Tag: AI mini PC

  • Ryzen AI 9 HX 470 Mini PC: 86 TOPS, OCuLink, 256GB RAM

    Ryzen AI 9 HX 470 Mini PC: 86 TOPS, OCuLink, 256GB RAM

    Key Takeaway

    – 86 TOPS total AI compute power (55 TOPS NPU + 31 TOPS CPU/GPU) enables local AI model running without cloud token costs.
    – Powered by AMD Ryzen AI 9 HX 470 CPU, with up to 256 GB DDR5 RAM and up to 6 TB NVMe storage.
    – Integrated Radeon 890M GPU (≈ Nvidia 1650 performance), with OCuLink port for eGPU upgrade up to RTX 5090.
    – MSRP $1,599, currently on sale for $1,049 (32 GB RAM / 1 TB SSD model).


    New Mini PC Packs AI Power

    Bosgame has released the VTA-439 mini PC designed to run AI applications and services such as OpenClaw locally, saving owners the cost of purchasing AI tokens. The mini PC is powered by a 12C/24T AMD Ryzen AI 9 HX 470 CPU, with up to 256 GB of DDR5 SO-DIMM 5600MT/s RAM in two slots and up to 6 TB of M.2 2280 NVMe PCIe4.0 x4 storage in three slots (based on currently available 2 TB SSDs).

    AI Compute and Graphics Capabilities

    Importantly, the HX 470 offers a total of 86 TOPS of AI compute power (55 TOPS NPU + 31 TOPS CPU/GPU), enabling the VTA-439 mini PC to run large AI models, OpenClaw, and AI applications for image and video generation. For Windows 11 applications and games, the mini PC uses an integrated AMD Radeon 890M, which hovers around an Nvidia 1650 in performance, and can be boosted by connecting an external GPU (eGPU) to its OCuLink PCIe 4.0 x4 port, including cards as powerful as the latest Nvidia RTX 5090.

    Pricing and Storage Options

    The VTA-439 with 32 GB RAM and 1 TB SSD has an MSRP of $1,599.00, but is currently on sale for $1,049 on the Bosgame website. This device supports up to 256 GB of DDR5 SO-DIMM 5600MT/s RAM and up to 6 TB of M.2 2280 NVMe PCIe4.0 x4 storage, giving users lots of flexibility for demanding workloads. The Ryzen AI 9 HX 470 chip ensures fast processing for AI tasks, while the integrated Radeon 890M handles everyday graphics needs.

    External GPU Support and Performance

    The OCuLink PCIe 4.0 x4 port allows users to connect an eGPU, boosting graphical performance significantly. While the integrated Radeon 890M is comparable to a Nvidia 1650, adding a card like the RTX 5090 can transform this mini PC into a high-end gaming machine or workstation. This makes the VTA-439 a versatile option for both AI development and entertainment, all while eliminating the need for cloud-based AI tokens.

    Final Thoughts on the VTA-439

    Bosgame’s new mini PC offers a compelling package for enthusiasts who want local AI processing without recurring costs. With up to 86 TOPS of compute power, three M.2 slots, and eGPU support, it is positioned as a powerful yet compact solution. The current sale price of $1,049 provides good value for the hardware, though the MSRP of $1,599 is steep for a device with an integrated GPU.


    Sources

  • Bosgame VTA-439 Mini PC: Ryzen AI 9, 86 TOPS, 256GB RAM

    Bosgame VTA-439 Mini PC: Ryzen AI 9, 86 TOPS, 256GB RAM

    Key Takeaway

    – Runs AI applications like OpenClaw locally with 86 TOPS total AI compute (55 TOPS NPU + 31 TOPS CPU/GPU).
    – Powered by AMD Ryzen AI 9 HX 470 CPU, supports up to 256 GB DDR5 RAM and 6 TB NVMe storage.
    – Integrated Radeon 890M GPU (≈ Nvidia 1650 performance), upgradable via OCuLink eGPU port (supports up to RTX 5090).
    – Priced at $1,049 (sale) down from $1,599 MSRP for the 32GB RAM / 1TB SSD model.


    Bosgame Launches New AI Mini PC

    Bosgame has unveiled there VTA-439 mini PC, a device specifically engineered for running AI applicationns and services like OpenClaw locally, which potentially help owners circumvent the cost of purchasing AI tokens. This little machine is quite a powerhouse for its size, offering alot of compute capability for various tasks. It’s aimed at users who want to do heavy lifting without relying on cloud services.

