In the modern AI pipeline, the CPU no longer merely supports the model — it drives the model (agentic workloads invert historical CPU:GPU ratio).
Predictor: Jensen Huang
Prediction text
In the modern AI pipeline, the CPU no longer merely supports the model — it drives the model (agentic workloads invert historical CPU:GPU ratio). | the CPU no longer merely supports the model; it drives the model | Server CPU TAM growth rate 2026-2030
Key catalyst: Server CPU TAM growth rate 2026-2030
Watch events: Server CPU market growth vs GPU; NVIDIA Grace adoption; Intel/AMD agentic-optimized CPU launches.
Verbatim quote
the CPU no longer merely supports the model; it drives the model
Resolution evidence
Georgia Tech/Intel research: CPU = 50-90% of end-to-end latency in Agentic AI workloads; NVIDIA Grace CPU development validates.
Predictor: Jensen Huang
Calibration plot (stated vs observed)
Evidence about this node from Jensen Huang is multiplied by κ in /api/intake. Lower κ = less weight; floors at 0.10 (effectively silenced) and caps at 1.00 (full weight).
Reference class
This node isn't linked to a reference class. The Bayesian update applies without outside-view blending.
Probability over time
Milestone chain
- 2026-02-15hitNVIDIA standalone CPU deployed in Meta data centers signals new infrastructure categoryHow: NVIDIA announces standalone (non-paired-with-GPU) CPU deployment in Meta data centers for agentic workloadsSource: CNBC — Nvidia GTC AI pivot CPU center stageconf 90%
- 2026-03-07overdueQ1 window check-in (25%)
- 2026-03-25hitArm publishes AGI CPU thesis confirming CPU renaissance for agentic AIHow: Arm or major industry analyst publishes formal AGI CPU thesis treating CPU as essential agentic AI infrastructureSource: Shashi.co — Arm Bets on Silicon: The AGI CPU and the CPU Renaissanceconf 85%
- 2026-03-18hitJensen Huang at GTC publicly frames CPU as central to agentic AIHow: Jensen Huang at GTC 2026 keynote publicly frames CPU as core to agentic AI infrastructure (not merely supporting GPU)Source: CNBC — Nvidia GTC AI pivot CPU center stage (March 2026)conf 95%
- 2026-04-20hitMorgan Stanley publishes Rise of AI Agents thesis: GPU bottleneck shifts to CPUHow: Morgan Stanley publishes major research report on agentic AI shifting value from GPUs to CPUs/memory with $32.5-60B incremental CPU TAM by 2030Source: Morgan Stanley research / Bitget News — Rise of AI Agents Global Impact (April 2026)conf 99%Notes: HIT — Morgan Stanley published April 2026 report explicitly stating CPU bottleneck for agentic, $32.5-60B incremental TAM. Direct cross-reference.
- 2026-05-12overdueQ2 window check-in (50%)
- 2026-07-17pendingQ3 window check-in (75%)
- 2026-12-01 → 2027-12-31pendingCPU:GPU ratio in agentic workloads exceeds 1:1 cluster-level by 2027How: Hyperscaler procurement disclosures or analyst reports confirm CPU:GPU ratio ≥1:1 at cluster level for agentic AI deploymentsSource: AI2.work — Why Agentic AI Is Making CPUs the New Bottleneckconf 60%Notes: Cascade — CPU-side orchestration is 50-90% of agentic workload latency per Morgan Stanley.
- 2027-01-01 → 2030-12-31pendingServer CPU TAM grows ≥15% annually 2026-2030 per Morgan Stanley estimateHow: Annual server CPU TAM growth rate averages ≥15% across 2026-2030 per Morgan Stanley/IDC trackingSource: Morgan Stanley April 2026 report — $32.5-60B incremental CPU TAM by 2030conf 70%Notes: Direct measurement of source's quoted criterion.
What if this resolves?
Click a button to clamp this prediction and run a Gibbs sample. Returns the predictions whose marginals shift most. ~30s per run; ideal for stress-testing "if X resolves, what else moves?"
