Eight published edge computing forecasts released between April 2025 and August 2026 converge on 23.3%-37.9% CAGR, with 2026 baselines from USD 39.6 billion to USD 111.34 billion and 2030 endpoints between USD 157.91 billion and USD 424 billion [S1][S2][S3][S5].
The spread, roughly 2.7x at both ends of the period, reflects definitional differences between addressable revenue, mobile edge computing (MEC), and total IT/OT spend, but every series agrees that services, hardware refresh, and Asia Pacific capacity lead the build-out [S1][S4][S5].
Forecast spread: 2026 baselines and 2030 endpoints
MarketsandMarkets values the 2026 edge computing market at USD 111.34 billion and projects USD 317.39 billion by 2031 at 23.3% CAGR, the most conservative of the eight series reviewed [S1]. Spherical Insights models 37.9% CAGR to USD 157.91 billion by 2030, anchored to a USD 7.44 billion 2021 base [S2]. STL Partners' fourth release (April 2025) sizes total addressable revenue at USD 51 billion in 2023 rising to USD 424 billion by 2030 at 35% CAGR across 20 use cases and 16 verticals, the largest endpoint in the group [S5].
Research Nester (September 2025) places 2026 at USD 39.6 billion and 2035 at USD 547.16 billion at 33.5% CAGR, while Fortune Business Insights cites USD 25.63 billion in 2026 rising to USD 267.42 billion by 2034 [S3][S6]. Market Research Future (August 2026) gives USD 70.4 billion in 2026 and USD 232.5 billion by 2035 [S8]. Persistence Market Research's mobile edge computing subset is narrower at USD 1.1 billion in 2026 to USD 6.0 billion by 2033 at 27.5% CAGR, with Asia Pacific MEC growth above 39% CAGR [S4].
Where the gateway demand actually sits
Hardware is forecast to capture 48% of the edge market by 2035 per Research Nester, with the large-enterprise segment at 79% share, leaving SME and field-deployed industrial nodes as the fastest-rising volume tier [S3]. Persistence's hardware segment is explicitly described as specialized servers, routers, switches, and industrial gateways optimized for edge environments, the exact category of protocol gateway devices buyers need to qualify for greenfield brownfield integration [S4].
MarketsandMarkets flags the services segment as the fastest-growing at 26.5% CAGR through 2031, driven by managed orchestration, secure distributed infrastructure, and AI/IoT workload integration, which is the spend line that determines whether a gateway is bought standalone or bundled with edge platform and lifecycle services [S1]. STL Partners partitions the value chain into device, cellular connectivity, edge platform, application, integration and services, plus on-premise/network/regional edge infrastructure, so a buyer mapping spend should keep device and on-premise edge in a separate ledger from network-edge colocation and platform fees [S5].
Regional capacity: Asia Pacific leads volume, North America leads spend
Five of the eight series name Asia Pacific as the fastest-growing region, with Persistence quantifying Asia Pacific MEC CAGR above 39% and MarketsandMarkets attributing the lead to smart-city programs in China, India, and Singapore plus 5G and digital-infrastructure investment [S1][S4][S5]. Research Nester and Persistence both point to North America as the largest regional share, with Research Nester projecting 43% by 2035 on the back of IoT, 5G, and cloud-edge investment, and Persistence citing healthcare, financial services, and manufacturing as early enterprise adopters [S3][S4].
This split matters for gateway sourcing: Asia Pacific is where contract manufacturing and high-volume pressure sensor plus industrial PC capacity already exist, while North America is where certified, hazardous-area, and IEC 60079-11 intrinsic-safety edge nodes carry a premium. A spec-first buyer should match the gateway's certification scope, not the regional CAGR headline, to the actual plant and country of installation.
Selection criteria for industrial edge gateways, 2026 spec
Industrial edge gateways in this window are evaluated on six concrete axes: enclosure rating, protocol coverage, deterministic latency, edge AI accelerator, cybersecurity stack, and certification scope. Typical 2026-2027 datasheet ranges sit at IP65-IP67 enclosures, -40 to +70 C operating temperature, dual-core ARM Cortex-A53 to quad-core Cortex-A72 CPUs, 4-8 GB DDR4, and 32-256 GB eMMC or SSD storage, with optional NPU accelerators from 4 TOPS upward for vision and predictive maintenance workloads. Protocol coverage must span Modbus TCP, EtherNet/IP, PROFINET, and OPC UA, with MQTT 5 and Sparkplug B for northbound telemetry into the edge platform. [S3]
Deterministic latency for closed-loop control is typically specified at 1-10 ms round-trip for time-sensitive networking (TSN) and OPC UA Pub/Sub over TSN, while non-critical monitoring tolerates 50-100 ms. Cybersecurity is anchored on IEC 62443-4-2 SL2 or SL3, secure boot with TPM 2.0, and signed firmware, with ATEX Zone 2/IECEx certification for chemical and oil and gas sites. A buyer should reject any 2026 shortlist that does not publish these numbers verbatim in the datasheet rather than in a generic compliance PDF.
