Razer joins AI bandwagon with external AI accelerator backed by iconic AMD chip architect

• Razer introduces local AI hardware focused on developers and on premise work
• Tenstorrent accelerator adds portable AI compute via Thunderbolt connected devices
• AI device supports daisy chained units for local multi accelerator workloads

Razer has revealed an expansion beyond gaming hardware with an external AI accelerator and a new workstation platform aimed at developers working locally on advanced models.

Launched at CES 2026, the Razer Forge AI Dev Workstation is a high performance system designed for training, inference, and simulation workloads without relying on cloud services.

The on premises solution is for developers who want direct control over datasets, models, and experiments while avoiding subscription fees.

Tenstorrent external AI accelerator

The Razer Forge AI Dev Workstation supports up to four professional graphics cards from Nvidia or AMD, allowing large pooled VRAM configurations for multi GPU workloads.

Processor options include AMD Ryzen Threadripper PRO and Intel Xeon W chips, paired with support for eight DDR5 RDIMM slots for large memory capacity.

Networking is handled through dual 10Gb Ethernet ports, while storage includes up to four PCIe Gen5 M.2 NVMe drives and eight SATA bays.

Cooling is designed for sustained loads, with multiple high pressure fans intended to maintain airflow across dense internal components.

The workstation can operate as a standalone tower or transition into rack environments, allowing it to scale from individual desks to clustered deployments.

Alongside the workstation, Razer has been working with Tenstorrent on a compact external AI accelerator aimed at portable development workflows. Tenstorrent is led by Jim Keller, best known for his work on AMD’s Zen CPU architecture and early self driving silicon at Tesla.

The accelerator connects over Thunderbolt 4 or Thunderbolt 5 and is designed to add local AI compute to laptops and other compatible systems.

It is based on Tenstorrent’s Wormhole architecture and supports the company’s open source software stack for running LLMs, image generation models, and other AI workloads.

Multiple units can be connected together, with up to four devices forming a small local cluster for larger models.

“A device anyone can plug into their laptop unlocks the next generation of developers building on our open platform,” said Christine Blizzard, chief experience officer at Tenstorrent. “Our goal is to make AI more accessible and we trust Razer to deliver products that developers love.”

“AI developers on the edge demand power, flexibility, and mobility – and this collaboration delivers all three,” said Travis Furst, head of notebook and accessories division at Razer. “Our partnership with Tenstorrent combines their cutting-edge AI acceleration technology with Razer’s expertise in high-performance engineering and external enclosure design. Together, we’re advancing edge AI development as part of Razer’s broader vision for AI – bringing portable, uncompromising compute to developers.”

Pricing and availability for the external AI accelerator have yet to be announced.

TechRadar will be extensively covering this year's CES, and will bring you all of the big announcements as they happen. Head over to our CES 2026 news page for the latest stories and our hands-on verdicts on everything from wireless TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI. You can also ask us a question about the show in our CES 2026 live Q&A and we’ll do our best to answer it.

And don’t forget to follow us on TikTok and WhatsApp for the latest from the CES show floor!