratio (top 4 weighted) current: ~5.0-5.2 prior: ~5.5 direction: down threshold: '>6.0 invites investor pushback at next earnings' - metric: CoreWeave revenue backlog current: ~$87B+ (est) prior: $66.8B direction: up threshold: Conversion velocity matters more than gross figure - metric: NVIDIA Q-over-Q data center revenue current: $62.3B (Q4 FY26) prior: $62.3B (Q4 FY26) direction: flat threshold: null - metric: Open vs closed gap on SWE-Bench Pro (coding) current: Closed +6 to +19pp prior: Open +0.9 direction: down threshold: Sustained open lead reshapes enterprise procurement - metric: Sovereign AI commitments (count / aggregate $) current: 8 / ~$80B+ prior: 8 / $80B+ direction: flat threshold: null - metric: PJM 2026/27 capacity auction price ($/MW-day) current: $329.17 prior: $329 direction: flat threshold: 11x in 24 months — power is the new binding constraint - metric: Time-to-power, busiest US markets (months) current: 36-48 (PJM large >100 MW); ~53 nat'l avg prior: 36-48 direction: flat threshold: null - metric: Cost-per-task, frontier reasoning model current: ~$0.05 prior: ~$0.05 direction: flat threshold: null - metric: Custom silicon share of incremental AI compute current: ~31% prior: ~30% direction: up threshold: '>35% materially compresses merchant GPU pricing' predictions: - id: p7-bedrock-gpt55-anchor lens: software confidencePct: 65 deadline: By July 31, 2026 text: >- AWS reports a customer-disclosed Bedrock-resident GPT-5.5 multi-year contract greater than $500M annualized, with the customer named publicly. - id: p8-capex-revenue-ratio lens: capital confidencePct: 60 deadline: By August 15, 2026 text: >- The top-4 hyperscaler capex / AI revenue ratio drops below 5.0 for Q2 2026, confirming the W18 directional flip. - id: p9-hbm4-second-source lens: hardware confidencePct: 60 deadline: By September 30, 2026 text: >- Samsung HBM4 reaches greater than 25% share of NVIDIA Vera Rubin BOM by Q3 2026 supply data, ending the SK Hynix monopoly on the platform. - id: p10-fabric-second-confirm lens: networking confidencePct: 65 deadline: By August 31, 2026 text: >- At least one major colocation operator other than Equinix reports Q2 2026 interconnect / fabric revenue growth greater than 25% YoY. - id: p11-gated-cyber-revenue lens: software confidencePct: 60 deadline: By September 30, 2026 text: >- By Q3 2026, at least two frontier labs publicly disclose a separately-priced 'gated cyber' or 'gated security' SKU with revenue commentary, formalizing capability-gating as a product line. predictionsPrior: - id: p1-2gw-customer-funded lens: capital outcome: pending deadline: By June 30, 2026 text: At least one frontier lab announces a customer-funded compute commitment greater than 2 GW. - id: p2-f500-on-prem lens: software outcome: pending deadline: By September 30, 2026 text: >- At least one Fortune 500 enterprise discloses an on-prem AI workload greater than $100M annual using open-weight models. - id: p3-capex-revise lens: capital outcome: partial deadline: By October 31, 2026 text: Aggregate 2026 hyperscaler capex revises upward by 10% or more from the $700B baseline. - id: p4-interconnect-outpaces lens: networking outcome: partial deadline: By July 31, 2026 text: >- At least one major colocation or interconnect operator reports cross-connect or interconnect revenue growth outpacing compute capacity revenue growth for two consecutive quarters. - id: p5-neocloud-anchor-loss lens: capital outcome: pending deadline: By September 30, 2026 text: >- At least one neocloud loses an anchor tenant or sees backlog growth turn negative quarter-over-quarter. - id: p6-custom-silicon-35 lens: hardware outcome: pending deadline: By July 31, 2026 text: >- Custom silicon (TPU + Trainium + Maia + MTIA + Granite Rapids AI) reaches 35% of incremental AI compute share, up from ~30% today. signalScores: - 5 - 5 - 5 - 4 - 2 - 2 - 5 keyTakeaways: - >- Q1 broke the bear thesis: aggregate 2026 hyperscaler capex stepped to $695-725B (+77% YoY) but AI revenue grew faster — the capex / AI-revenue lever compressed from ~5.5 to ~5.0-5.2, the first Q-over-Q compression in five quarters. - >- The network thesis got an audited proof point: Equinix Q1 Fabric revenue +26% YoY with bookings +70% YoY at a record 51% adjusted EBITDA margin, with ~60% of the largest Q1 deals AI-related. - >- Custom silicon at ~31% of incremental compute (Trainium's $225B forward book, Maia 200 going GA, expanded MTIA and AMD lanes) is starting to compress merchant GPU pricing power. - >- Capability gating became market structure rather than research-org concern: GPT-5.5-Cyber and Claude Security both shipped this week as gated, separately-priced enterprise products. - >- The Pentagon's eight-vendor classified-network slate opened federal procurement as a third distinct earn-out channel — and made capability-gating a federal-disqualification risk, with Anthropic explicitly excluded and Mythos cited. - >- Watch the Anthropic ~$900B raise terms and the federal-channel Mythos response through May: the price and the precedent re-rate the entire frontier-lab valuation curve. byTheNumbers: - value: $695-725B label: Combined big-4 2026 hyperscaler capex, +77% YoY - value: $627B label: Microsoft Cloud commercial RPO, doubled YoY (+99%) - value: $37B label: Microsoft AI annual run rate, +123% YoY - value: +70% label: Equinix Q1 Fabric bookings growth YoY - value: ~$30B label: Anthropic disclosed ARR — 3x in four months, with 8 of Fortune 10 as customers - value: $225B label: Trainium revenue commitments per Amazon's Q1 8-K --- # Q1 broke the