Depreciation Models Fail: Nvidia B200 Used Residual Value Reaches 158% of Launch Price, AI Compute Demand Continues to Drive GPU Premiums
Data analysis shows that Nvidia B200's used residual value has reached 158% of its launch price, while A100 and H100 residual values also far exceed traditional depreciation expectations, reflecting the strong boost AI compute demand gives to GPU premiums.
Key Takeaways
- Key Highlight：Data analysis shows that Nvidia B200's used residual value has reached 158% of its launch price, while A100 and H100 residual values also far exceed traditional depreciation expectations, reflecting the strong boost AI compute demand gives to GPU premiums.
- Innovation & Tech：Highlights advancements in Depreciation, Models, Fail, demonstrating rapid progress in model capabilities.
- Industry Impact：Reported via IT之家 (CN), offering actionable signals for developers and technology leaders.
Data analysis firm Silicon Data's latest report points out that Nvidia's Blackwell-based B200 GPU, approximately one year after shipment, has seen its used residual value soar to 158% of its launch MSRP. This means the AI accelerator card has seen roughly a 58% premium on the used market, breaking the conventional rule of year-over-year depreciation for traditional hardware equipment.
Not only the latest B200, but the earlier released A100 and H100 GPUs also currently have residual values far exceeding the levels expected under traditional 3-year or 5-year straight-line depreciation schedules. In traditional financial and accounting models, hardware equipment costs are typically amortized evenly over the useful life, but this rule has clearly failed in the current AI accelerator market. The core reason is that the demand for compute power from large model training and inference remains in a high prosperity cycle.
This phenomenon has a dual impact on compute supply and demand as well as asset management in the AI industry. On one hand, the high used residual values and premiums directly reflect the market's extreme thirst for high-performance AI chips, indirectly confirming that compute supply remains in a tight balance. On the other hand, the traditional method of accounting for AI compute assets using straight-line depreciation is facing challenges. In the future, when tech companies invest in large model R&D and data centers, they may need to re-evaluate the financial models and asset value retention rates of AI chips.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.
Industry Insights & Analysis
As artificial intelligence rapidly evolves, breakthroughs surrounding Depreciation, Models, Fail, Nvidia are shifting toward scalable, robust real-world implementations.
Driven by both open-source ecosystems and proprietary model architectures, the integration between compute optimization, data engineering, and agentic workflows is accelerating. This development provides a strategic benchmark for upcoming AI tooling and developer workflows.