MacBook Neo is Apple’s $599 laptop powered by A18 Pro

Apple has introduced MacBook Neo, a new entry-level MacBook built around the A18 Pro SoC . Apple says pre-orders start today, with retail availability set for Wednesday, March 11, 2026.

MacBook Neo features a 13-inch Liquid Retina display rated at 2408×1506 and 500 nits, plus support for 1 billion colors. Apple lists four

AMD confirms Ryzen AI 400G desktop APUs offer up to 12 usable PCIe 4.0 lanes

Last week, AMD announced the Ryzen AI PRO 400 series for laptops and desktops, along with the Ryzen AI 400 series for desktops. These are Zen 5-based processors positioned as successors to the Ryzen 8000G series, budget-focused APUs with integrated graphics. AMD has not clearly communicated which models are launching and what the final specifications are.

AMD has now listed the

Chuwi CoreBook X reportedly ships with Ryzen 5 5500U while advertised as Ryzen 5 7430U

© NotebookCheck

Notebookcheck reports that its Chuwi CoreBook X review unit, advertised as shipping with an AMD Ryzen 5 7430U, actually contained a Ryzen 5 5500U. The outlet says the discrepancy was not only present at the hardware level, but also hidden by system reporting so the laptop identifies itself as the newer 7430U in common tools. Worth noting that

Only Radeon and RTX 5060 Ti price 8GB remains unchanged

Remember ASUS’s January notice about an upcoming price change? ASUS later tried to pull those posts down, saying the message was meant for partners, not consumers. We left our coverage up because the result is the same, pricing gets adjusted. The original note also did not mention which products would be affected, but graphics cards were an obvious candidate.

Prices have moved up since January. You can see it across most GeForce listings, without

DDR5 scalping surge: DataDome says bots tuned request rates to evade alarms

DataDome says DDR5 supply constraints tied to AI demand are creating room for scalpers that rely on automated scraping. The company says bots are hitting DDR5 product pages about six times more often than real shoppers.

The report examines one operation that DataDome says generated more than 10 million scraping requests before being blocked. In a one-hour sample, it saw about 50,000 requests across 91 listings, averaging roughly 55

EA focusing on ARM

With the NVIDIA N1 series imminent, one of the largest publishers has started efforts to ensure that their anticheat system is native on ARM.

EA has posted a job opening that points to native ARM64 support for its kernel-level EA Javelin Anticheat, a move that lines up with the wider push toward Windows on Arm gaming hardware. The timing also matches growing attention around NVIDIA’s rumored N1 and N1X PC processors, which are expected to bring Arm CPU cores into more

Highguard to be unreleased 45 days after launch

Wildlight Entertainment says it will permanently shut down Highguard on March 12, 2026 . The free-to-play squad shooter launched on January 26, 2026, making the shutdown less than two months after release.

In its announcement, the studio said more than 2 million people played Highguard, but it could not build a “sustainable player base” to support the game long term.

Highguard will receive one final update before it goes offline

MSI developed RTX 5090 D v2 LIGHTNING for Chinese market

© Bilibili

A Bilibili user going by “Hardware Patrick Star” has shared what appears to be an MSI GeForce RTX 5090 D v2 LIGHTNING. The video shows a 24GB model tied to the China-only RTX 5090 D v2, which uses cut-down RTX 5090 specifications.

MSI previously said only 1,300 units of its LIGHTNING card would be made, and the

Micron memory for AI (obviously)

Having exited the consumer market , Micron is currently concentrating on improving the appeal of its AI product lineup. In this case, capacity is quite an important spec that may attract hyperscalers. And large capacity they delivered, on a single module.

Today Micron introduced a 256GB LPDRAM SOCAMM2 module designed for data center and AI server platforms. The company says the new module targets memory capacity and bandwidth requirements for modern workloads, including machine learning and in-memory analytics.