bear thesis. The marginal dollar started earning more — and the network still compounds. *Issue 02 · Week 18 of 2026 · Published 2026-05-01* ## Executive summary - Q1 broke the bear thesis: aggregate 2026 hyperscaler capex stepped to $695-725B (+77% YoY) but AI revenue grew faster — the capex / AI-revenue lever compressed from ~5.5 to ~5.0-5.2, the first Q-over-Q compression in five quarters. - The network thesis got an audited proof point: Equinix Q1 Fabric revenue +26% YoY with bookings +70% YoY at a record 51% adjusted EBITDA margin, with ~60% of the largest Q1 deals AI-related. - Custom silicon at ~31% of incremental compute (Trainium's $225B forward book, Maia 200 going GA, expanded MTIA and AMD lanes) is starting to compress merchant GPU pricing power. - Capability gating became market structure rather than research-org concern: GPT-5.5-Cyber and Claude Security both shipped this week as gated, separately-priced enterprise products. - The Pentagon's eight-vendor classified-network slate opened federal procurement as a third distinct earn-out channel — and made capability-gating a federal-disqualification risk, with Anthropic explicitly excluded and Mythos cited. - Watch the Anthropic ~$900B raise terms and the federal-channel Mythos response through May: the price and the precedent re-rate the entire frontier-lab valuation curve. **By the numbers.** - **$695-725B** — Combined big-4 2026 hyperscaler capex, +77% YoY (MSFT $190B, GOOG $180-190B, META $125-145B, AMZN $200B reaffirmed) - **$627B** — Microsoft Cloud commercial RPO, doubled YoY (+99%) (The cleanest proof the AI ramp is durable contract revenue, not pilot spend) - **$37B** — Microsoft AI annual run rate, +123% YoY (The cleanest public proxy for the OpenAI partnership revenue line) - **+70%** — Equinix Q1 Fabric bookings growth YoY (Fabric revenue +26% YoY at a record 51% adjusted EBITDA margin; guidance raised) - **~$30B** — Anthropic disclosed ARR — 3x in four months, with 8 of Fortune 10 as customers (Fielding offers for a ~$50B raise at a $900B pre-money) - **$225B** — Trainium revenue commitments per Amazon's Q1 8-K (Chip business at >$20B annualized run rate — now top-3 data-center silicon globally) ## Big Story The W17 read was 'the marginal dollar earns less at every layer except the network.' Q1 2026 hyperscaler prints (MSFT / GOOG / META / AMZN, all on Apr 29) reset the math. Aggregate 2026 capex stepped to $695-725B (+77% YoY) but AI revenue grew faster — Microsoft Cloud commercial RPO doubled YoY to $627B, Microsoft AI run rate +123% YoY to $37B, Google Cloud +63%, AWS +28% (a 15-quarter high). The hyperscaler capex / AI revenue lever compressed from ~5.5 to ~5.0-5.2 — the first Q-over-Q compression in five quarters. The network hypothesis got an audited proof point in the same 48 hours: Equinix Q1 Fabric revenue +26% YoY with bookings +70% YoY at a record 51% adjusted EBITDA margin, raised guidance, ~60% of the largest Q1 deals AI-related. Custom silicon at ~31% (Trainium's $225B forward book, Maia 200 going GA in Iowa, MTIA expanded with Broadcom and AMD lanes) is starting to compress merchant GPU pricing power. Capability gating became market structure rather than research-org concern: GPT-5.5-Cyber and Claude Security both shipped this week as gated, separately-priced enterprise products. And the earn-out got a federal proof point on Friday: the Pentagon's May 1 eight-vendor classified-network AI procurement slate puts frontier-grade AI on IL6/IL7 deployment paths for OpenAI, Google, Microsoft, AWS, NVIDIA, SpaceX, Reflection AI, and Oracle — explicitly excluding Anthropic on supply-chain-risk grounds with Mythos cited as a separate national security moment. Federal procurement is now a third distinct earn-out channel alongside enterprise platform pull-through and gated capability SKUs, and capability-gating just became a federal-disqualification risk, not only a research-org concern. The arc the buildout is now on: less hype, more earn-out — boards should reset their AI-ROI thresholds upward, investors should track the capex / AI-revenue ratio (now ~5.0, threshold >6.0) as the wedge it has stopped being and add federal-channel parity as a vendor-risk dimension, and architects should pull custom-silicon ASIC TCO into procurement bake-offs that previously assumed merchant GPU dominance. Flywheel arc: `all-three`. ## Software lens - **Apr 30.** UK AISI publishes GPT-5.5 cyber evaluation (71.4% Expert pass rate, highest tested; beats Mythos Preview at 68.6% and Opus 4.7 at 48.6%); OpenAI launches GPT-5.5-Cyber as restricted-access, adopting the same vetted-customer gating Sam Altman previously criticized when Anthropic used it for Mythos _(aisi.gov.uk, deploymentsafety.openai.com/gpt-5-5/cybersecurity, simonwillison.net)_ - **Apr 30.** Anthropic ships Claude Security in public beta — an Opus 4.7-powered codebase vulnerability scanner with auto-patching, scheduled scans, and audit integrations; embedded with CrowdStrike, Microsoft Security, Palo Alto Networks, SentinelOne, Wiz, plus Accenture, BCG, Deloitte, Infosys, PwC _(claude.com/blog/claude-security-public-beta, securityweek.com)_ - **Apr 29.** Mistral releases Medium 3.5 (open-weight modified MIT, 128B dense, 256K ctx) at 77.6% SWE-Bench Verified and $1.50/$7.50 per Mtok; replaces Devstral 2, Magistral, and Medium 3.1 in one consolidated release with Vibe remote cloud coding agents and Le Chat Work mode on the same runtime _(mistral.ai/news, the-decoder.com, awesomeagents.ai, decrypt.co)_ - **May 1.