The SuperX XN9160-B300 AI Server is purpose-built to accelerate large-scale distributed AI training and AI inference workloads, providing extreme GPU performance for intensive, high-demand applications. Optimized for GPU-supported tasks, it excels in foundation model training and inference, including reinforcement learning (RL), distillation techniques, and multimodal AI models, while also delivering high performance for HPC workloads such as climate modeling, drug discovery, seismic analysis, and insurance risk modeling.

Designed for enterprise-scale AI and HPC environments, the XN9160-B300 combines supercomputer-level performance with energy-efficient, scalable architecture, offering mission-critical capabilities in a compact, data-center-ready form factor.

The launch of the SuperX XN9160-B300 AI server marks a significant milestone in SuperX’s AI infrastructure roadmap, delivering powerful GPU instances and compute capabilities to accelerate global AI innovation.

XN9160-B300 AI server

The SuperX XN9160-B300 AI Server, unleashing extreme AI compute performance within a 8U chassis, features Intel Xeon 6 Processors, 8 NVIDIA Blackwell B300 GPUs, up to 32 DDR5 DIMMs, and high-speed networking with up to 8 × 800 Gb/s InfiniBand.

High GPU power and memory

The XN9160-B300 is built as a highly scalable AI node, featuring the NVIDIA HGX B300 module housing 8 NVIDIA Blackwell B300 GPUs. This configuration provides the peak performance of the Blackwell generation, specifically designed for next-era AI workloads.

Crucially, the server delivers a massive 2,304GB of unified HBM3E memory across its 8 GPUs (288GB per GPU). This colossal memory pool is essential for eliminating memory offloading, supporting larger model residence, and managing the expansive Key/Value caches required for high-concurrency, long-context Generative AI and Large Language Models.

Extreme inference and training throughput

The system leverages the B300 Ultra’s superior FP4/NVFP4 precision and second-generation Transformer Engine to achieve monumental performance leaps. According to NVIDIA, Blackwell Ultra delivers a decisive leap over Blackwell by adding 50% more NVFP4 compute and 50% more HBM capacity per chip, enabling larger models and faster throughput without compromising efficiency.[1] Scaling is effortless, thanks to eight 800Gb/s OSFP ports for InfiniBand or dual 400Gb/s Ethernet. These ports allow for the high-speed, low-latency communication necessary to connect the servers into vast AI Factories and SuperPOD clusters. The fifth-generation NVLink interconnects further ensure that the 8 on-board GPUs communicate seamlessly, acting as a single, potent accelerator.

Robust CPU and power foundation

The GPU complex is supported by a robust host platform featuring Dual Intel Xeon 6 Processors, providing the efficiency and bandwidth required to feed the accelerators with data. The memory subsystem is equally formidable, utilizing 32 DDR5 DIMMs supporting speeds up to 8000MT/s (MRDIMM), ensuring the host platform never bottlenecks the GPU processing.

For mission-critical reliability and sustained performance, the XN9160-B300 is equipped with 12 × 3000W 80 PLUS Titanium redundant power supplies, ensuring extremely high energy efficiency and stability under continuous peak load. The system also includes multiple high-speed PCIe Gen5 x16 slots and comprehensive storage options, including eight 2.5″ Gen5 NVMe hot-swap bays.

Technical specifications:

Market positioning

The XN9160-B300 is built for organizations pushing the boundaries of AI, where maximum scale, next-generation models, and ultra-low latency are core requirements:

• Hyperscale AI Factories: For cloud providers and large enterprises building and operating trillion-parameter foundation models and highly demanding, high-concurrency AI reasoning engines.
• Scientific Simulation & Research: For exascale scientific computing, advanced molecular dynamics, and creating comprehensive industrial or biological Digital Twins.
• Financial Services: For real-time risk modeling, high-frequency trading simulations, and deploying complex large language models for financial analysis with ultra-low latency demands.
• Bioinformatics & Genomics: For accelerating massive genome sequencing, drug discovery pipelines, and protein structure prediction at scales requiring the B300’s immense memory capacity.
• Global Systems Modeling: For national meteorological and governmental agencies requiring extreme compute for global climate and weather modeling and highly detailed disaster prediction.

For more information, visit superx.sg.

The post SuperX launches AI server powered by Blackwell GPU appeared first on Engineering.com.

