• Which processors will the GPD BOX use?

  There are two processor options: the Intel Core Ultra X7 358H and the Intel Core Ultra 7 356H. Both are part of Intel’s Core Ultra Series 3, also known as Panther Lake.

• What is the difference between the Core Ultra X7 358H and the Core Ultra 7 356H versions?

  Both chips have 16 cores, but the higher-tier Core Ultra X7 358H is listed with up to 4.8GHz boost, 12 Xe3 graphics cores, and up to 180 TOPS of AI performance. The Core Ultra 7 356H is listed with up to 4.7GHz boost, 4 Xe3 graphics cores, and up to 100 TOPS.

• How much RAM does the GPD BOX have?

  A choice of 32GB and 64GB LPDDR5x 8533 MT/s memory options are available for the GPD BOX.

• Can the storage be upgraded?

  Yes, thre are two M.2 slots. One PCIe Gen 5 x4 slot and one PCIe Gen 4 x2 slot.

• Does the GPD BOX support PCIe Gen 5 storage?

  Yes. There is one dedicated M.2 PCIe Gen 5 x4 slot, which means the BOX supports very fast NVMe storage

• What ports does the GPD BOX have?

  There are two USB4 v2.0 ports, one DisplayPort 2.1 output, one HDMI 2.1 port, four USB 3.2 Gen 2 Type-A ports, two 2.5G RJ45 Ethernet ports, and one MCIO 8i port.

• Does the GPD BOX support multiple monitors?

  There are HDMI 2.1, DisplayPort 2.1, and dual USB4 v2.0 for connecting to multiple monitors.

• Why does the GPD BOX have dual 2.5G LAN ports?

  Dual 2.5G Ethernet gives the BOX more flexibility than most mini PCs. It can be useful for advanced networking, creator workflows, office setups, direct device links, router or firewall-style use cases, lab environments, or anyone who simply wants faster wired networking with more options

• What is MCIO on the GPD BOX?

  MCIO is the high-bandwidth expansion connection based on PCIe 5.0 x8, and presents it as a major part of the BOX’s appeal for external graphics and future expansion

• How is MCIO different from OCuLink?

  The MCIO offers much more bandwidth in this implementation. The BOX’s MCIO 8i link provides up to 512Gbps bidirectional bandwidth and offers four times the bandwidth of OCuLink based on PCIe 4.0 x4.

• Does the GPD BOX support external GPUs?

  Yes, that is one of its main selling points. The GPD BOX as a compact desktop that can scale up with external GPU support, especially through MCIO and also through USB4.

• Can the GPD BOX connect to the GPD G2 eGPU dock?

  Yes. In fact, the GPD BOX and GPD G2 are being presented as a matched pairing. The BOX acts as the compact base system, while the G2 adds desktop-class graphics expansion and extra connectivity.

• Do you need the GPD G2 to use the GPD BOX?

  No. The GPD BOX is still a complete mini PC on its own. The G2 is an optional expansion path for users who want much stronger graphics performance or a more modular desktop-style setup.

• Can the GPD BOX use an eGPU over USB4 as well as MCIO?

  GPD BOX hastwo USB4 v2.0 ports and also promotes MCIO as the higher-bandwidth option. In practical terms, USB4 should offer broader compatibility, while MCIO is the more performance-focused path with the GPD G2.

• Which connection is likely to be better for eGPU performance, MCIO or USB4?

  GPD is specifically promoting MCIO as the higher-bandwidth solution and one of the key reasons the BOX stands out versus more conventional mini PCs that rely only on USB4 or Thunderbolt-class external graphics support

• What gaming performance can you expect from the GPD BOX without an eGPU?

  The higher-end model can deliver 60 to 120 FPS in most AAA titles on integrated Xe3 graphics, while the lower-tier model is framed as being around Radeon 780M-class performance. Those claims are promising, but not yet independently verified.

• Is the GPD BOX good for emulation?

  It looks promising for emulation, especially in the higher-tier configuration, because it combines a modern 16-core CPU design with stronger integrated graphics than older mini PCs. Still, final expectations should wait for real-world testing across different emulators

• Is the GPD BOX suitable for content creation work?

  The combination of modern Core Ultra processors, fast storage options, strong connectivity, and possible eGPU pairing could make it attractive for editing, media work, and general creative workloads.

• Can the GPD BOX handle local AI workloads?

  You can expect up to 180 TOPS on the higher-end model and up to 100 TOPS on the lower-tier model, with marketing around local LLM inference, multimodal agents, and scientific or simulation workloads. As always, those use cases will need independent testing to judge practical performance.

• What operating system will the GPD BOX run?

  It will run Windows 11 and likely come with it pre-installed. Other OS should run but not officially supported.