    Powerful Specifications Under the Hood

    This mini PC is driven by a 12C/24T AMD Ryzen AI 9 HX 470 CPU, which can be paired with up to 256 GB of DDR5 SO-DIMM 5600MT/s RAM installed across two slots. For those needing ample storage, it supports up to 6 TB of M.2 2280 NVMe PCIe4.0 x4 storage across three slots, based on currently available 2 TB SSDs, providing plenty of space for large model files. The system is designed to not bottleneck even demanding workflows, making it a serious contender for local AI workloads.

    Importantly, the HX 470 offers a total of 86 TOPS of AI compute power, broken down into 55 TOPS from the NPU and 31 TOPS from the CPU/GPU. This enables the VTA-439 to smoothly run large AI models, OpenClaw, and various applications for image and video generation directly on the device. The levels of performance here allows for local inferencing without relying on internet connectivity.

    Graphics Performance and Connectivity

    For Windows 11 applications and games, the system utilizes an integrated AMD Radeon 890M graphics, which provides performance comparable to an Nvidia GTX 1650. Users can significantly boost this by connecting an external GPU (eGPU) to its OCuLink PCIe 4.0 x4 port, which supports powerful cards including the latest Nvidia RTX 5090. This makes it a very flexible setup for both productivity and occasional gaming, provided you have the budget for an eGPU enclosure.

    Pricing and Availability Details

    The VTA-439 configuration with 32 GB of RAM and a 1 TB SSD has an MSRP of $1,599.00, but is currently available for a lower price of $1,049 on the Bosgame website. This significant discount makes it a more accessible option for developers and hobbyists alike, though it’s still a considerable investment for most people. The pricing reflects its positioning as a premium device in the mini PC market, focusing on AI capabilities rather than being a budget option.


    Sources

  • Asus Ascent QN10 Mini PC: Qualcomm Snapdragon X2 Elite AI

    Asus Ascent QN10 Mini PC: Qualcomm Snapdragon X2 Elite AI

    Key Takeaway

    – 18-core Snapdragon X2 Elite with 80 TOPS AI compute power
    – Up to 32GB LPDDR5x RAM and 4TB NVMe SSD storage
    – Runs Windows on ARM with native OpenClaw support
    – No Nvidia GPU, limiting AI model compatibility
    – Wi-Fi 7, Bluetooth 5.4, 2.5 GbE, and three USB4 ports


    Asus Ascent QN10 Mini PC Launches with Snapdragon Power

    The Asus Ascent QN10 mini PC has arrived targeting businesses and AI developers with an 18-core Snapdragon X2 Elite X2E-88-100 CPU offering 80 TOPS of AI compute power, up to 32 GB of LPDDR5x 9600 RAM, and up to 4 TB of M.2 2280 NVMe SSD storage. Developers have access to the Qualcomm AI Hub, a repository of select AI models and sample apps to use in training and developing customized ones, and can run OpenClaw natively on the CPU.

    Software and Compatibility Challanges

    The Ascent runs Windows on ARM, a variant of regular Windows 11 for x64 CPUs (AMD or Intel). The Snapdragon SoC offers performance at potentially less power than AMD or Intel offerings by using a design based on ARM CPUs commonly found in modern smartphones. The recently released MacBook Neo that Notebookcheck reviewed also uses a smartphone processor in a computer.

    Limitations Without Nvidia Graphics

    However, the AI mini PC does not have an Nvidia GPU, so it can’t run the much larger library of AI models, sample apps, and other AI services that Nvidia offers. Windows on ARM is also limited in its app, driver, and peripheral compatibility versus regular Windows 11 for x64 (AMD or Intel).

    Connectivity and Port Selection

    The QN10 comes with Wi-Fi 7, Bluetooth 5.4, and 2.5 GbE connectivity. It also has three USB4 ports, three USB-A 3.2 ports, one USB-A 2.0 port, one HDMI 2.1 FRL port, and a headphone jack.

    • Wi-Fi 7 and Bluetooth 5.4
    • 2.5 GbE Ethernet
    • Three USB4 ports
    • Three USB-A 3.2 ports
    • One USB-A 2.0 port
    • One HDMI 2.1 FRL port
    • Headphone jack

    Dimensions and Weight Details

    The mini PC measures 130 x 130 x 40 mm (5.1 x 5.1 x 1.6 in.) and weighs 720 g (1.6 lbs.).