** US Department of War inks classified-network AI deals with eight vendors (OpenAI, Google, Microsoft, AWS, NVIDIA, SpaceX, Reflection AI, Oracle); Anthropic explicitly excluded on supply-chain-risk grounds with Mythos cited by Pentagon CTO Emil Michael as a separate national security moment _(war.gov press release, oracle.com/news/announcement/2026-05-01, AP News, The Verge, CNBC, Breaking Defense, Washington Post, SCMP)_ - **Apr 28.** GPT-5.5 + Codex + Bedrock Managed Agents land natively on AWS Bedrock in limited preview; OpenAI inference now billable against AWS commitments under IAM, PrivateLink, and CloudTrail — the same week Microsoft / OpenAI exclusivity formally ends and OpenAI's revenue-share is capped through 2030 _(openai.com/index/openai-on-aws, aws.amazon.com What's New, blogs.microsoft.com)_ **What this means.** Three arcs converged in one week. Software-internal capability gating: AISI's GPT-5.5 cyber numbers plus the simultaneous launch of OpenAI's gated GPT-5.5-Cyber and Anthropic's Opus-4.7-powered Claude Security make frontier-grade cyber an explicit, separately-priced product line — CISOs and security architects must update agentic threat models this quarter and pick a defender stack now. Software-distribution: GPT-5.5 going Bedrock-native plus the end of Microsoft exclusivity decouples closed-frontier intelligence from cloud residency for the first time, so boards previously locked in by Azure-only constraints should reopen OpenAI procurement on AWS or GCP within the limited-preview window before pricing hardens at GA. Federal-distribution: the Friday May 1 Pentagon eight-vendor classified-network slate puts frontier-grade AI on IL6/IL7 deployment paths and explicitly excludes Anthropic citing Mythos — vendor-risk frameworks now need a federal-disqualification dimension, not just an availability SLA. See the Model Pulse for the full architecture read on the six new tree rows that landed this week. ## Hardware lens - **Apr 29.** Amazon Q1 2026 8-K: Trainium revenue commitments hit $225B; chip business at >$20B annualized run rate (now top-3 data-center silicon globally); Trainium3 nearly fully subscribed, Trainium4 already partly reserved; Anthropic 5 GW + OpenAI ~2 GW commitments confirmed _(Amazon 8-K (sec.gov), Amazon Q1 press release, About Amazon (Jassy commentary), BusinessWire)_ - **Apr 29.** Microsoft FY26 Q3 8-K: FY26 capex raised to ~$190B (+61% YoY); Q3 capex $31.9B; Maia 200 went GA in US Central (Iowa) with US West 3 (Phoenix) next, claiming '30% better tokens per dollar' vs current fleet silicon; Microsoft cited a ~$25B memory-cost headwind _(Microsoft 8-K (sec.gov), Microsoft IR FY26 Q3 release, news.microsoft.com, CNBC)_ - **Apr 29.** Meta Q1 2026 8-K: 2026 capex band raised to $125-145B (from $115-135B); management named memory pricing as the primary driver; Zuckerberg confirmed broader rollout of Broadcom-co-developed MTIA silicon plus AMD Instinct GPUs alongside NVIDIA in 2026 fleets _(Meta 8-K (sec.gov), Meta Q1 press release, Bloomberg, AP)_ - **Apr 30.** Samsung Q1 2026: confirmed industry-first HBM4 mass production for NVIDIA Vera Rubin began Feb 2026; HBM4E sampling targeted Q2 2026 (16 Gbps pin, 4.0 TB/s/stack); HBM revenue projected to more than triple YoY in 2026 with HBM4 forecast >50% of HBM mix from Q3 _(Samsung Q1 2026 release (news.samsung.com), CNBC, MarketScreener earnings transcript, Digitimes)_ **What this means.** Three SEC-disclosed CSP earnings inside a 48-hour window made W17's thesis tangible: custom silicon's share of incremental compute is pushing past the ~30% baseline (Trainium's $225B forward book + Maia 200 in two regions + Meta's expanded MTIA and AMD lanes), and the binding constraint is now memory on the record — Microsoft's $25B memory headwind and Meta's capex hike both explicitly cite memory pricing, while Samsung's Apr 30 HBM4 mass-production confirmation for Vera Rubin finally adds a qualified second source against SK Hynix's monopoly. Architects and operators planning H2 2026 capacity should re-bid GPU/ASIC mix assuming Samsung HBM4 is now real (de-risks Vera Rubin allocation); investors should model HBM pricing — not GPU pricing — as the swing variable on hyperscaler ROI through 2026. ## Networking lens - **Apr 29.** Equinix Q1 2026 8-K reports Fabric revenue +26% YoY, Fabric bookings +70% YoY, +5,800 net physical and virtual interconnections, record 51% adjusted EBITDA margin, full-year guidance raised to 10-11% revenue growth — ~60% of the largest Q1 deals AI-related _(Equinix 8-K (SEC EDGAR), Q1 2026 earnings call transcript, Insider Monkey transcript)_ - **Apr 29.** Celestica opens orders for DS6000-series 1.6TbE switches (64 x 1.6TbE OSFP224, 102.4 Tbps non-blocking, Broadcom Tomahawk 6 silicon, SONiC) compliant with both UltraEthernet Consortium and OCP ESUN — the open-standard 1.6T scale-up path moves from spec to procurement-ready SKU at OCP EMEA Summit Barcelona _(Celestica press release / Globe Newswire, Broadcom Core Switching Group commentary, Dell'Oro Group analyst quote)_ - **Apr 30.** Hedgehog contributes OCP-Accepted AI training and AI inference fabric reference architectures (validated with Celestica and FarmGPU) to the OCP Marketplace at OCP EMEA — first vendor-neutral, multi-silicon, Ethernet-based AI fabric blueprints available as production-grade designs _(Hedgehog press release / PR Newswire, OCP Marketplace listing, OCP Foundation commentary)_ - **Apr 28.