Connect with us

Tech

Why Investors See More Payoff in Microsoft’s AI Spending Than Meta’s

Published

on

Microsoft shares rose 15.5% on July 30, while Meta fell 8%, giving investors two sharply different ways to assess Big Tech’s AI infrastructure boom. Microsoft paired heavy spending with faster Azure growth, rising Copilot adoption and a larger commercial backlog. Meta’s advertising business grew strongly, but expenses climbed and free cash flow dropped.

Microsoft is not spending cautiously. It reported $41 billion in quarterly capital expenditures, while Meta expects total 2026 capital expenditures of $130 billion to $145 billion. The periods and accounting details differ, so the figures are not directly comparable. Investors responded more favorably to Microsoft because its results showed clearer links between infrastructure investment, cloud consumption, software subscriptions and contracted demand.

The July 30 market reaction followed Microsoft’s fiscal fourth-quarter results. Quarterly revenue and operating income both rose 18%. Azure revenue increased 43%, while full-year Azure revenue exceeded $100 billion for the first time.

Microsoft also reported more than 30 million paid Microsoft 365 Copilot seats, with net additions more than doubling from the previous quarter. Its commercial remaining performance obligation — contracted revenue not yet recognized — rose 84% to $678 billion. About 30% is expected to be recognized within 12 months.

Those disclosures give investors several measurable demand signals. Customers are buying Copilot licenses, consuming Azure services and making multiyear commitments. Microsoft also operates several commercial platforms that can use the same infrastructure, including Azure, Microsoft 365 and GitHub.

The buildout still carries financial pressure. Microsoft reported $19.6 billion in quarterly free cash flow as capital expenditures increased. Its company gross margin percentage fell year over year, driven partly by the sales mix shifting toward Azure and continued AI infrastructure investment. Microsoft forecasts Azure growth of approximately 45% in constant currency for its next quarter.

Vendor growth does not guarantee that each deployment will produce an acceptable return. Because AI agent cloud costs can vary sharply between similar workflows, IT teams need workload-level controls, adoption data and measurable operational results before expanding Azure AI or Copilot contracts.

Meta’s AI returns remain harder to separate

Meta raised the lower end of its 2026 capital-expenditure forecast from $125 billion to $130 billion, leaving the upper end at $145 billion. The forecast includes principal payments on finance leases. Quarterly free cash flow fell to $784 million from $8.55 billion a year earlier.

Meta’s core business remained strong. According to Meta’s second-quarter results, revenue increased 28% to $60.8 billion. Ad impressions rose 14%, while average ad prices increased 12%.

Meta says AI is improving its core business and supporting new products, but it does not report AI-attributable revenue separately. Total costs and expenses rose 55%, while operating income declined 8%. The quarter included $2.4 billion in legal charges and $1.18 billion in severance expenses, so AI investment alone did not cause the decline.

Meta expects its infrastructure to support larger models, personal agents and future enterprise products. It is also reportedly considering whether to sell access to excess AI compute, but it has not announced a commercial service, pricing, availability or financial targets.

Investors are applying a similar test elsewhere. Google recently raised its 2026 capital-spending forecast to as much as $205 billion as cloud demand accelerated. Microsoft currently provides stronger evidence of revenue growth and contracted demand, while Meta’s newer AI businesses remain financially less visible.

Read more: Larger capital budgets do not eliminate every infrastructure constraint; AI data center networking is becoming a separate bottleneck as GPU clusters place more pressure on switches, interconnects and congestion controls.

>

Continue Reading

Tech

5 Best Mac Studio Alternatives for 2026

Published

on

Key takeaway: Lenovo’s ThinkStation P3 Ultra SFF Gen 2 is the best overall option because it balances performance, upgrade options, and business support. The ASUS ROG NUC 16 is better for NVIDIA-powered creative work, while the Minisforum MS-S1 Max is designed for running large AI models locally.

Mac Studio delivers considerable power in a compact desktop, but some configurations have become harder to find. These five alternatives are worth considering if your preferred Mac Studio is unavailable, your applications require Windows or Linux, or you want more replaceable components.

Each system has a distinct strength. Lenovo and HP focus on professional applications, ASUS caters to creators, Minisforum prioritizes local AI, and Framework offers greater repairability. I compared their performance, features, pricing, ease of ownership, and customer support.

Best Mac Studio alternatives at a glance

Prices as of July 30, 2026.

Why you can trust TechRepublic

TechRepublic applies the same evaluation criteria to every product in a buyer’s guide. These ratings cover features, pricing, ease of ownership, and customer support. They do not represent hands-on benchmark testing.