    Sources
  • MSI RTX Spark Mini PC rivals Mac Studio, Asus, HP Mini PCs

    MSI RTX Spark Mini PC rivals Mac Studio, Asus, HP Mini PCs

    Key Takeaway

    – The MSI EdgeMesa N AI+ is a compact Mini PC designed for AI developers and data scientists.
    – It enables local AI processing, reducing latency, improving data privacy, and cutting cloud dependency.
    – It handles resource-heavy tasks like LLMs, generative AI, and real-time inferencing.
    – High-speed connectivity makes it suitable for industries like robotics, smart cities, healthcare, and finance.
    – Expected release window is August or September 2026 (no price or exact date announced yet).


    MSI EdgeMesa N AI+ Mini PC Officially Unveiled

    MSI has formally taken the wraps off the EdgeMesa N AI+ Mini PC, and this new compact desktop targets AI developers, data scientists and also power users. It’s engineered to bring heavy AI tasks closer to home.

    Hardware That Packs a Punch Despite a Clunky Name

    The EdgeMesa N AI+ joins the newly-announced HP OmniDesk Mini PC plus the Asus ProArt Mini PC, becoming one of the first RTX Spark Mini PCs announced at Computex 2026. Even if the name MSI has gone with doesn’t quite roll off the tongue, the hardware is definately worth talking about. Unsurprisingly, this system is built to move AI workloads from cloud to local desktop, which is a big shift for pro users.

    Local AI Processing for Privacy and Speed

    With local AI processing, the system tackles key pain points for professional users. Reduced latency, heightened data privacy and a significant reduction in reliance on external cloud infrastructure make it stand out. MSI claims the system is optimised to handle resource-heavy tasks, including Large Language Models (LLMs), generative AI and real-time inferencing with ease. This makes it a versatile tool for demanding workflows.

    High-Speed Connectivity and Industry Adaptability

    Beyond its raw computational prowess, MSI has packed the EdgeMesa N AI+ with high-speed connectivity options that make it adaptable to specialized industries. From robotics and smart city infrastructure to healthcare and financial analytics, the machine is designed to act as a robust local node for AI deployment, says MSI. The company is currently showcasing this mini PC at Computex 2026 if you happen to be there.

    Pricing and Availability Details

    The price and launch date will be announced at a later date. However, like with all other RTX Spark laptops and PCs, you might have to wait untill at least August or September to get your hands on this tiny powerhouse. No specific pricing or specification details have been confirmed yet, but MSI promises more information closer to launch.


    Sources

    • Jupiter 2 Mini PC: 10Gbps Network Port, 60 TOPS Performance

      Jupiter 2 Mini PC: 10Gbps Network Port, 60 TOPS Performance

      Key Takeaway

      – Robust developer-focused mini PC with AI capability: Spacemit K3 SoC (eight A100 cores) offering up to 60 TOPS AI performance and RVA23 support, suitable for local AI workloads.
      – Strong connectivity and expansion: SFP+ 10 Gbps, Gigabit Ethernet, WiFi 6, Bluetooth 5.2, UFS, M.2 2280 SSD over PCIe 3.0 x4, USB-C with 4K60 DP Alt Mode, and optional cellular module.
      – Flexible display and power options: eDP 2.5K@90Hz, USB-C DP Alt Mode 4K60, USB PD or 12V with ATX pins; Fedora and Ubuntu 26.04 supported.


      Overview

      The Jupiter2 is a new mini PC from Milk-V. It is currently available for pre-order from Arace for $300. Customers will therefore still need some patience. “Customers” is already a useful keyword here, since this is unlikely to be a system for typical consumers and is instead more of a developer platform for the RISC-V architecture.

      Processor and AI

      The SoC is a Spacemit K3, which has eight A100 compute cores. AI performance is listed at up to 60 TOPS, which means AI applications should also be able to run locally. The SoC supports the RVA23 instruction set.

      Networking and storage

      The configuration is fairly solid, and the networking in particular is very strong. SFP+ supports network connectivity at up to 10 Gbps, which means roughly 1.2GB can be transferred per second. Gigabit Ethernet is also included, while WiFi 6 and Bluetooth 5.2 are supported as well. UFS storage can be used, and an M.2 2280 SSD can be installed. It is connected via four PCIe 3.0 lanes.

      Ports and expansion

      There is an eDP port that can output 2.5K content at up to 90Hz. There is also a USB-C port, which supports 4K60 output via DisplayPort Alt Mode. A SIM card slot is available, and a cellular module can be added later. The system offers two USB 3.2 Gen 1 ports in USB-C form and four USB 2.0 Type-A ports. Power can be supplied via USB PD or 12 volts, in which case two ATX pins are used. Fedora and Ubuntu 26.04 are also supported as operating systems.

      Arace, Milk-V


      Sources