• Will the GPD BOX support Linux?

  Given the hardware class, Linux support is likely to be of interest, but buyers should wait for official confirmation and user testing.

• Has GPD confirmed the GPD BOX launch date?

  No launch date has been confirmed yet.

• How much will the GPD BOX cost?

  Prices have not yet been confirmed.

• How much power does the GPD BOX use?

  A configurable TDP range of 15W to 80W for both processor options. That gives a general idea of the BOX’s operating envelope, although full system power draw at idle and under load has not yet been published

• What warranty will the GPD BOX have?

  GPD STORE provides a two year warranty on all GPD products, and one year for others.

• Who is the GPD BOX best suited for?

  The GPD BOX looks best suited to buyers who want a premium mini PC with stronger-than-average connectivity, fast storage, and room to scale into a more powerful setup later with an eGPU. It should be especially interesting to gamers, creators, developers, and compact desktop enthusiasts who like small systems but do not want to give up expansion options.

• What makes the GPD BOX different from a typical mini PC?

  Its standout features are the use of Intel Panther Lake processors, dual USB4 v2.0, an MCIO 8i port for high-bandwidth expansion, dual 2.5G Ethernet, and support for pairing with the GPD G2 eGPU dock. GPD is clearly pitching it as a premium compact system rather than a basic office mini PC.

• What is the GPD BOX?

  The GPD BOX is a newly announced mini PC that GPD is positioning as its first entry into the compact desktop category. It is designed as a small, high-spec system that can work on its own or be paired with high-bandwidth external graphics expansion

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

Sorry, no FAQ matched your criteria.

• What is the GPD G2?

  The GPD G2 is an external GPU docking station with the world’s first MCIO dual-port eGPU dock. It is designed to connect compact systems to desktop graphics cards while also adding extra connectivity and storage options

• What is the main purpose of the GPD G2?

  Its main job is to let compatible devices use desktop-class PCIe graphics cards externally. It is a high-bandwidth graphics expansion platform rather than just a simple dock.

• What makes the GPD G2 different from a typical eGPU dock?

  The biggest differentiator is MCIO. The MCIO 8i plus USB4 v2.0 connectivity, along with a much more serious power design than most compact eGPU docks.

• Which devices is the GPD G2 designed to work with?

  The GPD G2 can be used for home PCs, high-end gaming handhelds, laptops, workstations, servers, and even some HPC-style use cases. The most relevant requirement is compatible connectivity, especially MCIO or USB4-class support depending on the device.

• Does the GPD G2 only work with GPD devices?

  No. It can be used with a range of PCs and handhelds, not only GPD hardware

• What is MCIO on the GPD G2?

  MCIO is the high-bandwidth connection GPD is using to push the G2 beyond a more conventional USB4-only eGPU dock. On the G2, the MCIO 8i link supports PCIe devices, including graphics cards, via PCIe Gen 4 x8

• What is the difference between using MCIO and USB4 on the GPD G2?

  MCIO is the more performance-focused path, while USB4 adds compatibility and dock-style functionality. The USB4 connection handles eGPU functionality when used independently, while simultaneous MCIO plus USB4 connection is needed to enable hub functionality alongside MCIO

• Can the GPD G2 work as more than just a GPU dock?

  Yes. Besides graphics support, it also adds USB ports, Ethernet, and an M.2 storage slot

• What ports does the GPD G2 have?

  There is one USB4 v2.0 port, two USB 3.2 Gen 2 Type-A ports, one RJ45 Ethernet port, one MCIO 8i connection, and one M.2 2280 slot

• Does the GPD G2 have Ethernet?

  Yes. It includes one RJ45 Ethernet port rated at 10/100/1000 Mbps auto-negotiation.

• Does the GPD G2 have USB-A ports?

  Yes. there are two USB 3.2 Gen 2 Type-A ports running up to 10Gbps.

• Can the M.2 slot on the GPD G2 be used as bootable external storage?

  Yes. the M.2 slot supports external storage booting.

• How fast is the M.2 storage slot on the GPD G2?

  It has a M.2 2280 slot running at PCIe 3.0 x2, converted from USB 3.2 Gen 2. That means it adds useful storage expansion, but it is not being positioned as the highest-speed NVMe implementation.

• Does the GPD G2 include a GPU?

  No, you would need to buy your own GPU.

• What graphics card slot does the GPD G2 have?

  It is a PCIe Gen 5 x16 slot.

• Does the GPD G2 support NVIDIA GeForce RTX 4090 or RTX 5090-class GPUs?

  Yes, including 4090 and 5090-class models. 

• Does the GPD G2 support AMD and Intel graphics cards too?

