WiCi One Debuts as First Wireless GPU Enclosure: RTX 5090 Gaming Over Wi-Fi 7 Starting at $1,999

WiCi One Introduces Wireless GPU Technology Using Wi-Fi 7

WiCi has launched the WiCi One, positioning it as the world's first wireless local-GPU solution that eliminates the need for physical display cables by transmitting GPU capabilities over Wi-Fi 71

1

2

2

. The wireless GPU enclosure comes in two configurations: a baseline model featuring an RTX 5060 Ti with 16GB VRAM and Core i3 100U CPU priced at $1,999, and a flagship model with an RTX 5090 with 32GB VRAM and Core Ultra 7 255H CPU at $2,5991

1

2

2

. Early signups receive discounts bringing the entry price down to $1,9992

2

.

Gaming Performance Wirelessly Reaches 110 FPS in Cyberpunk 2077

The eGPU over Wi-Fi 7 solution demonstrates impressive gaming performance through its virtual GPU driver, with the RTX 5090 model achieving up to 110 FPS in Cyberpunk 2077 at 4K Ultra settings with DLSS 4 enabled, and 90 FPS at 4K Cinematic with DLSS 41

1

. The RTX 5060 Ti configuration delivers 80 FPS and 60 FPS respectively under the same conditions1

1

. Wi-Fi 7 provides sufficient bandwidth with a theoretical maximum of 46Gbps, exceeding DisplayPort 1.4a's 32.4 Gbps capacity1

1

. WiCi modified standard Wi-Fi protocols to optimize for low latency by transmitting data asynchronously, making Wi-Fi 7 viable as a wireless replacement for USB4 or Thunderbolt connections1

1

.

Local AI and Gaming Without Subscription Fees or Cloud Dependencies

The wireless local-GPU solution targets users seeking to run large open-weight AI models entirely at home, supporting models up to 750B+ parameters including Kimi K3, DeepSeek v4.1, and Gemma 42

2

. This approach eliminates usage caps, subscription fees, and addresses data privacy concerns associated with cloud-based AI compute2

2

. The device operates in two modes: Driver Mode, which presents as a standard local GPU to the operating system, and API Mode, which allows existing AI applications to connect to local compute with zero code changes2

2

. Created by the minds behind Vincross Hexa robot and Sequoia-backed Vesta Sleep, the WiCi One enables thin clients and entry-level laptops to stream AAA titles at 70+ FPS or accelerate timeline rendering in professional applications like Premiere Pro and DaVinci Resolve2

2

.

TurboStream Technology Expands VRAM Pool Through NVMe Storage

The WiCi One includes NVMe storage—1TB on the RTX 5060 Ti model and 4TB on the RTX 5090 configuration—that serves dual purposes for AI workloads2

2

. Through TurboStream weight paging technology, the 4TB NVMe storage on the RTX 5090 model can function as an extended VRAM pool, enabling more complex AI model deployments that exceed the GPU's native 32GB memory capacity1

1

2

2

. The system adopts a Unified Hierarchical Memory approach to manage resources efficiently2

2

. WiCi showcased token-per-second performance metrics for various large language models in the trillion-parameter range, demonstrating the device's capability for demanding AI workloads2

2

.

Broader Device Compatibility Compared to Traditional eGPU Solutions

Unlike conventional eGPU enclosures requiring Thunderbolt, USB4, or OCuLink interfaces, the WiCi One works with virtually any device supporting Wi-Fi 7 connectivity, including laptops, desktops, tablets, and smartphones1

1

. This compatibility advantage eliminates the need for specific ports or adapter modifications that traditional eGPU solutions demand1

1

. The device utilizes 4x4 MIMO on the 320 MHz band for Wi-Fi 7 transmission, employing proprietary protocols and networking routines to address wireless connectivity challenges2

2

. While the specifications appear compelling, independent testing will be necessary to verify real-world performance claims and assess latency impacts inherent to wireless transmission versus wired connections2

2

. Watch for third-party benchmarks examining whether the wireless interface introduces noticeable input lag during gaming sessions or affects inference speeds for AI model execution compared to traditional PCIe-connected GPUs.