How I evaluated the best Mac Studio alternatives

I scored each computer using five weighted categories:

  • General features (35%): I compared performance, memory, storage, ports, cooling, and compact design.
  • Advanced features (25%): I considered local AI capabilities, creator tools, professional software certification, and room for expansion.
  • Pricing (20%): Products earned more credit for clear and affordable pricing. I also considered whether the starting price covers a complete, usable computer.
  • Ease of use (10%): I looked at setup, maintenance, upgrades, and everyday system management.
  • Customer support (10%): I compared warranty coverage, repair services, replacement parts, and driver support.

Lenovo ThinkStation P3 Ultra SFF Gen 2: Best overall

Lenovo logo
Image: Lenovo

Rating: 4.5 out of 5

Why I chose Lenovo

The ThinkStation P3 Ultra SFF Gen 2 is the best overall option because it balances power, upgrade options, and business support. It can be equipped with an Intel Core Ultra 9 processor, professional NVIDIA graphics, and up to 128GB of memory. This makes it suitable for engineering, 3D design, software development, and other demanding work.

Lenovo also provides more room to expand than most compact desktops. The system supports as many as four internal SSDs, Windows and Linux, and certifications for professional applications. High-end configurations can become pricey, and available ports, expansion slots, and warranty coverage vary.

Pros
Cons
Strong performance for professional applications. High-end configurations can become pricey.
Room for as many as four internal SSDs. Expansion options vary by configuration.
Supports certified professional software and Linux.

Pricing

Starts at $2,005.81.

Standout features

  • Professional graphics: Supports NVIDIA RTX PRO graphics with up to 24GB of video memory.
  • Generous storage: Accommodates as many as four internal SSDs with up to 16TB of total storage.
  • Business options: Available with Intel vPro, Linux, and support for certified professional applications.

ASUS ROG NUC 16: Best for creative and video workloads

Asus ROG logo
Image: ASUS

Rating: 4.4 out of 5

Why I chose ASUS

The ROG NUC 16 is the best option for creators who rely on NVIDIA graphics. Available configurations include an Intel Core Ultra 9 processor and an RTX 5080 Laptop GPU with 16GB of video memory. NVIDIA Studio drivers and dedicated video encoding make it suitable for editing, rendering, streaming, and other graphics-heavy work.

Windows 11 Home is included, and owners can access the memory and storage without tools. The computer also supports as many as five 4K displays. However, it is pricey, and its laptop-class RTX 5080 does not deliver the same performance as the desktop card with a similar name.

Pros
Cons
Strong NVIDIA performance for creative applications. It is pricey.
Easy memory and storage upgrades. Uses a laptop-class GPU.
Supports as many as five 4K displays. Lacks professional software certification.

Pricing

Starts at $3,199 and includes Windows 11 Home.

Standout features

  • Creator-focused graphics: Offers an RTX 5080 Laptop GPU with 16GB of video memory.
  • Simple upgrades: Memory and internal storage can be replaced without tools.
  • Multi-monitor support: Handles as many as five 4K displays.

HP Z2 Mini G1i: Best for certified professional applications

HP logo
Image: HP

Rating: 4 out of 5

Why I chose HP

The HP Z2 Mini G1i is a good choice for readers who use certified design, engineering, architecture, or media applications. It supports Intel Core Ultra processors, professional NVIDIA graphics, and up to 128GB of memory. HP also offers business support and onsite service options for organizations managing several workstations.

Its compact case can sit horizontally or vertically or mount behind a monitor. Memory, storage, and graphics are accessible without tools, while separate cooling modes let users favor performance or quieter operation. Its main drawback is its high starting price.

Pros
Cons
Certified for professional design and engineering applications. Expensive.
Tool-free access to major replaceable components. Lower storage capacity than several competitors.
Broad business support and onsite service options. Many desirable ports and components cost extra.

Pricing

Starts at $3,646.00.

Standout features

  • Professional certification: Suitable for applications that require approved hardware and drivers.
  • Flexible placement: Can sit on a desk or mount behind a monitor.
  • Adjustable cooling: Offers separate performance, quiet, and rack settings.

Minisforum MS-S1 Max: Best for large local AI models

Minisforum logo
Image: Minisforum

Rating: 4 out of 5

Why I chose Minisforum

The MS-S1 Max is the most AI-focused computer in this guide. Its AMD Ryzen AI Max+ processor combines computing, graphics, and AI acceleration with 128GB of shared memory. This large memory pool allows it to run local AI models that may be too large for a conventional graphics card.

It also provides fast wired networking, several high-speed USB connections, expandable storage, and room for an internal expansion card. Users can place it on a desk, install it in a rack, or connect several systems for larger projects. On the downside, its memory cannot be upgraded, and it lacks the professional software certification available from HP and Lenovo.