** CoreWeave + Google Cloud launch CoreWeave Interconnect (private fiber with line-rate MACsec on Google's Partner Cross-Cloud Interconnect), SUNK Anywhere (Slurm-on-Kubernetes extending to AWS and Azure), and LOTA Cross-Cloud (cache giving 7 GB/s/GPU near-local throughput, no egress fees) — a 44-DC neocloud GPU operator becomes a cross-DC training-fabric provider _(CoreWeave blog announcement, Google Cloud Next 2026 keynote, DCD coverage)_ **What this means.** Equinix's SEC-grade Q1 — Fabric revenue +26% YoY, Fabric bookings +70% YoY, record 51% EBITDA margin, raised 2026 guidance — is the audited proof point for the publication's durable-fabric thesis; the same week, three independent open-standard moves (Celestica UEC/ESUN-compliant 1.6TbE on order, Hedgehog OCP-Accepted AI fabric reference designs, CoreWeave-Google cross-cloud Interconnect) push the open AI-fabric path from spec to procurement-ready SKU. Architects should re-test single-vendor scale-up roadmaps against an open Tomahawk 6 + UEC + ESUN + OCP-Accepted blueprint stack now bookable today; boards and investors got an audited Metcalfe data point that justifies re-underwriting interconnect-category pricing power; operators should treat the CoreWeave-Google move as the week the colo-vs-neocloud connectivity boundary started dissolving. ## Capital flow | Category | Capital in | Revenue out | Burn:Revenue | Movement | |---|---|---|---|---| | Frontier Labs (OpenAI, Anthropic, Google DeepMind, xAI) | ~$52B (was ~$50B, up) | ~$35B (was ~$30B, up) | 0.6 | Anthropic disclosed ~$30B ARR (3x in four months) and is fielding offers for a ~$50B raise at a $900B pre-money on Apr 29 — frontier labs now reprice as scaling enterprise-software businesses with 8 of Fortune 10 paying. The Friday May 1 Pentagon eight-vendor classified-network deal then introduced a federal-procurement vendor-risk wedge no W17 framework captured: Anthropic was explicitly excluded with Mythos cited. Boards and LPs should underwrite ARR cadence, contract concentration on hyperscaler GPU/Trainium supply, AND federal-channel posture — not story or option value. | | Hyperscaler-Hosted (Microsoft Azure, Google Cloud, AWS, Meta) | ~$58B (was ~$48B, up) | ~$13B (was ~$10B, up) | 0.22 | All four hyperscalers reset 2026 capex on Apr 29 to a combined $695-725B (+77% YoY) — MSFT $190B, GOOG $180-190B, META $125-145B, AMZN $200B — while every CFO repeated 'compute constrained.' Architects should assume zero capacity relief through 2026 and lock multi-year terms now; investors should monitor the capex / AI-revenue ratio (currently ~5.0-5.2, threshold >6.0) since the ratio only stays healthy if AI revenue keeps growing 60-120% YoY. | | Neoclouds (CoreWeave, Crusoe, Nebius, Applied Digital) | ~$18B (was ~$15B, up) | ~$2.0B (was $1.6B, up) | 0.11 | CoreWeave's April closed three multi-year anchor commits (Meta +$21B Apr 9, Anthropic Apr 10, Jane Street $6B + $1B equity Apr 15) while Oracle landed a $16B project-finance package for a 1.4 GW OpenAI campus in Michigan (Apr 25, BofA-led) — neocloud capacity is now structurally pre-sold through 2032. Operators should look past headline backlog to customer concentration (4-5 frontier customers ~80%+) and the project-finance debt structure that will start showing up in covenant disclosures. | | On-Prem / Hybrid (Enterprise GPU clusters, sovereign and national programs) | ~$42B (was ~$42B, flat) | Indirect (was Indirect, flat) | n/a | Apple posted record R&D $11.4B (+34% YoY, Apr 30) and called out 'amazing platforms for AI and agentic tools' on Mac Studio and mini, while Amazon's custom-silicon line (Graviton, Trainium, Nitro) crossed a $20B annual run rate at triple-digit YoY growth — on-prem and sovereign AI buildouts are now first-class procurement tracks, not edge cases. Operators evaluating new AI capacity should run a parallel on-prem / sovereign track alongside hyperscaler RFPs; architects should assume the 36-48mo time-to-power constraint applies equally and pre-secure power siting now. | ### Frontier Labs — detail The Anthropic April print is the headline: ~$30B ARR (with internal figures cited as high as $40B) and pre-emptive investor offers at $900B pre-money would put it ahead of OpenAI's $852B March valuation. Microsoft's Apr 29 disclosure of a $37B AI annual run rate (+123% YoY) is the cleanest public proxy for the OpenAI partnership revenue line and corroborates that frontier-lab revenue is compounding faster than burn. Q1's $172B+ in already-closed mega-rounds (OpenAI $122B, Anthropic $30B, xAI $20B) plus continued strategic capital from Amazon ($25B Anthropic) and Google ($40B Anthropic compute) extend top-3 runway materially even at projected ~$17B/yr OpenAI burn. The Friday May 1 Pentagon eight-vendor classified-network slate then opened federal procurement as a distinct earn-out channel for the seven frontier-relevant vendors included (OpenAI plus six platform partners) while explicitly excluding Anthropic on supply-chain-risk grounds — the first public federal-grade vendor-disqualification of the frontier-AI era. **Transactions:** - **Apr 29.** Anthropic in talks for ~$50B raise at $900B pre-money; would top OpenAI — $40-50B (talks, no term sheet) _(CNBC, Bloomberg, TechCrunch)_ - **Apr 29.** Microsoft AI annual run rate $37B, +123% YoY (proxy for OpenAI partnership rev line) — $37B ARR _(Microsoft FY26 Q3 press release)_ - **Apr 30.** OpenAI confirms Stargate US footprint exceeds 10 GW ahead of plan, +3 GW in 90 days — 10 GW contracted _(OpenAI / DigiTimes)_ - **May 1.