AI Infrastructure

Pune – October 12, 2026 -- The global market for data center racks drawing more than 100 kW is forecast to grow from USD 37.2 billion in 2025 to USD 52.5 billion by 2034, a 3.5% CAGR, according to a new Credence Research Inc. report.

AI training and inference, accelerated computing and high-performance computing (HPC) workloads are driving demand. They require substantially more rack-level power and cooling than conventional enterprise infrastructure. Direct-to-chip liquid cooling, rear-door heat exchangers, immersion-ready enclosures and higher-capacity power distribution are becoming core design considerations for new deployments.

Rochester, N.Y. – October 12, 2026 -- Vuzix Corporation (NASDAQ: VUZI) has demonstrated 19,200 parallel optical pathways across an approximately 30 mm optical path on its Vuzix Causeway waveguide evaluation platform, and is now offering evaluation units to qualified partners exploring optical interconnects for AI data centers.

Vuzix Causeway routes thousands of optical channels through a compact substrate

The platform applies proprietary waveguide technology to route thousands of parallel optical pathways through a compact substrate. Vuzix targets co-packaged optics (CPO), near-package optics (NPO) and AI computing applications. Future communications implementations are projected to support multi-terabit aggregate bandwidth, subject to source, detector and system integration performance.

Waveguide structures can be tailored to source wavelength and illumination approach for custom architectures using microLED or VCSEL sources.

San Jose – October 12, 2026 -- Cisco and NVIDIA plan to develop an open industry specification for a 400Gb/s-per-lane pluggable optical module that is expected to deliver 4x the bandwidth density of 200Gb/s-per-lane optics. The companies announced the Ultra-Dense Small Form Factor Pluggable (USFP) at the Open Compute Project (OCP) Global Summit 2026.

Cisco also said the first systems powered by its 102.4T Silicon One G300 are shipping to customers. Its 51.2T P200-based systems are running in production at hyperscalers, including Microsoft and Alibaba.

USFP targets 4x bandwidth density and 1.3x power efficiency

Cisco and NVIDIA said USFP is expected to deliver 1.3x the power efficiency of 200Gb/s-per-lane optics. They say it relies on a proven operating model to reach 400Gb/s per lane, and that it preserves serviceability, deployment flexibility and the operational simplicity of today's pluggable optics.

Boston – October 12, 2026 -- Fortress will present a full-scale prototype of its NextGen SMART ILA modular optical infrastructure platform, designed for a 600 kW day-one load that scales to 1.2 MW, at an invitation-only event in Denver, Colorado, on October 20–22, 2026.

Platform targets 400 to 800 kW of ICT capacity across low-, high-density and mixed configurations

Fortress positions the NextGen SMART ILA as a standardized, modular alternative to traditional precast infrastructure. The platform supports low-density, high-density and mixed configurations, allowing consistent deployment across multiple sites as network requirements change.

San Jose – October 12, 2026 -- Flex (NASDAQ: FLEX) has introduced prefabricated AI Modules that integrate power, compute and liquid cooling, and it says they cut AI factory deployment timelines by up to 50% compared with traditional construction.

The modules are part of Flex's Cloud and Power Infrastructure (CPI) segment. Flex plans to separate that segment into Axiom Solutions International, Inc. (Axiom), an independent publicly traded company, in the first quarter of calendar 2027.

Axiom AI Modules combine three factory-built systems

The pre-engineered modules include an 800 VDC power rack for power conversion and distribution, advanced cooling distribution units (CDUs) for liquid cooling, and high-density IT racks. The IT racks integrate AI servers, networking and high-speed interconnects.

Dusseldorf – October 12, 2026 -- Envalior will present a materials toolkit for 800V DC data center infrastructure at Fakuma 2026 in Friedrichshafen, including a Stanyl® CR310 PA46 grade for switchgear arc chambers with a heat deflection temperature [HDT(A)] of 290°C, well above standard PA66. The show runs October 12 to 16, and the company's experts are at Hall B4, stand 4302.

Shift to 800V DC raises demands on arc quenching and connectors

Envalior argues that AI workloads are increasing power density and operating temperatures. Components in power paths and cooling systems must therefore handle higher voltages and thermal loads while protecting costly computing equipment. Martin Burg, Advanced Development Expert at Envalior, said the move to 800V DC requires materials with greater tracking resistance, heat resistance and dimensional stability at high temperatures.

Taipei – October 12, 2026 -- ASRock Rack Inc. is displaying a Vera Rubin NVL72 rack platform, which it says will be available for deployment in the near future, at booth D37 of the 2026 OCP Global Summit in San Jose, California. The event runs October 12 to 15.

ASRock Rack pairs NVIDIA Vera CPUs and HGX Rubin NVL8 with liquid cooling

The 2UXGM-VERA2 carries two NVIDIA Vera CPUs. ASRock Rack built it for agentic AI workloads, citing complex data pipelines and multi-step reasoning.

The 2U16X-GNR2/DLC Rubin is fully liquid-cooled and uses the NVIDIA HGX Rubin NVL8 platform. The company says it provides unified compute and memory bandwidth for massive-scale reasoning models.

The 6UXGM-GNR2 CX8/DLC is based on NVIDIA MGX architecture. It supports liquid-cooled NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs for enterprise AI, visual computing and inference deployments.

Espelkamp – October 12, 2026 -- HARTING Technology Group has joined the NVIDIA MGX ecosystem as a power system component provider for the NVIDIA Vera Rubin NVL72 platform, supplying power whips that span the shift from 480 VAC to 800 VDC data center architecture.

HARTING supplies two power whips covering both MGX power generations

Power whips are cable assemblies that distribute power from the building to each rack. For the 3RU 110kW power shelf of the Vera Rubin NVL72, HARTING provides a 100A/480VAC power whip. For the move to 800 VDC, it offers a next-generation 157A/800 VDC DC-DC power whip that connects the power infrastructure to the compute rack.

Rising compute density forces data centers to handle significantly more power in tighter spaces while maintaining reliability, efficiency and scalability. That puts new demands on the connectivity systems that feed AI infrastructure.

San Jose – October 12, 2026 -- Texas Instruments (Nasdaq: TXN) unveiled the ADC34DSO75, a 3 gigasamples-per-second analog-to-digital converter that it says enables 24/7, low-power, event-based remote monitoring at nanosecond resolution, thousands of times faster than current solutions.

The company debuted the device at the Open Compute Project (OCP) Global Summit alongside two other analog products aimed at AI data center power instability, inadequate real-time monitoring and signal degradation. TI says these problems can lead to costly downtime.

ADC34DSO75 targets nanosecond power-rail glitches that cause downtime

TI says existing monitoring tools often miss nanosecond-level power-rail glitches behind unplanned outages. The ADC34DSO75 combines ultra-high-speed sampling with on-chip digital signal processing in a small form factor to detect and capture transient fault events in real time.

Palo Alto – October 12, 2026 -- insidepacket has announced the AI Access Platform, which connects enterprises to GPU and neocloud providers through private network paths that can be activated in minutes. The platform is planned for introduction in Europe, first with Zayo Europe.

Platform comes in two models for GPU clouds and operators

The first model is a dedicated on-ramp that a GPU cloud sells under its own brand. The second is an operator layer that exposes many GPU clouds to the operator's enterprise customers. The underlying NaaS platform has been in production in the US since September 2025 and in Europe since mid 2026.