Pros
Cons
Large memory pool for local AI workloads. Memory cannot be upgraded.
Fast networking and extensive USB connectivity. Limited support for third-party software compatibility.
Supports desktop, rack, and multi-system setups. More specialized than a general-purpose workstation.

Pricing

Costs $3,639 with a Ryzen AI Max+ 395 processor, 128GB of memory, and a 2TB SSD.

Standout features

  • Local AI capacity: Its 128GB shared-memory pool can support unusually large AI models.
  • Fast connections: Includes dual 10GbE networking and high-speed USB4 ports.
  • Flexible setup: Can operate as a desktop, rack-mounted system, or part of a small cluster.

Framework Desktop: Best for DIY repairability and value

Framework logo
Image: Framework

Rating: 4 out of 5

Why I chose Framework

Framework Desktop is the best option for hands-on users who want a repairable and customizable compact PC. It supports AMD Ryzen AI Max processors, Radeon graphics, and up to 128GB of shared memory. The largest configuration can make as much as 112GB available to graphics and AI applications.

Framework provides replacement parts, repair guides, two internal SSD slots, and customizable front ports. Its starting price is affordable, but the DIY model does not include an operating system. Storage and a CPU fan also cost extra, while the memory cannot be upgraded after purchase.

Pros
Cons
Affordable starting price. The starting configuration is incomplete.
Large memory pool for graphics and local AI. Memory cannot be upgraded later.
Good repair support and replacement-part availability. Standard warranty lasts only one year.

Pricing

Starts at $1,269 for the DIY Edition. An operating system is not included, while storage and a CPU fan add to the final cost.

Standout features

  • AI-ready memory: The 128GB model can assign as much as 112GB to graphics and AI work.
  • Repair-friendly design: Replacement parts and step-by-step repair guides are available.
  • Custom front ports: Two modular bays let owners select the connections they use most.

Must-read Apple coverage

How to choose the best Mac Studio alternative

  • Check software compatibility: Start with the applications, plug-ins, peripherals, and operating system you need. Professional certification may be important for specialized design and engineering software.
  • Compare complete prices: Include the processor, graphics, memory, storage, operating system, and warranty you need. A low starting price may exclude essential components.
  • Choose the right graphics and memory: NVIDIA graphics may provide better support for creative and professional applications, while a large shared-memory pool can help with local AI.
  • Consider long-term ownership: Look at upgrade access, cooling, ports, storage capacity, warranty coverage, and replacement-part availability.

Frequently asked questions (FAQs)

Are AI TOPS figures directly comparable between compact workstations?

No. Vendors may calculate TOPS differently, so the largest figure does not always identify the fastest computer. Application support, usable memory, model compatibility, and independent benchmark results provide a more useful comparison.

When is ECC memory worth paying for?

ECC memory can be valuable for long calculations and production work where an error could corrupt a project or invalidate the results. Most general users and creators will benefit more from additional memory capacity or better graphics.

What does ISV certification mean for a workstation buyer?

ISV certification means a particular hardware and driver combination has been validated for professional software. It can reduce compatibility and support problems with CAD, engineering, and other specialized applications. Buyers should confirm that their intended configuration supports the software they use.

Apple’s new leasing program could change how US customers upgrade iPhones, Macs, iPads, and Apple Watches. 

>

Continue Reading

Tech

Cryogenic Firefighting Robot Could Fight Wildfires

Published

on

Yiannis Levendis, a professor of mechanical and industrial engineering at Northeastern University, has developed a prototype firefighting robot that sprays liquid nitrogen onto flames to extinguish them quickly. The project grew from a laboratory experiment in 2008, when Levendis was testing shredded tire rubber as a possible alternative fuel.

“Liquid nitrogen, when thrown on fire, quickly evaporates and expands into gas almost 1,000 percent,” Levendis said. “That suffocates the fire and creates a cooling effect on the fuel.”

He said the flames were extinguished almost immediately, prompting him to explore whether the same approach could be adapted for wildfire response. Northeastern Global News reported that the robot is designed to suppress small fires and hotspots before they grow into major wildfires.

How the robot works

The prototype, built with Northeastern students Hayden Hishmeh and Aobo Liu, resembles a compact all-terrain vehicle with caterpillar tracks, a liquid nitrogen tank and a spray hose. It can be operated remotely or potentially autonomously, allowing it to move into areas that may be too dangerous for firefighters.

Unlike many wildfire retardants, liquid nitrogen does not leave behind chemical residue. As it rapidly turns into gas, it cools burning vegetation while reducing the oxygen available for combustion. The robot is currently being tested on controlled fires using forest fuels common in wildfire-prone areas of the western United States.