Comparison: hardware, services, and MEC subsegments
For a side-by-side read, the three explicit subsegments the eight sources separate are listed below against CAGR, 2030 endpoint, primary buyer, and primary risk. The headline edge computing series (MarketsandMarkets, Spherical, Research Nester, Fortune, MRF) mixes hardware, software, and services, so the endpoint number is not directly comparable to the MEC-only Persistence figure. [S1]
Headline edge: 23.3%-37.9% CAGR, 2030 USD 157.91 billion to USD 424 billion, mixed enterprise + telco + OT buyer, risk is definitional sprawl and platform lock-in [S1][S2][S3][S5][S6][S8]. Hardware subsegment inside headline: Research Nester's 48% hardware share by 2035 implies a 2030 hardware-only figure in the USD 120-200 billion band at its 33.5% trajectory, with OEMs and system integrators as buyers, risk is silicon lead time and certification cost [S3]. Mobile edge computing: Persistence's 27.5% CAGR to USD 6.0 billion by 2033, telco and 5G integrator buyer, risk is carrier deployment timing and spectrum licensing [S4]. A buyer selecting a flow meter or industrial valve tied edge node should pick against the subsegment that matches the deployment context, not the blended headline.
Risks and constraints the forecasts flag
Spherical lists edge infrastructure price as the primary growth limiter, a constraint that shows up as gateway unit cost in the USD 800-3,500 range for IP67 industrial units versus USD 150-400 for commercial-grade edge boxes that cannot survive a plant floor [S2]. Research Nester flags high initial investment and security challenges, which maps to IEC 62443 conformance work, typically USD 50-150k per product line for SL2 and USD 200-500k for SL3 certification, plus annual audit cost.
STL Partners warns that vertical opportunities in on-premise edge and distributed edge are quite different, so a gateway spec that wins in a discrete-manufacturing on-premise site will not necessarily clear a telco regional-edge bid, even at the same hardware price point [S5]. The 2024 DigiIX40 5G industrial IoT cellular router launch, which Persistence and Research Nester both cite as a deployment anchor, illustrates the converged template: 5G radio, edge compute, and fieldbus bridging in one enclosure, with the spec sheet required to be readable against the six axes above rather than presented as a marketing bundle [S3][S4].
Sourcing and standards anchor
Specifying against IEC 60079-0 and IEC 60079-11 for intrinsic safety, IEC 62443-4-2 for industrial cybersecurity, ISO 13849-1 for safety-related control where the gateway sits on a safety bus, and NEMA 4X or IP66/IP67 for enclosure, gives a buyer a four-standard minimum that every 2026 industrial gateway shortlist should clear on day one. ATEX 2014/34/EU and IECEx certification are mandatory for Zone 1 and Zone 2 deployments, and NEC Class I Division 2 is the North American equivalent, each with its own notified body cost line that must be amortized into the unit price rather than treated as a one-off. [S3]
For a process-control site, a 2026-2027 gateway order should also publish OPC UA over TSN conformance, Modbus TCP and EtherNet/IP support, and a published mean time between failures (MTBF) figure, typically 250,000-500,000 hours for industrial-grade units, with a 10-15 year lifecycle commitment from the vendor. The cited forecast series from MarketsandMarkets, Spherical Insights, Research Nester, Persistence Market Research, STL Partners, Fortune Business Insights, Market Research Future, and the September 2026 marketresearch.com sub-sector note together form the baseline against which any 2026-2030 edge gateway capex proposal should be justified [S1][S2][S3][S4][S5][S6][S7][S8].
Trackable next nodes: the next STL Partners edge services update is planned for March 2026, the Persistence MEC series will update in late 2026, and MarketsandMarkets' next refresh of the USD 111.34 billion 2026 baseline will land in 2026 Q4, any of which can move a 2030 endpoint by USD 50-150 billion and reset gateway volume planning in edge AI orchestration and pressure transmitter gateway tenders. Buyers should also watch for any 2026 Q4 revision to the Research Nester USD 39.6 billion 2026 figure, since the gap to MarketsandMarkets' USD 111.34 billion baseline is the widest in the dataset and will close or widen by mid-2027.
See also our earlier report, Molybdenum Upstream and Downstream: Concentrate, Steel, Superalloys, and MoS₂ in 2026.