The collaboration will focus on co-developing and delivering an AI agent platform tailored for industry-specific AI agents in sectors such as healthcare, manufacturing, and robotics, along with an AI computing infrastructure that seamlessly integrates the FUJITSU-MONAKA CPU series and NVIDIA GPUs via NVIDIA NVLink Fusion. The combined AI agent platform and computing infrastructure will accelerate the AI industrial revolution with AI agents that continuously learn and improve. This will enable cross-industry, self-evolving, full-stack AI infrastructure, overcoming the limitations of general purpose computing systems.

Through this initiative, Fujitsu aims to transform industries starting in Japan, fostering widespread AI adoption that expands globally, thereby enhancing competitiveness across all sectors and

Vision for a full-stack AI infrastructure

Working with NVIDIA, Fujitsu aims to realize a human-AI co-creation cycle and continuous system evolution by integrating high-speed AI computing with human judgment and creativity. Specifically, it will accelerate manufacturing using digital twins, leverage physical AI, including robotics for operational automation to address labor shortages, and stimulate human innovation. The resulting cross-industry, full-stack AI infrastructure will deploy advanced AI to achieve transformative automation and intelligence previously unattainable.

For more information, visit global.fujitsu.

The post Fujitsu expands collaboration with NVIDIA on AI infrastructure appeared first on Engineering.com.

A report from the International Energy Agency (IEA) forecasts that electric car sales will exceed 20 million units in 2025, representing over 25% of global car sales. The report also states that by 2030, EVs are expected to account for more than 40% of all car sales globally. To meet this demand, automakers and EVSE manufacturers are accelerating development and deployment, but they face significant challenges:

• Fragmented system architectures across platforms increase design complexity and time-to-market.
• Lack of standardization in power conversion and charging interfaces leads to compatibility issues and higher integration costs.
• Uncertainty around future charging trends makes long-term investment protection difficult.
• High capital expenditure (CAPEX) for testing strains budgets and slows innovation.

Keysight’s expanded EOL portfolio directly addresses these challenges by offering manufacturers a standardized, unified solution that streamlines fragmented system architectures and simplifies integration across power conversion systems — including DC-DC converters, on-board chargers, and EVSE. With scalable, future-proof testing capabilities, the EV2020B and EV2020BE platforms enable manufacturers to confidently innovate across electrification domains. For end users, this translates into faster access to reliable, energy-efficient EV systems that deliver consistent performance and compatibility across vehicle platforms and charging environments.

EV2020B EV Manufacturing Functional Test Platform: Designed to adapt to the diverse and evolving needs of EV manufacturers. Whether testing a DC-DC converter, an on-board charger, or both, the platform offers customizable configurations to meet specific functional test requirements. With built-in flexibility and comprehensive toolsets, EV2020B enables manufacturers to validate performance, ensure compliance, and streamline production — all within a single, scalable solution.

EV2020BE EV Manufacturing Functional Test Platform for EVSE:Engineered to streamline EVSE testing with a unified platform that supports both AC and DC charging standards, eliminating the need for separate systems and reducing operational complexity. Its modular charging interface allows for rapid swap-outs, enabling quick turnaround and minimizing production downtime. With dynamic power sharing for the dual-gun EVSE simulation, it enables the realistic emulation of peak load conditions — critical for validating performance and optimizing energy distribution on dual-gun EVSE.

Keysight’s EOL solutions provide customers with a seamless experience from R&D to manufacturing, building familiarity and confidence in measurement practices across the entire product lifecycle. By aligning early design validation with production testing, Keysight ensures a consistent ecosystem of tools, workflows, and insights to empower manufacturers to innovate with speed and precision.

For more information, visit keysight.com.

The post Keysight expands EV end-of-line test portfolio appeared first on Engineering.com.

Trimble has been a long-term supporter of EWB’s mission in the US by providing technology donations for student chapters, financial grants for disaster relief and climate resiliency projects through its Trimble Foundation Fund, and championing STEM education initiatives. Now, the company is expanding its support to the international stage.

