Nvidia Rack

2026 NVIDIA AI Outlook: From GPU to LPU Racks & Inference | TrendForce

www.trendforce.com

NVIDIA's Product Roadmaps to Segment into GPUs, CPUs, and LPUs when Targeting the AI Market · Vera Rubin Rack Expected to Gradually Replace GB300 after Release in 2H26 · NVIDIA's Projected Quarterly Shipment of GB/VR Racks between 2025 and 2026 · NVIDIA Also Proposed Vera CPU to Strengthen Agentic AI Infrastructures · Overview of NVIDIA's Chip Generations and Rack Interconnect Architectures · Conclusion: Facing the Challenge of Rising Market Share for CSPs' In-House ASICs, NVIDIA Pushes Integrated Rack Solutions Across CPUs, GPUs, and LPUs to Accelerate Its Expansion from AI Training into Inference NVIDIA Leverages CPO to Enhance High-Speed Interconnect Performance of Rack-Scale AI Chip Solutions and Whole Rack Systems

2026 Global Data Center Trends: Rack Solutions | TrendForce

www.trendforce.com

This report focuses on NVIDIA data-center growth and rack-based solutions, highlighting large architectures and liquid cooling as core innovation drivers, with regional markets and risk management noted. NVIDIA data-center growth drives rack-based solutions and platform development. Liquid cooling remains central to heat management and efficiency. CSPs and edge applications push supply chain and cost-structure changes. Regional deployment and risk management are long-term priorities. ... NVIDIA Is Expected to Rely on CSPs to Sustain High Growth in 2026; With China Still Blocking Imports of H20, NVIDIA Will Shift Focus to HGX and Other Models to Address Some Inventory Issues NVIDIA's Roadmap for Main AI Server Offerings and Shipment Modes for Complete Racks · Iteration from Blackwell to Rubin in 2026 Expedites Penetration of Liquid-Cooling and Magnifies Liquid-Cooling Adoption among AI Servers · Liquid-Cooling Adoption among AI Chips Set to Rise with the Rollout of GB300/VR200 Servers · HBM Shipment Expected to See Significant Growth in 2026 Driven by R100 and GB200/GB300 · Penetration Rate of HBM Products in 2026 (Before and After) Major Sources of Demand for GB and VR Racks Will Come from Major CSPs and Tier-2 Operators like NeoCloud in 2026 · NVIDIA Will Focus on Major CSPs with GB and VR Racks in 2026, While HGX and MGX Will Target Customers in Edge AI Segment

NVIDIA Research & Analysis | TrendForce

www.trendforce.com

Diverging HBM4 certification timelines are reshaping 2026 market share, with Samsung gaining ground as SK hynix faces delays. ... NVIDIA advances its 800V Power Rack roadmap, with mass adoption expected from Rubin Ultra onward, while North American data center buildouts face delays from grid and equipment constraints. AI compute surges in 2026 as ASICs rise, NVIDIA defends its ecosystem, and bandwidth becomes the bottleneck. ... Hyperscaler capex stays strong, lifting AI server growth, while tight supply now shapes shipment pace. ... AI demand keeps HBM shipments strong as technology advances and tight supply shifts pricing power toward suppliers. ... NVIDIA's data center strength lifts GB/VR Rack demand, extending its reach into infrastructure and HBM supply. Global AI server momentum builds as CSPs expand rack buys, hyperscalers advance in-house chips, and OEM orders grow. ... During the first half of 2026, surging demand for KV Cache coupled with constrained memory supply resulted in severe memory bottlenecks. Strong CSP capex fuels steady AI server growth as cloud giants expand GPU rack orders and self-developed chip programs.