Tom Ryden, executive director of MassRobotics, said the approach could appeal to firefighting agencies or industry partners.

“A perfect use for robots is to send them into dangerous spots,” Ryden said, according to Northeastern Global News. “Even if we lose a robot, it’s just the cost of the machine that’s lost, not a person.”

Levendis also believes the cryogenic system could eventually be adapted for much larger firefighting vehicles. “I believe it can be converted to vehicles as large as firetrucks,” he said. “We are fairly close.”

More must-read AI coverage

Field deployment still faces major hurdles

The technology is still in the prototype stage, and significant engineering challenges remain before liquid-nitrogen firefighting robots could be deployed in active wildfire operations. Cryogenic storage, transportation logistics and the amount of nitrogen required for sustained firefighting would all need to be addressed.

Even so, the project points toward a different strategy for wildfire response: using robots to attack small ignition points quickly, especially in terrain that is too risky for human crews. If that approach proves practical, it could become a useful complement to existing wildfire suppression methods rather than a replacement.

Also read: NASA’s robotic rescue mission shows how remotely operated machines can take on high-risk work where direct human access is limited.

>

Continue Reading

Tech

Samsung Galaxy Buds Leak Reveals Ear-Hook Workout Design

Published

on

Leaked images published July 30 show an unreleased Samsung earbud design with hooks that wrap around the outside of the ear. Android Central reported that software graphics associate it with the internal label “Buds Canal 5” and model number SM-U600.

The wraparound form could provide a more stable fit during running, cycling, or gym sessions while leaving the ear canal relatively open. Samsung has not announced the product, confirmed its intended use, or said whether it will reach stores.

Two leaked designs leave the product unsettled

The images published by Android Central show separate left and right earbuds with full exterior hooks. They also appear to show a charging case with a USB-C port.

The SM-U600 identifier and recurring references in Samsung software suggest the device has progressed beyond a single concept image. Software listings can indicate that support is being prepared, but they do not confirm an immediate launch. Samsung’s unannounced Galaxy S26 FE and Galaxy Tab S12 recently surfaced in a Google app list without official release dates.

The relationship between Buds Canal 5 and the earlier Galaxy Buds Able reporting remains unclear. Android Authority associated both names with SM-U600, although earlier images showed a smaller clip-style design rather than a hook wrapping around the ear.

Reports have also differed over the audio technology. Early coverage mentioned bone conduction, while later reporting described an air-conduction system that directs sound toward the ear without blocking the canal. Samsung has confirmed neither description.

“Galaxy Buds Able” may be a development label rather than a retail name. Android Authority later reported that the product could be marketed as Galaxy Buds On, but Samsung has not confirmed either name.

Battery life, water resistance, active noise cancellation, controls, wireless codecs, and device compatibility remain unknown. Pricing has not leaked. A separate report pointed to a possible late-September or October 2026 launch, although Samsung has not announced a release date.

Workout stability comes with audio trade-offs

Ear hooks provide additional support during repeated movement because the earbuds rest around the outside of the ear instead of relying only on an in-ear or outer-ear fit. An open design can also help runners and cyclists hear vehicles, voices, and other nearby sounds.

The trade-offs include weaker passive isolation, reduced bass in noisy settings, and more sound leakage than sealed earbuds. Wind noise, pressure behind the ear, and comfort alongside glasses could also affect longer sessions.

Samsung’s standard Galaxy Buds4 use an open fit, while the Buds4 Pro use silicone tips and emphasize stronger isolation and adaptive noise cancellation. The leaked model would add a more exercise-focused shape rather than directly replacing either product.

That form factor could complement the Galaxy Watch9 and Watch Ultra2, which add workout, recovery, trail, and health features. Samsung’s broader wearable health tools also show how the company is expanding beyond basic activity tracking, although no watch or health integration has been confirmed for SM-U600.

Open hooks may suit outdoor exercise and people who dislike silicone tips. Sealed earbuds remain better suited to flights, offices, public transportation, and other settings where isolation is the priority.

For organizations standardizing Samsung mobile hardware, an open-ear option could suit field employees who need audio prompts while remaining aware of traffic, colleagues, or equipment. Procurement teams would still need confirmed durability, battery, compatibility, support, and pricing details before treating SM-U600 as a workplace accessory.

Samsung has not confirmed the earbuds’ name, specifications, price, or release date. The next useful evidence will need to show whether the hook and clip renders depict one changing SM-U600 design—or two separate products.

Also read: Earlier Galaxy Watch9 and Ultra2 leaks show which reported details matched Samsung’s eventual announcement — and why unreleased hardware specifications should still be treated cautiously.

>

Continue Reading

Trending

Copyright © 2017 Zox News Theme. Theme by MVP Themes, powered by WordPress.