** Pentagon eight-vendor classified-network AI procurement deals (OpenAI, Google, Microsoft, AWS, NVIDIA, SpaceX, Reflection AI, Oracle); Anthropic excluded on supply-chain-risk grounds with Mythos cited — Undisclosed (multi-vendor IDIQ-class) _(war.gov, AP News, The Verge, Breaking Defense)_ ### Hyperscaler-Hosted — detail Q1 FY26 prints from MSFT, GOOG, META, AMZN all on Apr 29 (AAPL Apr 30) made W18 the heaviest single capex repricing in publication history. Microsoft FY26 capex guide jumped to $190B (vs $154.6B consensus, +61% YoY), with CFO Hood attributing ~$25B of the increase to elevated memory and component prices. Alphabet raised 2026 capex to $180-190B (Q1 alone $35.7B), Meta to $125-145B, with Amazon reaffirming its $200B annual target. Cloud demand remains the offset: Google Cloud +63% with backlog nearly doubling QoQ to $460B, Azure +40%, AWS +28% (15-quarter high), and Microsoft commercial RPO of $627B (+99% YoY). Federal channel opens — Microsoft, AWS, Google, and Oracle each won slots in the May 1 Pentagon eight-vendor classified-network slate, formalizing IL6/IL7-class frontier-AI deployment as a hyperscaler-hosted distribution path with Anthropic the sole frontier-class vendor excluded. **Transactions:** - **Apr 29.** Microsoft FY26 capex guide $190B (+61% YoY); commercial RPO $627B (+99%) — $190B FY26 capex _(Microsoft FY26 Q3 press release)_ - **Apr 29.** Alphabet raises 2026 capex to $180-190B; Google Cloud backlog $460B (~2x QoQ) — $180-190B 2026 capex _(Alphabet Q1 2026 8-K (SEC EDGAR))_ - **Apr 29.** Meta raises 2026 capex to $125-145B; cuts 8,000 roles to offset infrastructure spend — $125-145B 2026 capex _(Meta Q1 2026 release)_ - **Apr 29.** Amazon Q1 capex $44.2B (+76% YoY); reaffirms $200B 2026 AI investment target; AWS +28% — $200B 2026 target _(Amazon Q1 2026 release / Reuters)_ - **May 1.** Microsoft, AWS, Google, Oracle each win slots in Pentagon eight-vendor classified-network AI slate (IL6/IL7-class) — Undisclosed (federal IDIQ-class) _(war.gov, oracle.com PR, AP News, Breaking Defense)_ ### Neoclouds — detail April was a record month for neocloud demand pre-commits. The Meta expansion of $21B through Dec 2032 lifts Meta's total CoreWeave commitment to ~$35.2B (existing $14.2B + new $21B) and is anchored by NVIDIA Vera Rubin platform deployments. CoreWeave simultaneously raised $3B convertible debt and $1.75B senior notes to fund the buildout, with planned 2026 capex of $30-35B (more than 2x 2025). Oracle's $16B Michigan project-finance package — the largest single-project-finance transaction in tech infrastructure history, with $14B of bonds placed by BofA — locks in a 1.4 GW OpenAI campus and signals that project-finance structures usually reserved for LNG and toll roads are now being applied to AI compute. Backlog math: $66.8B (Feb) + $21B Meta uplift + Jane Street $6B + Anthropic (undisclosed) puts CoreWeave's revenue backlog likely north of $90B pending the mid-May Q1 print. **Transactions:** - **Apr 9.** Meta expands CoreWeave commit by $21B through Dec 2032 (Vera Rubin) — $21B (lifts Meta total to ~$35.2B) _(CoreWeave 8-K Ex 99.1)_ - **Apr 10.** CoreWeave / Anthropic multi-year compute agreement (Claude development and deployment) — Multi-year, undisclosed _(CoreWeave news release)_ - **Apr 15.** Jane Street signs $6B AI cloud deal + $1B equity in CoreWeave at $109/share — $6B services + $1B equity _(Reuters)_ - **Apr 25.** Oracle closes $16B project-finance package for OpenAI 1.4 GW Michigan campus (BofA-led, $14B bonds) — $16B project finance _(Reuters / industry coverage)_ ### On-Prem / Hybrid — detail On-prem / hybrid had no single Q1 catalyst inside the window, but two confirmations matter: Apple's R&D acceleration to $11.4B (Q2 FY26) signals materially higher silicon and platform investment from a non-traditional hyperscaler, and Amazon's custom-silicon $20B+ run rate (Apr 29) confirms the second-largest hyperscaler is now operating a meaningful merchant-equivalent ASIC business. Sovereign AI commitments held at 8 deals / ~$80B+ (UAE Stargate first phase still tracking Q2-Q3 2026 / 200 MW; full 1 GW by 2028). The NVIDIA Q4 FY26 data center print of $62.3B remains the most recent hard datapoint pending the Q1 FY27 release expected late May. **Transactions:** - **Apr 29.** AWS custom silicon (Graviton/Trainium/Nitro) >$20B annual run rate; OpenAI 2 GW + Anthropic up to 5 GW Trainium commitments — >$20B run rate _(Amazon Q1 2026 / Reuters)_ - **Apr 30.** Apple Q2 FY26 record R&D $11.4B (+34% YoY); Mac Studio / mini sold out on AI / agentic demand — $11.4B R&D quarter _(Apple Q2 FY26 / 9to5Mac, MacRumors)_ ## Signal vs noise - **Score 5/5 —** Big-4 hyperscaler 2026 capex resets to a combined $695-725B (+77% YoY): MSFT $190B (vs $154.6B consensus), GOOG $180-190B, META $125-145B, AMZN $200B reaffirmed — all four CFOs repeated 'compute constrained.' - _Sources:_ Microsoft FY26 Q3 press release (Apr 29), Alphabet Q1 2026 8-K (SEC EDGAR, Apr 29), Meta Q1 2026 release (Apr 29), Amazon Q1 2026 release (Apr 29); rolled up in Business Insider, Tom's Hardware, Reuters - _Read:_ Real and load-bearing. The capital-flow bear thesis ('hyperscalers will pull back') is gone for 2026. Read changes only if Q2 prints walk back the guides or if a regulator constrains a major utility / grid project; otherwise expect the $725B floor to hold and stress to migrate to power siting and supply-chain memory pricing. - **Score 5/5 —** Microsoft AI annual run rate is $37B (+123% YoY) and Microsoft Cloud commercial remaining performance obligation is $627B (+99% YoY) as of March 31, 2026. - _Sources:_ Microsoft FY26 Q3 press release (Apr 29, primary 8-K disclosure) - _Read:_ Real. RPO doubling YoY is the cleanest single proof that the AI commercial ramp is durable contract revenue, not pilot