By: Network-Switch.com | July 28, 2026
- 1. NVIDIA Expands Agent Toolkit for Engineering
- 2. Wistron Opens U.S. Plant for NVIDIA AI Systems
- 3. Microsoft Plans Large-Scale AMD Helios Deployment
- 4. Supermicro Launches H15 EPYC 9006 Servers
- 5. Arista Adds Zero-Trust Security to VeloCloud SD-WAN
- 6. HPE Joins DOE Genesis Mission R&D Projects
- 7. Lenovo Expands U.S. Server Manufacturing
- 8. Frequently asked questions (FAQs)
- 9. Sources
Enterprise infrastructure news between July 20 and July 27 centered on rack-scale AI systems, accelerated server platforms, secure branch networking, scientific computing and regional manufacturing capacity. This edition contains seven newly selected developments and excludes previously covered Dell PowerStore, Dell PowerEdge, Juniper QFX5252 and HPE Aruba topics.
No qualifying standalone Dell, Juniper or HPE Aruba product announcement was confirmed for this review window. Older release-to-availability updates were not reused.
NVIDIA Expands Agent Toolkit for Engineering
NVIDIA expanded its Agent Toolkit with re-architected PhysicsNeMo libraries and updated CUDA-X tools. The additions give engineering agents access to AI physics models, sparse solvers and quantum-chemistry capabilities for chip, system and industrial design workflows.
NVIDIA also positioned Nemotron 3 Ultra and its ACE-RTL research agent for register-transfer-level coding and verification. Cadence, Siemens, Synopsys, Samsung and other engineering companies are applying the stack to electronic design automation, packaging, simulation and thermal analysis.
Wistron Opens U.S. Plant for NVIDIA AI Systems
Wistron opened a 324,000-square-foot manufacturing facility in Fort Worth, Texas, to produce NVIDIA AI systems. The plant is building Grace Blackwell Ultra hardware and is scheduled to manufacture Vera Rubin Superchips.
The facility is part of a $700 million investment, has created more than 500 jobs and is expected to reach 1,000 employees by the end of 2026. Wistron said production will scale to tens of thousands of advanced boards per month.
Microsoft Plans Large-Scale AMD Helios Deployment
AMD and Microsoft expanded their infrastructure partnership across GPUs, CPUs, networking and software. Microsoft plans to deploy the AMD Helios rack-scale platform on Azure for frontier-model inference, Azure AI services and customer workloads.
Helios integrates Instinct MI455X GPUs, EPYC Venice CPUs, Pensando networking and ROCm software. Azure will also introduce new EPYC-powered virtual machines and broaden its use of Pensando DPUs. AMD said Helios shipments will begin in the second half of 2026.
Supermicro Launches H15 EPYC 9006 Servers
Supermicro introduced its H15 server portfolio based on 6th Gen AMD EPYC 9006 processors, with systems supporting up to 256 cores and 512 threads. The company cited up to 1.7 times higher generational CPU performance, twice the PCIe bandwidth and 2.6 times the memory bandwidth.
The portfolio covers enterprise servers, cloud platforms, dense multi-node systems, blades and all-flash storage. Its Petascale Storage models support up to 4.8 PB per system, while Supermicro's AMD Helios implementation scales to 72 GPUs for AI training and inference.
Arista Adds Zero-Trust Security to VeloCloud SD-WAN
Arista launched AI-driven Edge Threat Management for VeloCloud SD-WAN. The software combines branch connectivity with firewalling, threat prevention, segmentation, Geo-IP filtering, DNS filtering and centralized policy management.
Arista AVA adds policy explanations and traffic simulation to help administrators understand rule behavior before changes are applied. ETM is scheduled for general availability in the fourth quarter of 2026 and will support current VeloCloud hardware and virtual edge platforms.
HPE Joins DOE Genesis Mission R&D Projects
HPE was selected for multiple research projects under the first phase of the U.S. Department of Energy's Genesis Mission. The work covers AI model applications, network optimization, cybersecurity, scientific workflows, fusion energy, weather forecasting and water availability.
Projects include SciNet, an AIOps platform for self-driving scientific networks, and SwarmSlicer, which explores network and GPU slicing for robotics over future 6G systems. HPE is also working with AMD and Oak Ridge National Laboratory on Lux, a dedicated AI system for scientific research.
Lenovo Expands U.S. Server Manufacturing
Lenovo doubled the footprint of its Whitsett, North Carolina, manufacturing facility to 890,000 square feet. The expansion adds 35 percent more production capacity and infrastructure designed for 20 megawatts of power.
The site includes additional direct-liquid-cooling capability and is prepared for the next two generations of high-performance computing systems. Lenovo said the investment will expand U.S. server output, shorten delivery paths and create approximately 1,000 jobs as production scales.
Frequently asked questions (FAQs)
What were the most important enterprise infrastructure developments between July 20 and July 27, 2026?
The strongest updates were AMD's Helios deployment plan with Microsoft, Supermicro's H15 server portfolio, Arista's integrated zero-trust security for VeloCloud SD-WAN, HPE's Genesis Mission research work, Lenovo's expanded U.S. server production, and two NVIDIA engineering and manufacturing developments.
What is the AMD Helios rack-scale platform?
AMD Helios is an integrated AI platform combining AMD Instinct MI455X GPUs, EPYC Venice CPUs, Pensando networking and ROCm software for large-scale AI training and inference. Microsoft plans to deploy it on Azure.
What does Arista Edge Threat Management add to VeloCloud SD-WAN?
It adds locally enforced firewalling, threat prevention, segmentation, Geo-IP filtering, DNS filtering and AI-assisted policy analysis to the branch SD-WAN edge, reducing the need for a separate firewall appliance.
Why is NVIDIA's expanded Agent Toolkit relevant to infrastructure engineering?
The toolkit connects engineering agents to PhysicsNeMo and CUDA-X capabilities for physics modeling, sparse solvers, quantum chemistry, chip design and system simulation, allowing more engineering workflows to run on accelerated computing.
Were there new Dell, Juniper or HPE Aruba product announcements in this review?
No qualifying standalone product announcement from Dell, Juniper or HPE Aruba was confirmed for this edition. Older availability updates and previously covered Juniper QFX and HPE Aruba topics were excluded.
Sources
- NVIDIA: Agent Toolkit With PhysicsNeMo and CUDA-X
- NVIDIA: Wistron Fort Worth AI Systems Manufacturing Plant
- AMD: Microsoft Azure AI Infrastructure Partnership
- Supermicro: H15 Server Portfolio With AMD EPYC 9006
- Arista: AI-Driven Zero Trust Branch
- HPE: DOE Genesis Mission R&D Projects
- Lenovo: North Carolina Server Manufacturing Expansion