Collaboration’s first initiative: unlocking critical infrastructure funding in Uganda

The collaboration’s first initiative is focused on an 18-month critical infrastructure project in Eastern Uganda with EWB East Africa, focused on a unique challenge: unlocking available local government funds. Up to 70% of the population lacks access to basic necessities such as safe water, quality sanitation, reliable energy and education facilities. While local funds exist to address these issues, district parishes currently cannot access them due to a lack of necessary data, tools and engineering capacity.

Integrated technology and community engagement

The project in Uganda aims to positively impact the lives of people in the community, beginning immediately with three distinct phases.

The initial phase includes crucial preparatory work:

• Technical baseline surveys and infrastructure mapping using Trimble solutions, GIS software and drone support.
• Recruitment and training of 40 local youth enumerators to conduct a population census.

The subsequent phases will encompass feasibility and technical audits, co-design with community leaders, financial modeling and cost-benefit analysis. This information will be utilized in stakeholder presentations to attract commercial investment, including public-private partnerships.

Focus on local capacity building

A component of phase two is a hands-on ‘Buildathon’ event designed to bring together Trimble teams, customers and partners to provide rapid, practical assistance. Trimble and its education partner Panelle, which connects brands with women in business and underrepresented industries, are forming a working group to gather female leaders from construction and engineering to help shape the initiative’s direction.

For more information, visit trimble.com.

The post Trimble expands partnership with Global Engineering Charity appeared first on Engineering.com.

Predicting and designing for success

ICMD simulates key aspects of the printing process—melt pool dynamics, microstructure formation, and resulting part strength—so manufacturers can evaluate materials digitally before production. These validated predictions have been recognized in the U.S. Air Force Research Laboratory AM Modeling Challenge (2020) and the NIST AM Bench Challenge (2022), underscoring their accuracy and industry relevance.

In addition, the Alloy Design Toolkit enables the creation of materials specifically engineered for printability, helping avoid cracking and porosity. By ensuring that alloys are both high-performing and printable, manufacturers can accelerate the path to lighter aerospace components, safer energy systems, and advanced medical implants.

Optimizing builds with printability maps

Scaling AM requires precision in process parameters. ICMD addresses this challenge with intuitive printability maps that highlight safe operating windows for laser power, scan speed, and other variables. These maps help manufacturers reduce defects, achieve fully dense parts, and move to production with greater confidence and efficiency.

Accelerating innovation across industries

From supporting longer fatigue life in aerospace engines to enabling custom alloys for medical devices, ICMD offers a digital approach to materials development. The new AM capabilities help companies manage costs, lower risks, and accelerate the production of high-performance parts. This AM toolkit is the latest addition to the ICMD software platform, following the release of a fatigue toolkit earlier this year.

For more information, visit questek.com.

The post QuesTek adds additive manufacturing modeling to ICMD software appeared first on Engineering.com.

The new iQpump605 pump drive follows the iQpump1000, offering improved sustainability, flexibility, and user-friendly features to enhance efficiency and energy savings. 

iQpump605 is designed with a Hand-Off-Auto (HOA) keypad with a bright sunlight-resistant display, dedicated function keys to help you easily navigate the screens, and data logging capability to record values while you are away from the drive. The pump-centric setup wizard will save you time by taking you step-by-step, asking the right questions to get you commissioned faster. Built-in application presets for submersible and above-ground applications simplify start-up further. An optional Bluetooth keypad is also available to allow you to start up, adjust, and monitor drives with your mobile device through our DriveWizard Mobile app.

iQpump605 drives are available in a variety of enclosure options, including IP20/UL Type 1 and Protected Chassis, IP55/UL Type 12. IP55 models are also available with an optional disconnect switch. Available drive models:

• IP20/UL Type 1: 3 to 60 HP at 240 V | 3 to 125 HP at 480 V
• IP20/Protected Chassis: 75 to 150 HP at 240 V | 150 to 600 HP at 480 V
• IP55/UL Type 12 (with or without disconnect switch): 3 to 60 HP at 240 V | 3 to 125 HP at 480 V

Configured or Bypass Drive Packages are available in UL Type 1, 12, or 3R enclosures. Available configured or bypass drive package models:

• Enclosed Configured (UL Type 1, 12, or 3R):
  208 V up to 150 HP | 240 V up to 150 HP | 480 V up to 600 HP
• Enclosed Bypass (UL Type 1, 12, or 3R):
  208 V up to 100 HP | 240 V up to 100 HP | 480 V up to 250 HP

For more information, visit yaskawa.com.