Evidence chain
Raw metadata
{
"trf": 0.5881123843731557,
"kappa": 0.8077,
"base_rate": null,
"predictor": "Jensen Huang",
"total_llr": -0.4054651081081644,
"grace_days": 7,
"bayesian_v2": true,
"prior_logit": -0.3494317766421371,
"bayes_factor": "1.4:1 against",
"blend_reason": "no reference_class linked",
"inside_prior": 0.413520220545267,
"kappa_source": "predictor_table",
"n_milestones": 1,
"blend_applied": false,
"contributions": [
{
"llr": -0.4054651081081644,
"kind": "quartile_checkpoint",
"kappa": 0.8077,
"label": "Q2 window check-in (50%)",
"weight": 0.05,
"strength": "weak",
"confidence": null,
"source_url": null,
"adjusted_llr": -0.32749416781896434,
"expected_date": "2026-05-12",
"measurement_criterion": null
}
],
"evidence_kind": "metadata_milestone_miss_sweep",
"inside_source": "history_v2",
"inside_weight": 0.5883213309387909,
"outside_weight": 0.4116786690612091,
"posterior_prob": 0.3369477449161395,
"posterior_logit": -0.6769259444611014,
"predictor_brier": 0.01276,
"inside_posterior": 0.3369477449161395,
"blended_posterior": 0.3369477449161395,
"reference_class_id": null,
"total_adjusted_llr": -0.32749416781896434,
"predictor_n_resolved": 8
}Raw metadata
{
"trf": 0.6650500679109432,
"kappa": 0.8077,
"base_rate": null,
"predictor": "Jensen Huang",
"total_llr": -0.4054651081081644,
"grace_days": 7,
"bayesian_v2": true,
"prior_logit": 1.265666373331276,
"bayes_factor": "1.4:1 against",
"blend_reason": "no reference_class linked",
"inside_prior": 0.78,
"kappa_source": "predictor_table",
"n_milestones": 1,
"blend_applied": false,
"contributions": [
{
"llr": -0.4054651081081644,
"kind": "quartile_checkpoint",
"kappa": 0.8077,
"label": "Q1 window check-in (25%)",
"weight": 0.05,
"strength": "weak",
"confidence": null,
"source_url": null,
"adjusted_llr": -0.32749416781896434,
"expected_date": "2026-03-07",
"measurement_criterion": null
}
],
"evidence_kind": "metadata_milestone_miss_sweep",
"inside_source": "prior_prob",
"inside_weight": 0.5344649524623397,
"outside_weight": 0.4655350475376603,
"posterior_prob": 0.7187303036352096,
"posterior_logit": 0.9381722055123117,
"predictor_brier": 0.01276,
"inside_posterior": 0.7187303036352096,
"blended_posterior": 0.7187303036352096,
"reference_class_id": null,
"total_adjusted_llr": -0.32749416781896434,
"predictor_n_resolved": 8
}Network propagation neighbors
No propagation data yet. Run inference/.venv/bin/python scripts/ops/run_loopy_belief_propagation.py on the droplet, or wait for the Sunday 02:00 UTC weekly cron.
Ticker exposure
Beneficiaries (22)
Prerequisites (1)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| correlate | S_COMPUTE_100GW_2030 | Compute: 100GW national-scale by Dec 2030 | compute_scale | — |
Dependents (0)
| Type | Pred | Title | Domain | Lag |
|---|---|---|---|---|
| No dependents | ||||
Validations (1)
| Observed at | Status | By | Notes |
|---|---|---|---|
| 2026-04-29 | partial | thesis_timeline_v1.0_import | Georgia Tech/Intel research: CPU = 50-90% of end-to-end latency in Agentic AI workloads; NVIDIA Grace CPU development validates. |
Linked documents (10)
| Sim | Source | Title | Market prob | Polarity | Reviewed | Published |
|---|---|---|---|---|---|---|
| 0.667 | github_release | tensorflow/tensorflow v2.17.0 | — | mentions | pending | 2024-07-11 |
| 0.662 | github_release | tensorflow/tensorflow v2.17.0-rc0 | — | mentions | pending | 2024-06-18 |
| 0.660 | github_release | tensorflow/tensorflow v2.17.0-rc1 | — | mentions | pending | 2024-07-02 |
| 0.649 | arxiv | AMDP: Asynchronous Multi-Directional Pipeline Parallelism for Large-Scale Models Training | — | mentions | pending | 2026-05-28 |
| 0.623 | github_release | tensorflow/tensorflow v2.16.1 | — | mentions | pending | 2024-03-07 |
| 0.615 | github_release | tensorflow/tensorflow v2.15.0 | — | mentions | pending | 2023-11-14 |