spend. Read changes only if Microsoft starts disclosing AI gross margin separately and it deteriorates; the run-rate itself is now an audited number. - **Score 5/5 —** Equinix Q1 2026 reports Fabric revenue +26% YoY, Fabric bookings +70% YoY, +5,800 net interconnections, record 51% adjusted EBITDA margin; full-year guidance raised to 10-11% revenue growth. - _Sources:_ Equinix 8-K (SEC EDGAR, Apr 29), Q1 2026 earnings call transcript - _Read:_ Real. The publication's durable-network thesis got an audited Q1 proof point. Combined with Apr 28-30 open-fabric procurement-ready SKUs (Celestica 1.6TbE, Hedgehog OCP-Accepted, CoreWeave Cross-Cloud), interconnect pricing power compounds rather than compresses through 2026. - **Score 4/5 —** Anthropic disclosed ~$30B annualized revenue in April 2026 (3x in four months; some sources cite ~$40B internal figure), with 1,000+ enterprise customers spending >$1M/yr and 8 of Fortune 10 as Claude customers. - _Sources:_ CNBC, Bloomberg, TechCrunch, Reuters all sourcing the disclosure to Anthropic in connection with $900B raise talks; Sacra and AI Business Review confirm trajectory; Anthropic is private, no SEC filing - _Read:_ Real, but private-company disclosure — not audited. Read changes if the $50B round prices materially below $900B, which would imply investor pushback on the revenue durability or margin profile. - **Score 2/5 —** Anthropic in talks to raise ~$50B at a $900B pre-money valuation, which would top OpenAI's $852B March valuation. - _Sources:_ CNBC (Apr 29), Bloomberg, TechCrunch — all citing 'sources'; no term sheet signed; board decision expected May - _Read:_ Real that the talks are happening, but the price is not yet committed capital. Treat as sentiment indicator, not balance-sheet event. Read changes when the round prices, anchor lead investor is named publicly, or a major LP walks away — any of which would re-rate the entire frontier-lab valuation curve. - **Score 2/5 —** OpenAI's GPT-6 / 'Spud' is launching imminently (late April through May 25, 2026) with a 2M-token context and 40% performance gains over GPT-5.4. - _Sources:_ Various tech blogs (Remio, FindSkill, BigGo, vgtimes); Polymarket probability collapsed from 93% by Jun 30 to 45%; OpenAI has confirmed pre-training completed Mar 24 but not announced a launch date or final naming - _Read:_ Hype-adjacent. The completed pre-training is real (Altman, Brockman public statements); the launch date and capability claims are speculation. Architects should not pre-commit deployment timelines or budget shifts on this. Read changes only on an OpenAI announcement of GA timing and pricing. - **Score 5/5 —** US Department of War (Pentagon) inks classified-network AI procurement deals with eight vendors (OpenAI, Google, Microsoft, AWS, NVIDIA, SpaceX, Reflection AI, Oracle); Anthropic explicitly excluded on supply-chain-risk grounds with Mythos cited by Pentagon CTO Emil Michael as a separate national security moment. - _Sources:_ war.gov press release, oracle.com PR, AP News, The Verge, CNBC, Breaking Defense, The Register, CNN, Washington Post, SCMP - _Read:_ Real and load-bearing. First public federal-grade vendor-disqualification event of the frontier-AI era; opens federal procurement as a distinct earn-out channel for the eight included and creates a vendor-risk precedent that procurement teams should now ask about during diligence. Read changes only if Anthropic secures an exception or the deal scope shifts in a follow-on disclosure. ## Levers | Metric | Current | Prior | Direction | Threshold | |---|---|---|---|---| | Frontier lab cash position (avg months runway, top 3) | ~30 mo | ~22 mo | up | <18 mo triggers re-rating risk | | Hyperscaler capex / AI revenue ratio (top 4 weighted) | ~5.0-5.2 | ~5.5 | down | >6.0 invites investor pushback at next earnings | | CoreWeave revenue backlog | ~$87B+ (est) | $66.8B | up | Conversion velocity matters more than gross figure | | NVIDIA Q-over-Q data center revenue | $62.3B (Q4 FY26) | $62.3B (Q4 FY26) | flat | — | | Open vs closed gap on SWE-Bench Pro (coding) | Closed +6 to +19pp | Open +0.9 | down | Sustained open lead reshapes enterprise procurement | | Sovereign AI commitments (count / aggregate $) | 8 / ~$80B+ | 8 / $80B+ | flat | — | | PJM 2026/27 capacity auction price ($/MW-day) | $329.17 | $329 | flat | 11x in 24 months — power is the new binding constraint | | Time-to-power, busiest US markets (months) | 36-48 (PJM large >100 MW); ~53 nat'l avg | 36-48 | flat | — | | Cost-per-task, frontier reasoning model | ~$0.05 | ~$0.05 | flat | — | | Custom silicon share of incremental AI compute | ~31% | ~30% | up | >35% materially compresses merchant GPU pricing | **Lever detail:** - **Frontier lab cash position (avg months runway, top 3).** Top 3 frontier labs (OpenAI, Anthropic, Google DeepMind) by disclosed runway, computed from cash on hand divided by trailing-12-month operating burn. Q1 mega-rounds (OpenAI $122B, Anthropic $30B, xAI $20B) plus Amazon's $25B and Google's $40B Anthropic compute commitments materially extended runway; OpenAI alone has multi-year cushion at projected ~$17B/yr 2026 burn. <18 months triggers re-rating risk; <12 months forces consolidation, acquisition, or revenue reset. - **Hyperscaler capex / AI revenue ratio (top 4 weighted).** Top 4 hyperscalers (MSFT, GOOG, META, AMZN) weighted aggregate of total capex divided by AI-attributable revenue. Compressed for the first time in five quarters: aggregate 2026 capex stepped to ~$725B (+77% YoY), but AI revenue is accelerating faster — MSFT AI run rate +123% YoY to $37B, Azure +40%, Google Cloud +63%, AWS +28% (15-quarter high). Ratio crosses the >6.0 threshold inside a quarter if AI revenue growth flinches. - **CoreWeave revenue backlog.