OpenAI has launched its first physical hardware product, the Codex Micro, a specialized macropad designed to control AI-driven coding workflows. Developed in partnership with boutique manufacturer Work Louder, the $230 device is a modified version of the existing Creator Micro 2 platform. It features 13 low-profile mechanical keys, a rotary dial, and a joystick housed within a CNC-machined aluminum and polycarbonate body.

Source

A regression in the open-source AMDGPU driver is causing significant performance degradation across several major Linux distributions, including Ubuntu, Fedora, and Arch. The issue stems from a bug in the Linux 7.0 kernel series that mismanages GPU power profiles and clock settings. This flaw can allow GPU core frequencies to exceed manufacturer limits, triggering the kernel to throttle the device to prevent damage.

Source

IBM has introduced the Power S1112, a compact one-socket server designed for edge computing and distributed environments. Built on the Power11 architecture, the entry-level system features on-chip Matrix Math Acceleration to support local AI inference and analytics. The hardware is available in rack or tower configurations and allows administrators to run Linux or AIX partitions alongside traditional IBM i environments.

Source

OpenAI is developing a screenless, mobile smart speaker designed to function as a human-like AI companion for the home. The device features mechanical elements that allow it to move autonomously, creating a more lifelike presence than traditional smart speakers. It utilizes integrated cameras and sensors to maintain contextual awareness of its surroundings and the user's activities.

Source

Apple is exploring a partnership with PrismML, a Caltech spinout, to integrate extreme model-compression technology into future iPhone hardware. The startup utilizes ultra-low-bit weight architectures, such as 1-bit and ternary representations, to drastically reduce the memory footprint of large language models. This mathematical approach allows a 27-billion-parameter model like Alibaba’s Qwen 3.6 to shrink from 54 GB to under 4 GB, making it compatible with the 8 GB of RAM found in current Pro-tier iPhones.

Source

Nvidia is transitioning from a GPU manufacturer to a comprehensive AI infrastructure provider by launching the Vera CPU platform. This new processor is specifically engineered to handle agentic AI systems and intensive data center workloads. By moving beyond graphics accelerators, the company aims to capture a larger share of the traditional server market currently held by x86 competitors.

Source

Microsoft has revoked 11 legacy, Microsoft-signed UEFI shim bootloaders that allowed attackers to bypass Secure Boot protections. These outdated binaries, primarily version 0.9 and earlier, remained trusted by the "Microsoft Corporation UEFI CA 2011" certificate despite containing long-known vulnerabilities. By utilizing a "Bring Your Own Vulnerable Bootloader" (BYOVB) attack, threat actors could execute unauthorized code during the early boot phase before the operating system initializes.

Source

Nokia and Nvidia have announced a commercial AI-native radio access network (AI-RAN) platform designed to significantly increase data transmission capacity. The solution integrates Nokia’s AirScale base-station architecture with Nvidia’s accelerated computing silicon and CUDA-based AI libraries. This partnership aims to transform traditional wireless networks into software-defined environments capable of processing data closer to the end user.

Source

The release of Debian 12.15 marks the final point update for the "Bookworm" series as it transitions to long-term support. This milestone signifies the end of mainstream support for the native 32-bit x86 (i386) architecture within the Debian ecosystem. While Debian 12 will receive security maintenance until mid-2028, no further point releases are planned for this version.

Source

Apple’s federal lawsuit against OpenAI is strategically designed to gain visibility into the startup’s secretive hardware development through the legal discovery process. Legal experts suggest that Apple lacks direct evidence of how its stolen trade secrets are being utilized behind OpenAI's closed doors. By initiating litigation, Apple can legally compel the disclosure of internal documents and engineering plans that would otherwise remain confidential.

Source

ESET researchers have identified 11 vulnerable UEFI shim bootloaders that allow attackers to bypass Secure Boot on nearly any UEFI-based system. These shims, some dating back to 2013, remained signed by Microsoft despite containing critical flaws or authorizing vulnerable secondary components. Because these binaries were never revoked, attackers can use them to execute untrusted code and deploy persistent bootkits like BlackLotus or BootKitty.

Source

OpenAI is developing its first consumer hardware product, a portable smart speaker designed to function as a human-like AI companion. The screenless device will utilize ChatGPT and GPT-Live models to provide personalized assistance, media playback, and smart home control. Internal descriptions suggest the product is a new category of computer built specifically for AI, featuring a camera and sensors to maintain contextual awareness of its surroundings.

Source

Apple has filed a federal lawsuit against OpenAI and several former employees, alleging a systematic campaign to steal trade secrets for unreleased hardware products. The complaint claims OpenAI recruited over 400 former Apple staff and encouraged candidates to bring physical components to "show and tell" interview sessions. This legal battle highlights a shift in the industry where competitive advantage is moving from raw compute access toward proprietary institutional knowledge and human expertise.

Source

Microsoft has launched the commercial editions of the Surface Pro 13 and Surface Laptop 8, featuring Qualcomm’s Snapdragon X2 processors. These business-specific models ship with Windows 11 Pro and are designed for enterprise procurement channels and volume licensing. Administrators can manage these endpoints using standard commercial tools while benefiting from integrated security features like hardware-based privacy screens on select laptop models.

Source

Organizations are increasingly viewing AI-enabled PCs as a strategic tool to mitigate the unpredictable and soaring costs of cloud-based generative AI. While early adoption was slow, the rise of "tokenomics" has forced businesses to scrutinize expensive API bills where every prompt adds to a recurring operational expense. This shift marks a transition from a perpetual rental model to a capital investment in edge hardware that runs workloads locally for the cost of electricity.

Source

Microsoft has implemented a compatibility hold on the July 2026 Patch Tuesday update, KB5101650, for specific Dell computers. This decision follows reports of severe stability issues, including unexpected system shutdowns and significant performance degradation. The problem is currently limited to Intel-based Dell devices running Windows 11 versions 24H2 and 25H2.

Source

Apple is restructuring its silicon roadmap to prioritize artificial intelligence, reportedly skipping planned M6 Pro and Max variants to fast-track the M7 generation. While a base M6 chip is expected for entry-level Macs in late 2026, the company is shifting resources toward an AI-optimized M7 family slated for 2027. This strategic pivot aims to transform the Mac into a powerhouse for local AI inference, reducing the need for cloud-based processing.

Source

Researchers at Peking University have developed an all-optical interconnect system that links standard electronic chips to significantly accelerate artificial intelligence processing. This breakthrough utilizes specialized hardware to replace traditional electrical connections between chips with high-speed optical "joints." By integrating these optical links with custom algorithms, the system achieves AI inference speeds up to 100 times faster than conventional methods.

Source

OpenAI CEO Sam Altman and Elon Musk recently engaged in a public dispute regarding the viability of space-based computing infrastructure. Altman characterized the promotion of orbital data centers as a short-term sales pitch aimed at public market investors rather than a realistic technological milestone. He previously noted that space-based compute is unlikely to provide meaningful capacity for AI workloads for at least another decade.

Source