China’s Six-Network Pipeline Accelerates—Can Compute Infrastructure Stay on Top?

China is accelerating the development of a nationwide computing network and high-quality data sets for AI, with a focus on expanding foundational data resources and improving application support. The 15th Five-Year Plan emphasizes advancing the East Data, West Computing initiative and establishing an integrated nationwide computing power network, with real-world applications already demonstrating progress. While power grids and traditional networks are also receiving significant investment, the computing and telecom networks are being prioritized for digital infrastructure development, suggesting they are likely to lead in terms of new pipeline projects, publicly promoted projects, project starts, and completed investment.

Through the information cutoff, the measurable 2026 indicators point to compute-and-telecom retaining leadership on most of the resolution criteria, with power grids a close second and traditional networks clearly trailing on growth. NBS data for January-July 2026 show national fixed-asset investment down 6.7% and infrastructure investment down 3.6%, yet information-transmission investment up 26.0% and internet-and-related-services investment up 41.3% — the fastest growth of any six-network-related category. Power supply investment grew 16.0% and State Grid completed over 310 billion yuan in H1 (+12.6%), while the traditional group contracted: water management -11.0%, public-facilities management -9.1%, road transport -7.5%. On new pipeline projects, roughly 412 new intelligent-computing-centre projects worth about 893 billion yuan were added in H1 2026 (+72.3% by count, +68.5% by value), against 25 major new water projects and 33,000 water projects in total with completed water investment of 515.1 billion yuan and no YoY gain disclosed. On listed-company orders, the compute chain shows broad-based order and earnings expansion (Inspur H1 net profit +226-288%; most CPO/optical-module names growing), while grid equipment shows the sharpest percentage jump (State Grid UHV tenders of 29.26 billion yuan through July, +674% off a depressed 2025 base). NDRC's 3 September 2026 coordination meeting accelerates project libraries and public promotion for all six networks and puts 15th Five-Year Plan investment above 26 trillion yuan (grids 5+ trillion, underground pipes ~5 trillion, compute 4 trillion direct), so traditional networks will remain larger in absolute yuan — but the resolution compares growth against baselines, and on that basis water, utilities and logistics are the laggard, not the catch-up story. My position: COMPUTE_AND_TELECOM_LEAD, with meaningful probability on POWER_GRIDS_LEAD.

I select COMPUTE_AND_TELECOM_LEAD. This is a forecast, not a resolved fact. As of the cutoff, digital infrastructure has the strongest growth in completed investment and procurement-ready demand: information-transmission investment rose 26.0% through July, while the three major telecom operators sharply increased computing-network spending. Traditional networks have more physical projects, and grids are executing reliably, but their aggregate investment growth currently trails compute. I therefore expect compute and telecom to lead a majority of the measurable indicators by year-end, particularly completed investment, winning orders, and projects advancing from pipeline to procurement.

As of the 6 September 2026 cutoff, power grids show the strongest verifiable movement from pipeline into construction. State Grid's Q1 2026 fixed-asset investment rose 37% YoY to 129.0bn yuan (over 1.8bn yuan/day), 2026 UHV/transmission tenders cumulatively exceeded 36bn yuan by late July/August with 2026 orders already above all of last year, and listed equipment makers (中国西电, 许继电气, 平高电气, 国网信通) repeatedly disclosed large grid wins—国网信通 alone pre-won 2.905bn yuan on 4 September 2026. The compute network is large (4tn yuan over 2026-2030) but earlier-stage: mega data-center projects are announced, yet listed-company compute orders are scattered and some 100,000-card clusters face utilization/overbuild concerns. Traditional networks show volume but not superior growth: H1 water completed investment (515.1bn yuan) is the largest absolute figure but fell roughly 3% YoY, and underground-utility awards are fragmented municipal contracts. On project starts, completed-investment growth and listed-company winning orders—a majority of the resolution indicators—grids lead.

China's accelerated rollout of the 'Six Networks' (水网、新型电网、算力网、新一代通信网、城市地下管网、物流网) infrastructure initiative in mid-2026 reflects a decisive policy push to generate immediate physical work volume ('实物工作量') to support economic growth. Traditional networks—comprising water conservancy, urban underground utility pipelines, and logistics networks—account for over 15 trillion RMB (more than 55%) of the estimated 26 trillion RMB total 'Six Networks' investment under the 15th Five-Year Plan framework. With targeted capital deployment via 800 billion RMB in new policy-based financial tools, 160 billion RMB in ultra-long special treasury bonds allocated for underground pipe networks in 2026, and the NDRC issuing the first formal implementation plan for logistics networks in late August 2026, traditional networks are rebounding sharply from a low 2024–2025 baseline. Consequently, traditional physical infrastructure is projected to lead overall growth across new project pipeline disclosures, project starts, and completed physical investment indicators by late 2026.

Insufficient contemporaneous official data could be retrieved for the listed NDRC six-network pipeline disclosures, so the year-end lead cannot be verified from primary project-level indicators. On the available pre-cutoff policy framing, compute and telecom still appear more likely to lead on pipeline promotion, starts, and listed-company orders through 2026, but this is a medium-confidence inference rather than a completed-investment fact.