| 0.612 | github_release | tensorflow/tensorflow v2.21.0-rc0 | — | mentions | pending | 2026-02-09 |
| 0.612 | github_release | tensorflow/tensorflow v2.13.1 | — | mentions | pending | 2023-09-26 |
| 0.610 | github_release | tensorflow/tensorflow v2.21.0-rc1 | — | mentions | pending | 2026-03-02 |
| 0.610 | github_release | tensorflow/tensorflow v2.15.0-rc0 | — | mentions | pending | 2023-10-25 |
Raw metadata
{
"nia": false,
"mode": "THESIS",
"role": "Cited-Executive",
"context": "Huang 2026 observation validates Morgan Stanley 'Rise of the AI Agent' thesis; agentic workflows demand sequential CPU logic for scheduling/orchestration.",
"to_year": 2030,
"verbatim": "the CPU no longer merely supports the model; it drives the model",
"conv_cues": "declarative; CEO",
"direction": "HAPPEN",
"from_year": 2026,
"timeframe": "2026+",
"conv_level": "HIGH",
"milestones": [
{
"kind": "llm_pre_event",
"label": "NVIDIA standalone CPU deployed in Meta data centers signals new infrastructure category",
"source": "CNBC — Nvidia GTC AI pivot CPU center stage",
"status": "hit",
"weight": 0.4,
"ordinal": -7,
"source_id": null,
"confidence": 0.9,
"source_url": "https://www.cnbc.com/2026/03/13/nvidia-gtc-ai-jensen-huang-cpu-gpu.html",
"expected_date": "2026-02-15",
"observed_date": "2026-02-15",
"research_origin": "deep_research",
"measurement_criterion": "NVIDIA announces standalone (non-paired-with-GPU) CPU deployment in Meta data centers for agentic workloads"
},
{
"kind": "quartile_checkpoint",
"label": "Q1 window check-in (25%)",
"status": "overdue",
"weight": 0.05,
"ordinal": -6,
"source_id": null,
"expected_date": "2026-03-07",
"observed_date": null,
"miss_emitted_at": "2026-05-02T22:07:21.384228+00:00",
"miss_emitted_by": "metadata_milestone_sweep"
},
{
"kind": "llm_pre_event",
"label": "Arm publishes AGI CPU thesis confirming CPU renaissance for agentic AI",
"source": "Shashi.co — Arm Bets on Silicon: The AGI CPU and the CPU Renaissance",
"status": "hit",
"weight": 0.4,
"ordinal": -5,
"source_id": null,
"confidence": 0.85,
"source_url": "https://www.shashi.co/2026/03/arm-bets-on-silarm-bets-on-silicon-agi.html",
"expected_date": "2026-03-25",
"observed_date": "2026-03-25",
"research_origin": "deep_research",
"measurement_criterion": "Arm or major industry analyst publishes formal AGI CPU thesis treating CPU as essential agentic AI infrastructure"
},
{
"kind": "llm_pre_event",
"label": "Jensen Huang at GTC publicly frames CPU as central to agentic AI",
"source": "CNBC — Nvidia GTC AI pivot CPU center stage (March 2026)",
"status": "hit",
"weight": 0.4,
"ordinal": -4,
"source_id": null,
"confidence": 0.95,
"source_url": "https://www.cnbc.com/2026/03/13/nvidia-gtc-ai-jensen-huang-cpu-gpu.html",
"expected_date": "2026-03-30",
"observed_date": "2026-03-18",
"research_origin": "deep_research",
"expected_date_range": {
"to": "2026-04-15",
"from": "2026-03-15"
},
"measurement_criterion": "Jensen Huang at GTC 2026 keynote publicly frames CPU as core to agentic AI infrastructure (not merely supporting GPU)"
},
{
"kind": "llm_pre_event",
"label": "Morgan Stanley publishes Rise of AI Agents thesis: GPU bottleneck shifts to CPU",
"notes": "HIT — Morgan Stanley published April 2026 report explicitly stating CPU bottleneck for agentic, $32.5-60B incremental TAM. Direct cross-reference.",
"source": "Morgan Stanley research / Bitget News — Rise of AI Agents Global Impact (April 2026)",
"status": "hit",
"weight": 0.4,
"ordinal": -3,
"source_id": null,
"confidence": 0.99,
"source_url": "https://www.bitget.com/amp/news/detail/12560605375963",
"expected_date": "2026-04-20",
"observed_date": "2026-04-20",
"research_origin": "deep_research",
"measurement_criterion": "Morgan Stanley publishes major research report on agentic AI shifting value from GPUs to CPUs/memory with $32.5-60B incremental CPU TAM by 2030"
},
{
"kind": "quartile_checkpoint",
"label": "Q2 window check-in (50%)",
"st
... (truncated)