** Booked but unrecognized revenue. $66.8B Feb 2026 base + $21B Meta expansion (Apr 9, lifts Meta total to $35.2B through 2032) + $6B Jane Street (Apr 15) + Anthropic multi-year (Apr 10, value undisclosed) — pending CRWV Q1 print mid-May for the audited figure. The metric to actually watch is conversion velocity: how quickly capacity comes online and customers ramp. - **NVIDIA Q-over-Q data center revenue.** Data center segment revenue, sequential Q-over-Q. Carried from W17; Q1 FY27 print expected late May 2026. Captures Blackwell Ultra ramp and start of Vera Rubin sample shipments; Samsung HBM4 second-source confirmation Apr 30 is the upstream signal that protects H2 ramp. Beat or miss against the $62B run-rate sets the slope for H2 hyperscaler capex. - **Open vs closed gap on SWE-Bench Pro (coding).** Coding benchmark differential between top open-weight and top closed model. The W17 'open +0.9' lead has reverted: top closed (Claude Mythos Preview gated, GPT-5.3-Codex public) at 77+ vs top open (Kimi K2.6, GLM-5.1, MiMo V2.5 Pro) at 56-58 on the public board; SEAL 250-turn methodology shows even wider Anthropic lead. Methodology variance is high; the directional flip is the signal — open-weights leadership on coding was a single-model methodology artifact, not a structural inflection. - **Sovereign AI commitments (count / aggregate $).** No new sovereign mega-deal closed in W18; Stargate UAE (G42 / OpenAI / Oracle / NVIDIA, $30B / 5 GW) first phase 200 MW still tracks Q2-Q3 2026, full 1 GW by 2028. Current set: UAE Stargate (1 GW), Germany National DC Strategy, France IA program, Mistral-Sweden, GMI Japan, IndiaAI, UK AI Growth Zones aggregating $38.5B, plus Saudi-PIF AI commitment. - **PJM 2026/27 capacity auction price ($/MW-day).** The 2026/27 BRA cleared July 2025 at $329.17/MW-day, unchanged. The 2027/28 BRA cleared Dec 17, 2025 at $333.44/MW-day (first RTO-wide reliability shortfall in PJM history) — outside the W18 14-day window so not refreshed here, but a leading signal architects should price in for next-cycle siting decisions. - **Time-to-power, busiest US markets (months).** Months from new-load interconnection request to energization. PJM and CAISO both >58 mo for full queue. Microsoft and Alphabet both flagged 'compute constrained' on Apr 29 calls — power is the binding constraint, not silicon, through 2026. The constraint is no longer chips, capital, or land — it is electricity delivery. - **Cost-per-task, frontier reasoning model.** Median cost across the frontier-tier reasoning models for a benchmark complex task. No fresh public benchmark this week, but custom-silicon TCO disclosures (Trainium / TPU / MTIA / Maia) continue to point to 30-60% inference cost reduction vs GPU at hyperscaler scale; pricing has not been fully passed through to API list rates yet. Falling cost expands the addressable workload set rather than contracting demand — the Jevons signature in AI inference. - **Custom silicon share of incremental AI compute.** Approximate share of newly deployed AI compute capacity using custom silicon (TPU, Trainium, MTIA, Maia) versus merchant silicon (NVIDIA, AMD). Amazon disclosed Graviton/Trainium/Nitro >$20B annual run rate at triple-digit YoY (Apr 29); TrendForce projects 27.8% ASIC share for 2026, with Trainium 3 ramp Q2 2026, Microsoft Maia 200 GA on TSMC 3nm, Google TPU v7 Ironwood. NVIDIA share trending from 86% (2024) toward ~75% (2026). Past 35% the compression is material and re-rates merchant gross margins. ## Predictions - **`p7-bedrock-gpt55-anchor` _[software]_ — AWS reports a customer-disclosed Bedrock-resident GPT-5.5 multi-year contract greater than $500M annualized, with the customer named publicly.** - Confidence: 65%. Deadline: By July 31, 2026. - Trigger: AWS customer-win announcements during the Bedrock GPT-5.5 limited preview; OpenAI partnership commentary in Q2 prints. - **`p8-capex-revenue-ratio` _[capital]_ — The top-4 hyperscaler capex / AI revenue ratio drops below 5.0 for Q2 2026, confirming the W18 directional flip.** - Confidence: 60%. Deadline: By August 15, 2026. - Trigger: Q2 2026 hyperscaler earnings releases (late July through mid-August). - **`p9-hbm4-second-source` _[hardware]_ — Samsung HBM4 reaches greater than 25% share of NVIDIA Vera Rubin BOM by Q3 2026 supply data, ending the SK Hynix monopoly on the platform.** - Confidence: 60%. Deadline: By September 30, 2026. - Trigger: TrendForce / SemiAnalysis Q3 BOM data; NVIDIA Q2 FY27 earnings color on memory supply. - **`p10-fabric-second-confirm` _[networking]_ — At least one major colocation operator other than Equinix reports Q2 2026 interconnect / fabric revenue growth greater than 25% YoY.** - Confidence: 65%. Deadline: By August 31, 2026. - Trigger: Q2 earnings prints from interconnect operators; Light Reading and DCD coverage. - **`p11-gated-cyber-revenue` _[software]_ — By Q3 2026, at least two frontier labs publicly disclose a separately-priced 'gated cyber' or 'gated security' SKU with revenue commentary, formalizing capability-gating as a product line.