Lishuan Technology is indeed related to the AI chip concept, but it is not a pure AI chip company. Its core positioning is as a high-performance graphics GPU design enterprise, whose products maintain strong graphics processing capabilities while also possessing considerable AI computing power.
Below, I will analyze the relationship between Lishuan Technology and the AI chip concept from several key aspects:
🎯 Core Products and AI Capabilities
Lishuan Technology’s first GPU chip, the 7G100 series, is based on its self-developed TrueGPU TianTu architecture, with the instruction set and computing cores entirely designed in-house. Although this chip primarily targets the graphics rendering market, its design also integrates AI computing capabilities:
-
Supported Computing Precision: The 7G100 series supports multiple computing precisions including FP32 (single precision floating point), FP16 (half precision floating point), and INT8 (8-bit integer). This allows it to perform high-precision graphics rendering and scientific computing (FP32), as well as handle AI inference tasks that require higher efficiency (FP16, INT8).
-
AI Performance: According to tests, Lishuan Technology’s GPU chip has an AI computing power several times that of mainstream AI PC devices, efficiently supporting complex AI tasks such as code assistants and retrieval-augmented generation (RAG).
🔄 Strategic Positioning: A Gradual Path from Graphics to AI
Lishuan Technology’s main strategic positioning remains that of a GPU rendering chip company, focusing on the graphics card market in the short term. Although its technical architecture supports AI workloads, it has not yet launched dedicated AI chips (such as pure AI training cards or inference cards).
Choosing a path of starting from graphics GPUs and gradually enhancing AI capabilities is a pragmatic choice:
-
Graphics Market as a Starting Point: There is clear market demand and returns in areas such as gaming and professional visualization, which can help the company achieve commercialization.
-
AI as a Future Extension: Integrating AI units into graphics GPUs can provide value to existing markets (such as game super-resolution technology) while laying a technical foundation for future expansion into broader AI computing markets.
📊 Summary of Lishuan Technology’s AI Attributes
| Dimension | Description |
|---|---|
| Core Business | High-performance graphics GPU design |
| AI Computing Capability | Supports FP16, INT8, etc., with strong AI inference capabilities |
| Product Positioning | Graphics-focused, AI as a supplement; not a dedicated AI training chip |
| Technical Architecture | Self-developed TrueGPU architecture, capable of developing AI computing chips |
| Future Potential | Currently developing GPGPU and AI acceleration chips, with potential to enter the AI training/inference market |
💎 Conclusion
So, is Lishuan Technology related to AI chip concepts?
Yes, but it requires an accurate understanding of its connotation.
You can view it as a high-performance graphics GPU company with strong AI computing capabilities, serving as an important hardware supplier for empowering AI applications within the AI chip concept. Its current products are not specifically designed for AI training, but they have potential in AI inference, edge computing, and AI applications combined with graphics rendering (such as gaming, digital twins, and intelligent cockpits).
If investors are focused on the concept of “pure” AI training chips, they may pay more attention to companies like Cambricon and Haiguang Information. However, if they are interested in the opportunities of “GPU + AI” integrated computing and domestic high-performance graphics processors in broader AI application scenarios, then Lishuan Technology is undoubtedly an important market participant.
Investing in AI chip concept companies requires attention to risks such as rapid technological iteration, high ecological barriers, and potential performance fluctuations. It is recommended to closely monitor Lishuan Technology’s subsequent progress in product mass production, market promotion, and especially in the development of AI-specific architectures.