Skip to content

Chips ready now, and a path to one of your own.

The first-generation ASICs cover lightweight language models, heavy-load inference and multimodal vision. The same cores and routes can build a chip around your model.

First-generation chips side by side
Fig. 1 · First-generation ASICs
Forge lightweight edge LLM inference chip with the Fanshi mark laser-etched on the package

Forge

Edge LLM

Lightweight edge chip

A complete language model inside a watt-class handheld device.

17,000 token/s

Fig. 2 · Lightweight edge chip

Trident flagship LLM inference chip with the Fanshi mark on its metal lid

Trident

Heavy-load LLM

Flagship heavy-load chip

Run a 35B model locally at 15,000 token/s on a 70 W system budget.

15,000 token/s

Fig. 3 · Flagship heavy-load chip

Argus multimodal vision inference chip with the Fanshi mark, next to a camera module

Argus

Vision

Multimodal vision chip

Pixel-level, real-time vision inference at 120 FPS.

120 FPS

Fig. 4 · Multimodal vision chip

Die surface with memory arrays and logic regions (illustrative)

Custom ASIC

Your model

Built for your model

Your model, your chip. The same cores and routes, redefined around your model, power budget and volume.

0.6B → 35B+ model range

Fig. 5 · Built for your model

Figures on this page are current product definitions and design targets. Final specifications, test conditions, availability and delivery versions are confirmed in project documents agreed by both parties.

Qwen3.5 → 3.6 → 3.8

Newer Qwen models, no new architecture.

Qwen3.6 and Qwen3.8 keep the Qwen3.5 hybrid layout, which Fanshi inference cores already map. A newer model in the same family is a new chip build, not a redesign.

From an email to a chip

  1. 01

    Pick the route

    Target model, performance and power goals, update cadence, interfaces. The first question is always: how often does the model change?

  2. 02

    Evaluate

    Run the model on the evaluation board, measure accuracy and throughput, and get a working engineering sample before any silicon is committed.

  3. 03

    Design in

    Interfaces, system software and field validation built around the evaluation results, turning the sample into a product prototype.

  4. 04

    Deliver the ASIC

    Once the model is frozen: a production chip, or the IP integrated into your own silicon.

Request an evaluation kit

Send the target model, performance goal, power budget and interfaces. We confirm the route and reply with an evaluation plan.

Request an evaluation kit