How will the rise of RISC-V-based AI chips affect Nvidia?
EVAS's RISC-V chip architecture challenges Nvidia's GPU dominance The emergence of RISC-V-based AI chips, such as EVAS Intelligence's Epoch series, presents a structural alternative to Nvidia's GPU model. EVAS secured nearly RMB 2 billion in financing to develop this chip, which utilizes a TPU-style accelerator approach that differs substantially from Nvidia's general-purpose shader cores. Because the RISC-V architecture is governed by a Swiss nonprofit, it is structurally outside the reach of US Commerce Department export controls, offering a viable path for companies seeking to bypass restrictions on Nvidia hardware.
- Effect
- Mild negative
- How direct
- 2 steps, all reported
- When
- Within weeks
- The story
- Gone quiet
How it reaches Nvidia
-
Beijing-based AI chip startup EVAS Intelligence closed a financing round of nearly RMB 2 billion on September 18, 2026, reaching a post-money valuation of nearly RMB 15 billion. This follows the RISC-V open-standard governance body relocating to Switzerland in 2019 to avoid US trade regulations. EVAS's core product, the Epoch series, is built on RISC-V and departs substantially from Nvidia's GPU model.
The full event1independent outlet -
The RISC-V International organization relocated from the United States to Switzerland in 2019, citing concerns over US trade regulations. This structural move ensures the open ISA cannot be subjected to US Commerce Department export controls like proprietary US chips.
1 report connects these two. Brind only summarizes; follow a link to read the reporting itself.
- techtimes.com Sep 20
-
country in Central Europe
Everything about Switzerland -
The structural independence provided by the Swiss governance body enables Beijing-based EVAS Intelligence to build its Epoch series on RISC-V. This chip uses a TPU-style accelerator that departs substantially from Nvidia's GPU model, offering a non-restricted alternative to Nvidia's dominant architecture.
1 report connects these two. Brind only summarizes; follow a link to read the reporting itself.
- techtimes.com Sep 20
-
American multinational technology company
Everything about Nvidia
Tap any step to see the evidence behind it.
The facts so far
As reported. Each one links to where it comes from.
- EVAS Intelligence closed a financing round of nearly RMB 2 billion (approximately $295 million USD) on September 18, 2026.techtimes.com
- RISC-V International relocated from the United States to Switzerland in 2019 to avoid exposure to US trade regulations.techtimes.com
- EVAS's Epoch chip uses a TPU-style approach, unlike Nvidia's GPU model which uses thousands of general-purpose shader cores.techtimes.com
- The Epoch chip natively supports block-quantized FP8 precision, a format that roughly halves memory bandwidth requirements compared to FP16.techtimes.com
Why it matters
Nvidia currently dominates the AI accelerator market, largely due to its proprietary GPU architecture and the massive ecosystem built around CUDA. This ecosystem lock-in makes migrating away from Nvidia extremely costly for existing customers. The rise of RISC-V-based competitors like EVAS directly challenges this dominance by offering an open-standard alternative that is architecturally distinct and structurally outside the reach of US export controls.
This development is highly significant in the context of the US-China tech rivalry. While Huawei's proprietary Ascend chips face ongoing scrutiny regarding US leverage points, EVAS's use of RISC-V provides an architecture-level exposure gap. The success of EVAS's capital formation, which included Zhongxin Juyuan (a fund linked to SMIC), reflects a deliberate, state-aligned push to build a domestic, non-restricted AI infrastructure.
What we don't know yet
- Will independent third-party benchmarks confirm EVAS's claim of performance parity with leading vendors?
- How mature is the RISC-V software ecosystem for demanding production workloads outside of China?
What would change this answer
— Other considerations
Beyond this explainer
Questions worth asking
Ideas the web ties closely to this story that the explainer doesn't mention. These are prompts for your own research: Brind hasn't answered them.
-
About Intel Arc
How might Intel's Arc GPUs compete with both Nvidia's proprietary architecture and the open-standard approach of RISC-V chips?
-
About Jetson AGX
Given that EVAS's Epoch chip natively supports block-quantized FP8 precision, how does this compare to the capabilities offered by Nvidia's Jetson AGX platform?
-
About Samsung/SK Hynix
Since the Epoch chip reduces memory bandwidth requirements, how might the availability and performance of high-speed memory from companies like Samsung or SK Hynix affect its real-world deployment?
Explore connected ideas
Each opens in Related concepts: what the web connects to it, with Brind's news on it.
The ideas come from how millions of web pages place them near each other. The questions are Brind's AI reasoning, written only from this explainer and Brind's own reporting.
Reporting
- techtimes.comSep 20
Keep going
Brind's analysis is written by AI from the reporting linked above and can be wrong. It is not investment advice; do your own research.