NVIDIA’s AI Rack Transition Accelerates; NVL72 Output Value to ...

www.trendforce.com

TrendForce’s recent AI server industry report reveals that GB300 rack-scale solutions are set to be NVIDIA’s leading driver of shipments in 2026, with demand anticipated to remain robust through the first half of 2027. Following that, the next-gen VR200 racks will increase production and account for the largest portion of shipments in 2027. By incorporating their proprietary chips into financing arrangements, the companies are seeking to secure demand for their products over the longer term. For more information on TrendForce’s semiconductor reports and market data, please visit the Report Page, leave a Message, or Email (SR_MI@trendforce.com) the Sales Department. For more on the latest technology industry news and trends, please visit News. ... Secondary and Repair Markets Provide Support; Global Smartphone Panel Shipments to Decline Just 2.5% in 2026, Says TrendForce The customer mix for rack-scale solutions is also shifting. Although the five major North American CSPs remain the primary buyers, their share of procurement is expected to decline from approximately 70% in 2025 to 60% in 2026. TrendForce estimates that shipments of NVL72 racks, including GB and VR platforms, will grow by more than 50% YoY in 2027. At the same time, the ASP of the Vera Rubin systems is expected to be roughly double that of GB300, while prices for key upstream components such as wafers and HBM could also rise. As a result, the combined output value of GB300, VR200, and VR300 NVL72 rack systems is projected to exceed US$710 billion in 2027, surging 214% YoY.

Rack-Scale GPU Infrastructure Market Size, Share & 2031 Growth ...

www.mordorintelligence.com

The Rack-Scale GPU Infrastructure Market worth USD 9.58 billion in 2026 is growing at a CAGR of 33.92% to reach USD 41.27 billion by 2031. NVIDIA Corporation, Advanced Micro Devices, Inc., Intel Corporation, Super Micro Computer, Inc. and Hewlett Packard Enterprise Company are the major companies operating in this market. The rack-scale GPU infrastructure market is benefiting from new open system designs that reduce the complexity of building large AI clusters. The Open Compute Project published the XOC-N and its Open Cluster Designs for AI white paper in January 2026, providing operators with a vendor-neutral blueprint for both air-cooled pod designs and liquid-cooled AI factory builds. NVIDIA also brought Spectrum-X Ethernet Photonics into production with Vera Rubin in 2026, extending the push toward higher-efficiency optical networking inside large clusters. The rack-scale GPU infrastructure market has a split competitive structure. NVIDIA still sets a large share of platform specifications through its Vera Rubin roadmap and DSX reference architecture, which gives it unusual influence over how commercial AI factories are configured. It's the May 2026 production update, named 30 system builders in full-scale production and more than 350 supply chain partners across 30 countries, showing that manufacturing capacity is broad even while platform control stays concentrated. May 2026: NVIDIA announced that the Vera Rubin platform had entered full production, with production shipments scheduled for fall 2026. The pod-scale platform, integrating 5 purpose-built racks as a single AI supercomputer delivering 10x agent throughput versus Grace Blackwell, engaged over 350 supply chain partners across 30 countries, with 150 based in Taiwan.

Rack-Scale GPU Market Size, Share & 2031 Growth Trends Report

www.mordorintelligence.com

The Rack-Scale GPU Market worth USD 9.27 billion in 2026 is growing at a CAGR of 34.37% to reach USD 40.60 billion by 2031. Dell Technologies Inc., Hewlett Packard Enterprise Company, Super Micro Computer, Inc, Lenovo Group Limited and Inspur Group Co., Ltd. are the major companies operating in this market. Services are projected to expand at a 34.96% CAGR through 2031, making it the fastest-growing offering and showing how quickly operating complexity is moving beyond the comfort level of many buyers. CoreWeave's June 2026 bring-up of NVIDIA Vera Rubin NVL72 involved liquid-cooled storage, custom software-defined cooling control, and unified rack management, which illustrates the depth of coordination required before a rack enters production use. Enterprises remain important, but many of them are still approaching the rack-scale GPU market through hosted compute, dedicated cloud capacity, or managed on-premises solutions rather than through direct full-rack ownership. IBM's expanded collaboration with NVIDIA at GTC 2026 reflected that pattern by combining storage, open AI services, and accelerated compute for organizations moving from pilot AI programs into production use. June 2026: AMD and Rackspace Technology signed a definitive agreement for a phased deployment of 30 MW of AMD AI compute, delivering enterprise AI cloud, inference-as-a-service, and bare metal AMD Instinct MI-series infrastructure as a governed alternative to hyperscaler GPU offerings for enterprise workloads. May 2026: NVIDIA announced Vera Rubin has ramped into full production across 350+ manufacturing facilities in 30 countries, introducing NVIDIA Spectrum-X Ethernet Photonics, the first co-packaged-optics-based switches in production, providing 5x better power efficiency and forming the interconnect fabric for million-GPU AI factories.