The post Yaskawa adds iQpump605 drive to enhance efficiency and flexibility appeared first on Engineering.com.

Trimble Technology Outlets are a new and important part of Trimble’s civil construction distribution strategy. By adding authorized resellers representing a wide variety of manufacturers to the Trimble distribution channel, it is easier for users of a wide range of machine types to purchase, install and utilize Trimble technology for improved jobsite productivity and profitability.

As an authorized Trimble Technology Outlet, 4Rivers will offer Trimble grade control solutions — including site positioning systems and correction services — to customers in Colorado, New Mexico and the El Paso, Texas area. 4Rivers customers can expect to receive the same superior level of support, including installation, service and training, of which the Trimble dealer channel is known to deliver.

For more information about Trimble, visit trimble.com.

The post 4Rivers Equipment named Trimble Technology Outlet appeared first on Engineering.com.

SLT has become an integral part of the high-volume manufacturing process, as the complex devices used in AI and cloud infrastructure deployments push the boundaries of power consumption and thermal output. Teradyne Titan HP, currently in production at a number of large customers, is purpose-built to test these devices in real-world conditions. Currently supporting up to two kilowatts of power, with a roadmap to support four kilowatts in the near future, customers can be assured that their investment today will meet the demanding requirements of tomorrow’s devices as well.

Teradyne Titan HP’s superior thermal control ensures devices meet rigorous quality standards at or above the expected yield. This includes:

• A multi-branch cold plate architecture, which cools both the device under test (DUT) and non-DUT components for comprehensive thermal management.
• Asynchronous cooling control with proportional, integral and derivative (PID)-tuned, per-site thermal control, which provides precise, efficient cooling to prevent overheating.
• A DUT heater option, which enhances the accuracy of automated temperature control (ATC) for more precise thermal testing.

Teradyne Titan HP is the world’s leading SLT solution, enabling mission mode testing for next-generation AI and cloud infrastructure devices. As part of Teradyne’s comprehensive portfolio of test equipment, which delivers superior test for every device across every insertion, Teradyne Titan HP ensures that our customers’ capital equipment investments are future-proofed for tomorrow’s evolving market demands.

For more information, visit teradyne.com/titan-hp.

The post Teradyne unveils Titan HP system level test platform appeared first on Engineering.com.

This new partnership combines Mitsui Seiki’s renowned positional accuracy and repeatability technologies with MMT’s premium service and expertise to deliver significant technological and service benefits to manufacturers in the aerospace, mold and die, medical, automotive and optics industries.

Mitsui Seiki’s complete line of vertical and horizontal machining centers, jig borers, jig grinders, and thread grinders will be available to MMT customers, tailored to their requirements. MMT’s experienced staff will guide companies through the customization process, as well as automation configuration, application engineering, and after-sale support.

For more information, visit mitsuiseiki.com.

The post Mitsui Seiki USA partners with Mitcham Machine Tools appeared first on Engineering.com.

Through the partnership, BUNDLAR’s Knowledge Now platform will enable enterprises and public-sector organizations to create, manage, and publish immersive XR content directly to Vuzix smart glasses without writing a single line of code. By combining world-class wearable hardware with scalable, no-code content creation, Vuzix and BUNDLAR are closing the XR ecosystem gap and unlocking new revenue streams in training, field service, security, and defense.

Strategic growth opportunities

• Enterprise Training: Quick-service restaurant (QSR) staff using Vuzix M400 devices with BUNDLAR content modules reduced training time by up to 40 percent, a measurable ROI driver for high-turnover industries.
• Field Service Support: Technicians working in disconnected or challenging environments can follow hands-free step-by-step workflows, reducing downtime, minimizing errors, and improving operational efficiency.
• Defense and Government: Personnel can access immersive content to operate, maintain, repair, and replace mission-essential equipment in the field, even in offline or air-gapped settings — enhancing readiness and reducing support costs.
• Public Safety and Security: Security personnel using Vuzix devices running BUNDLAR content can overlay real-time visual information on top of live video streams, receiving alerts and responding quickly while remaining mobile.

For more information, visit vuzix.com.

The post Vuzix, Bundlar partner on no-code XR for enterprise, defense appeared first on Engineering.com.