GlobalFoundries and Marvell are scaling up silicon photonics efforts as data centers outgrow copper wiring
AI's data problem isn't compute anymore, it's how fast bits can move between chips
Enterprise articles
Apple could return to the server market with M8 Ultra hardware and Nvidia networking
The system could use NVLink Fusion to connect two or four M8 Ultra chips
Microsoft warns that killing VBScript could break Windows activation for businesses
VBScript is being phased out in stages after nearly 30 years as part of Windows
Magnetic tape storage is booming again, thanks to AI's unquenchable hunger for data
New LTO Ultrium LTO-10 tapes can store up to 100TB of compressed data per cartridge
Magnetic Storage Bubble: The Linear Tape-Open (LTO) format was designed back in the 90s as an open-standard alternative to proprietary tape formats. Now, the magnetic storage technology is growing in both relevance and shipment volumes because AI companies can't seem to stop gobbling up digital data for LLM training and inference workloads.
IBM is building a dual-architecture mainframe chip that will run Arm and Z software on the same cores
The chip will use 11 IBM Z cores built on a 2-nanometer process, running above 5.7 GHz
How "tokenomics" might save AI PCs
AI PCs finally have a reason to exist: cutting your company's AI bill
Nvidia may have just built its own hyperscaler without owning a single data center
The neoclouds are burning cash by the billion. Nvidia just lined up the money to keep them building
Why it matters: Nvidia has quietly assembled every piece of a cloud giant, and it doesn't even need to own the buildings. That is the theory making the rounds, and once you see it laid out, the last two years stop looking like a chip company having a very good run and start looking like a plan. The financing arrangement Nvidia signed this month with BlackRock, Goldman Sachs, and four other Wall Street heavyweights, aimed at funneling more than $500 billion of outside capital into AI data centers, would be the last piece clicking into place. Most outlets filed it under the now usual "more AI money" headline, but there is a more interesting read.
Enterprise SSDs now consume 48% of global NAND flash supply
AI is turning the NAND market upside down, with training making way for inference
Cargo thieves are staging car crashes to steal AI data center hardware
Authorities have recovered millions in stolen equipment, including $550,000 in Meta server switches
TL;DR: Cargo thieves are targeting equipment used in AI data centers, from server switches to cooling systems and networking hardware. In two recent California cases, criminals appear to have used car crashes to separate trucks from their security escorts before the shipments vanished. The incidents highlight how organized groups are using carrier-identity fraud, GPS manipulation and, increasingly, aggressive roadside tactics to steal high-value tech moving through the freight network.
Kioxia's new PCIe 6.0 SSDs want to feed GPUs data at a fraction of HBM's price
GP1 SSDs are optimized for direct GPU access through the PCIe interface
Thousands of server motherboards are vulnerable to controller flaws that could give attackers hardware-level control
Baseboard management controllers remain riddled with exploitable flaws
AMD's Helios rack is finally shipping, and it's coming for Nvidia's AI dominance
AMD claims 20% better performance than Nvidia's Vera CPU with new Epyc chips
Nvidia is challenging 20 years of datacenter CPU design with new Vera chip
Nvidia joins the battle of making the CPU interesting again
The big picture: Nvidia lifted the embargo on its Vera CPU deep dive this week, and the disclosure amounts to a direct challenge to two decades of x86 datacenter design philosophy. This is the most detail the company has shared on the chip since it first appeared on the Rubin roadmap, and it confirms something I have suspected for a while: Nvidia is not treating the CPU as an attach story anymore. It is treating it as a battleground with a lot of potential dollars at play.
HPE is quietly pivoting from servers to networking, and Cisco should be paying attention
HPE focuses on the network side of agentic AI
Microsoft Teams is getting Wi-Fi location "check-in," but it's less creepy than it sounds
A powerful surveillance tool or just another "nothingburger" from Redmond?
Windows Server is getting new network safety capabilities with DNS over HTTPS
The feature has been available on client Windows editions for quite some time
Microsoft took OpenClaw, wrapped it in enterprise security, and called it Scout
3,000 Microsoft employees are already using Scout to book travel, schedule meetings, and handle paperwork
Intel launches Xeon 6+ Clearwater Forest with 288 E-cores on 18A node, teases Diamond Rapids for 2027
Intel just shipped its most ambitious server chip yet, Diamond Rapids promises PCIe 6.0 and twice the memory bandwidth next year
Tech executives are suffering from "AI psychosis," says Box CEO
Your CEO is vibe-coding prototypes and calling it an AI strategy
TL;DR: A growing backlash against the forced adoption of AI is building across the industry, but enterprise executives remain as bullish as ever on the LLM-driven future. According to at least one prominent tech leader, that enthusiasm may signal how out of touch those at the top have become.
Nvidia Vera CPU impresses in early Nvidia-sanctioned benchmarks
Controlled tests show strong results against Intel Xeon and AMD Epyc
Dell makes its strongest case yet for keeping AI out of the cloud
Dell counts 5,000 enterprise customers for its AI Factory, moving from pitch to proof
A single Micron SSD can now store 245TB and use half the power of hard drives
The future of AI infrastructure can also be measured in rack space and power bills
Google is assembling the pieces for enterprise AI agents, now adoption becomes the challenge
From agent monitoring to cross-cloud data access, Google is trying to solve the barriers to large-scale deployment
72 sticks of server RAM were headed for the trash. They're now worth $20,000
Enterprise hardware lifecycles run on accounting rules, not market value – and that's how perfectly good hardware ends up as e-waste
AI infrastructure boom pushes AMD, Intel and Arm to new valuation heights
Nvidia still dominates, but AI demand is lifting the rest of the chip market