The NDRC's acceleration machinery for the Six Networks — project libraries, bank-enterprise financing linkage, and public promotion of projects to social capital (Sept 2-3 coordination meeting) — is structurally aimed at traditional networks (water, underground pipelines, logistics), which depend on government funding and promotion channels, unlike privately-funded compute or SOE-financed grids. On the five measurable indicators, traditional networks likely lead on new pipeline project counts (2,000+ cities compiling pipeline plans; 77万km/5 trillion yuan 15th-FYP pipeline program; water starts in H1 2026 already near 2025's full-year count), publicly promoted projects (the classic social-capital promotion category, backed by 160B yuan in 2026 ultra-long special bonds, +25B YoY), and sheer physical investment volume (urban renewal ~3T yuan/yr). Power grids are the strongest rival, with State Grid H1 investment +12.6% and UHV orders already exceeding full-year 2025, and the compute+telecom group shows the fastest growth rates (carrier compute capex +62-97%, national intelligent compute 2.5x YoY) but is diluted by declining overall telecom capex (-9.5% at China Mobile) and weaker visibility in disclosed project/order channels. On a majority-of-indicators basis, traditional networks catching up is the modal outcome, though the margin over power grids is thin.

As of September 2026, China's 'Six Networks' infrastructure initiative is accelerating across all sectors, but the evidence strongly indicates that **compute and telecom networks are leading in project materialization, investment growth, and policy prioritization**. The assessment is based on the following measurable indicators: (1) disclosed new pipeline projects, (2) publicly promoted projects, (3) project starts, (4) completed investment, and (5) listed-company winning orders. The compute and telecom sectors are demonstrating the fastest materialization and growth, supported by direct investment commitments, policy coordination mechanisms, and tangible physical indicators such as 5G base stations, computing power capacity, and ultra-high-voltage (UHV) project awards. While power grids and traditional networks (water, logistics, underground utilities) are also receiving substantial investment, their progress is either more gradual or focused on incremental upgrades rather than transformative growth.

POWER_GRIDS_LEAD is the most likely resolution. By the cutoff, the grid segment has the clearest combination of a large forward pipeline, rapid physical execution, completed investment growth, and listed-company orders: State Grid reported more than RMB310 billion of first-half fixed-asset investment, up 12.6% year on year, while it had already started 160,000 upgrade and construction initiatives. Its 2026–2030 investment plan is RMB4 trillion, about 40% above the prior five-year period. [web:75][web:76][web:79] Compute has a substantial RMB4 trillion five-year direct-investment expectation and strong bidding activity, but its disclosed project and investment evidence is narrower; communications infrastructure is comparatively mature, with 5G coverage and fiber deployment already extensive. [web:59][web:69][web:126] Traditional networks are a serious challenger: water conservancy completed RMB515.1 billion in first-half investment and started 25 major projects, while underground pipelines have a large five-year physical target and RMB160 billion of 2026 special-treasury-bond support. [web:89][web:46] Nevertheless, the grid has the strongest cross-indicator evidence and the clearest listed-company order flow, including multi-billion-yuan State Grid awards to China XD and other listed equipment suppliers. [web:109][web:117] The NDRC’s September coordination meeting accelerates project databases, bank linkage, and public promotion across all six networks, but does not yet disclose a comparable sector-by-sector scoreboard; therefore this is a forecast rather than a definitive count-based resolution. [web:0]

As of September 2026, the 'Six Networks' initiative is driving massive infrastructure investment, but Power Grids are demonstrating the most immediate and tangible construction momentum. The National Development and Reform Commission (NDRC) highlighted that investment in the upgraded power grid alone is expected to exceed 5 trillion yuan during the 15th Five-Year Plan (2026-2030), with significant projects like the Gansu-Zhejiang UHVDC line already under active construction in early 2026 [2][19]. While compute and telecom networks are critical for AI development, their growth is characterized by rapid technological iteration and service expansion (e.g., 5G-A, AIDCs) rather than the sheer volume of new physical pipeline starts seen in the energy sector. Traditional networks (water, underground logistics) are part of the plan but lack the same scale of disclosed, high-value project commencements and state-backed financing urgency as the power grid, which is essential for stabilizing the energy supply amidst the renewable transition [2][20]. Therefore, Power Grids lead in terms of visible construction activity and committed capital deployment in the near term.

As of September 2026, verified NDRC disclosures show compute and telecom networks already have far above baseline growth metrics, with intelligent capacity at 2.5 times the prior year level, rapid 5G-A deployment, large year-over-year investment surges, and a mature national spatial layout making near-term project execution and order materialization the likeliest leading outcome against power grids and traditional water, utility and logistics networks.

As of the Sep-6 cutoff, NDRC disclosures show compute and telecom with the sharpest, most quantifiable momentum—intelligent compute capacity 2.5x YoY (188.2 EFLOPS) and internet-services investment +39.9% in H1—and explicitly frame them as the 'new engines'/'new quality productive forces', while traditional infra is described as having diminishing marginal returns. The new power grid's flagship 15th FYP projects are themselves compute-enabling (green-power direct supply to seabed/coastal compute centers). On the resolution indicators, compute-telecom leads most clearly on growth and narrative, so it should still dominate, though hard order/investment-by-group data for year-end is not yet published.

China's six-network push is being layered on top of an extraordinary AI capex supercycle. Since early 2025, hyperscalers (Alibaba alone committed at least RMB 380 billion over three years to cloud and AI infrastructure, with Tencent and ByteDance following), the carriers' shift of capex from 5G radio into compute, and the East-Data-West-Computing hub build-outs have produced the fastest order momentum of any network group — flowing to IDC builders, server, optical-module, liquid-cooling and data-center power-equipment suppliers. Grid investment is large and rising (State Grid capex moving from roughly RMB 600 billion toward RMB 650 billion) but grows at high single digits, while water, underground-utility and logistics pipelines rely more on special-bond funding with slower, more diffuse order conversion to listed companies. On growth-based pipeline and order indicators through the resolution window, compute should dominate even though grids may lead on absolute completed investment.

Compute and telecom networks will lead China's Six-Network Pipeline implementation by December 2026, driven by unprecedented investment commitments, rapid project materialization, and exceptional listed-company order growth that outpaces power grids and traditional networks across most measurable indicators.