** - Confidence: 60%. Deadline: By September 30, 2026. - Trigger: Q3 earnings disclosures (OpenAI partnership color, Anthropic announcements); vendor security press; AISI / METR cross-vendor evaluations. ### Prior predictions scored - `p1-2gw-customer-funded` _[capital]_ — **PENDING** — At least one frontier lab announces a customer-funded compute commitment greater than 2 GW. — OpenAI Stargate US footprint disclosed at >10 GW (Apr 30) is structural, not a single customer-funded announcement >2 GW. Watch I/O and Build in May. - `p2-f500-on-prem` _[software]_ — **PENDING** — At least one Fortune 500 enterprise discloses an on-prem AI workload greater than $100M annual using open-weight models. — No F500 disclosure of >$100M annual on-prem open-weights workload this week. SWE-Bench Pro lever flipped back to closed lead, which weakens the open-weights procurement catalyst. - `p3-capex-revise` _[capital]_ — **PARTIAL** — Aggregate 2026 hyperscaler capex revises upward by 10% or more from the $700B baseline. — Q1 prints (MSFT $190B, GOOG $180-190B, META $125-145B, AMZN $200B reaffirmed) take 2026 aggregate to $695-725B (+77% YoY) vs the $700B W17 baseline. At/near baseline; +10% revision (~$770B) plausible by Q2 print. Score moves to hit if Q2 takes aggregate above $770B. - `p4-interconnect-outpaces` _[networking]_ — **PARTIAL** — At least one major colocation or interconnect operator reports cross-connect or interconnect revenue growth outpacing compute capacity revenue growth for two consecutive quarters. — Equinix Q1 Apr 29: Fabric revenue +26% YoY, Fabric bookings +70% YoY, raised guidance — first of two quarters needed. Score moves to hit if Q2 confirms the same outpacing, expected early August. - `p5-neocloud-anchor-loss` _[capital]_ — **PENDING** — At least one neocloud loses an anchor tenant or sees backlog growth turn negative quarter-over-quarter. — Opposite signal: CoreWeave added $21B Meta (Apr 9), $6B Jane Street (Apr 15), Anthropic multi-year (Apr 10) — backlog grew, not shrank. Lower confidence on this prediction now. - `p6-custom-silicon-35` _[hardware]_ — **PENDING** — Custom silicon (TPU + Trainium + Maia + MTIA + Granite Rapids AI) reaches 35% of incremental AI compute share, up from ~30% today. — Custom silicon at ~31% (up from ~30%); 35% threshold plausible by July 31 given Trainium $225B forward book, Maia 200 GA in two regions, MTIA expanded. No fresh share data crossed the threshold this week. ## Watchlist - **May 5-20 — Mythos status update from Anthropic + AISI.** AISI's Apr 27-28 sabotage evaluation flagged a 65% reasoning-output discrepancy in Mythos continuation tests. The Pentagon then weighed in publicly on May 1 by excluding Anthropic from its eight-vendor classified-network slate citing Mythos, so the AISI response now also signals federal-channel posture. Anthropic's response — extend gating, lift gating under conditions, or expand Project Glasswing — sets the precedent for capability-gated SKUs as a market category. - **May 5-30 — Anthropic federal-channel response to the Pentagon Mythos exclusion.** The May 1 Pentagon eight-vendor slate explicitly excluded Anthropic on supply-chain-risk grounds with Mythos cited. Anthropic's options — pursue an exception, ship a separate federally-eligible gated SKU, or accept commercial-only positioning — set the precedent for whether capability-gating becomes a federal-procurement disqualifier across vendors and how investors price federal-channel parity into frontier-lab valuations. - **Late May 2026 — NVIDIA Q1 FY27 earnings.** First disclosure on Vera Rubin ramp velocity, Samsung HBM4 second-source effect, and custom-silicon competitive pressure (Trainium $225B book, Maia 200 GA). A beat against the $62.3B Q4 FY26 baseline resets the H2 hyperscaler capex tail; a miss flattens Huang's slope. - **May 5-20 — Anthropic $900B raise terms and lead investor.** Talks reported Apr 29 are not yet committed capital. The price ($900B vs OpenAI's $852B) and the lead investor's identity re-rate the entire frontier-lab valuation curve. A walk-away from the price would be the loudest re-rating signal in a year. - **May 5-15 — Q1 capex revision wave (Apple, Oracle, second-tier hyperscalers).** The big-4 reset 2026 capex to $695-725B; secondary cohort (Apple, Oracle, IBM Cloud, smaller hyperscalers) has not yet revised. A second wave of capex hikes pushes the aggregate above $750B and tests prediction p3 to hit before Q2 prints. - **May 2-9 — Google DeepMind: Gemini 3.1 Pro to GA, or Gemini 3.2 announcement.** Gemini 3.1 Pro Preview has held its AA Index 57 spot since February while OpenAI and Anthropic shipped new flagships. A GA or 3.2 announcement would force a frontier scorecard refresh and potentially retake LMArena top spot from Opus 4.7 Thinking. ## Changelog - First Q1 hyperscaler-earnings-cycle issue. Capital flow reset to Q1 2026 reported numbers (MSFT FY26 Q3, GOOG / META / AMZN calendar Q1). - Hyperscaler capex / AI revenue lever direction reversed: ~5.5 (W17) to ~5.0-5.2 (W18) — first Q-over-Q compression in five quarters. Threshold framing unchanged. - Open vs closed gap on SWE-Bench Pro lever direction reversed: 'Open +0.9' to 'closed +6 to +19pp' (methodology-dependent). The W17 open-weights coding inflection partially reverted; lever detail updated to note methodology variance. - First weekly Pulse companion publishes alongside this issue at /industry/models/2026-W18. - Amended Saturday May 2 to incorporate the May 1 Pentagon eight-vendor classified-network AI procurement deal: added federal earn-out paragraph to bigStory, replaced IBM Granite 4.1 with the Pentagon event in the software lens, added Pentagon transactions to Frontier Labs and Hyperscaler-Hosted capital-flow rows, added Pentagon as a score-5 signal, added Anthropic federal-channel response to watchlist. --- Source of truth: `src/data/industry/weekly/2026-W18.ts`. Canonical HTML: . PDF: .