  Yes. There is broad support across AMD Radeon and Intel Arc product families as well as NVIDIA cards.

• Does the GPD G2 support workstation or professional GPUs?

  The GPD G2 supports borth gaming and professional GPUs. NVIDIA’s lineup spans multiple generations, starting from the GeForce GTX 10 Series (Pascal) and RTX 20 / GTX 16 Series (Turing), progressing through the RTX 30 Series (Ampere) and RTX 40 Series (Ada Lovelace), and concluding with the latest RTX 50 Series (Blackwell). AMD’s roster traces its evolution from the older Radeon RX 400 / 500 Series (GCN 4.0) and RX Vega Series (GCN 5.0) into its modern era with the RX 5000 (RDNA 1.0), RX 6000 (RDNA 2.0), RX 7000 (RDNA 3.0), and RX 9000 Series (RDNA 4.0). Finally, Intel is represented by its dedicated Arc graphics families: the Arc A Series (built on the Xe HPG architecture) and the Arc B Series (utilizing the newer Xe2 HPG architecture). 

  For professional GPUs. NVIDIA’s professional evolution spans from the earlier Quadro K (Kepler), M (Maxwell), and P (Pascal) series to the modern RTX A Series (Ampere) and RTX PRO Series (Blackwell). AMD’s workstation offerings transition from the legacy Radeon Pro WX and W Series (GCN architecture) through the modern Radeon Pro W6000 (RDNA 2) and W7000 (RDNA 3) series, culminating in the Radeon AI PRO R9000 Series (RDNA 4). Finally, Intel is represented by its dedicated professional GPUs: the Arc Pro A Series built on the Xe architecture and the succeeding Arc Pro B Series utilizing the newer Xe2 architecture. 

  NVIDIA’s enterprise solutions span from the legacy Tesla Series (encompassing multiple architectures including Kepler, Maxwell, Pascal, Volta, and Turing) to its modern, high-performance PCIe accelerators, such as the A100 (Ampere), H100/H200 (Hopper), and the B100 (Blackwell). AMD’s compute-focused Instinct lineage evolves from its earlier GCN-based cards (MI6, MI8, and MI25) into its purpose-built CDNA architectures, represented by the MI100 (CDNA 1), MI200 Series (CDNA 2), MI300 Series (CDNA 3), and the MI350 Series (CDNA 4). Finally, Intel’s data center portfolio includes the Flex and Max Series GPUs (utilizing Xe-HPG and Xe-HPC architectures), the specialized Gaudi 3 AI Accelerators, and the “Crescent Island” platform powered by the Xe3P architecture. 

• Can the GPD G2 be used for AI, compute, or server-style workloads?

  Yes, workstation, server, and HPC-style scenarios rather than only gaming or creator use.

• How much power can the GPD G2 deliver?

  The GPD G2 uses a Gold-rated ATX 3.1 power design with 800W continuous total 12V output.

• What power connector does the GPD G2 use for GPUs?

  It has a 12V-2×6 GPU power connector and complies with ATX 3.1 GPU requirements.

• Does the GPD G2 need its own power supply?

  Yes. This is not a bus-powered accessory. The power supply supports 100 to 240V AC input, a Gold-rated ATX 3.1 design, and an IEC C14 power socket aka ‘kettle lead’ as you would find on a standard desktop PC power supply.

• Can the GPD G2 charge a handheld or laptop?

  Yes, up to a point. The USB4 v2.0 port supports PD 3.0 charging up to 100W.

• What does the USB4 power delivery support on the GPD G2 mean in practice?

  It means the dock can potentially power or charge a compatible handheld or laptop over USB4 while also acting as an expansion device. There is support for 5V, 9V, 12V, 15V, and 20V power delivery profiles up to 100W.

• What performance can you expect from the GPD G2?

  GPD’s headline claim is that an RTX 4090 would see only around a 2% performance loss on the G2. That sounds impressive, but it should be treated carefully until fuller benchmark methodology and independent reviewers validate it.

• Is the GPD G2 better than a typical USB4 eGPU dock?

  The MCIO-focused design and much more aggressive bandwidth and power positioning suggest a step above more limited USB4-only docks

• What operating systems does the GPD G2 support?

  Windows 11 and Linux distributions 

• What are the GPD G2’s dimensions and weight?

  The dimensions are 157.3 × 119.8 × 182 mm (6.19 x 4.71 x 7.16 inches) with a net weight of 1605 g (3.5 lbs).

• What materials is the GPD G2 made from?

  It is made from aluminum-magnesium alloy and PC 94V-0.

• When is the GPD G2 release date?

  No release date has been announced yet.

• Has the GPD G2’s price been announced?

  Currently not yet. We hope to have final pricing soon.