# Enterprises Favor Non-Nvidia Chips in AI Accelerator Evaluations

> As enterprises evaluate their AI accelerator options, a notable 14-point advantage emerges for non-Nvidia chips, signaling a transformative moment in the AI infrastructure landscape.

**Source**: venturebeat.com | **Published**: 2026-09-02 | **Type**: article

## Key Facts

- 39.4% prefer non-Nvidia chips, signaling a shift in AI accelerator strategies.
- Azure's production adoption surged 18.1 points, highlighting its growing market strength.
- Urgency for platform changes dropped 9.5 points, indicating a cautious approach to infrastructure.
- Neocloud interest rose to 38%, suggesting a viable alternative to traditional hyperscale providers.
- Open-source production use jumped from 3.7% to 12.9%, reflecting a trend towards greater control.

## Summary

Recent findings from VentureBeat's July VB Pulse survey reveal a significant shift in enterprise preferences for AI accelerators, with a growing inclination toward non-Nvidia chips. The survey, which included 170 AI infrastructure respondents, indicated that 39.4% are likely to evaluate alternatives such as AWS Trainium, Google TPU, AMD Instinct, and Intel Gaudi over Nvidia's next-generation GPUs, which garnered only 25.3% interest. This 14-point gap marks a notable change in the competitive landscape, suggesting that enterprises are increasingly seeking options beyond Nvidia, which has long dominated the market.

Despite Nvidia's established position as the default choice in many production environments, organizations are building flexibility into their accelerator strategies. The survey results reflect a broader trend wherein enterprises are optimizing their existing AI infrastructure rather than hastily switching platforms. Notably, the percentage of respondents expecting to change platforms within three months dropped from 38.3% in June to 28.8% in July, even as production adoption and usage of various accelerators increased. This indicates a more deliberate approach to infrastructure planning, with companies focusing on enhancing their current capabilities.

In terms of specific platforms, Microsoft Azure experienced the most significant growth, with production adoption rising from 29% in June to 47.1% in July. This increase can be attributed to a respondent base that skewed towards larger organizations, where Azure adoption tends to be higher. Google’s Gemini also saw a rise in production use, from 41.1% to 47.6%, while OpenAI and Anthropic gained traction as well. The survey highlighted that enterprises are not only expanding their infrastructure but are also becoming more effective operators, as evidenced by a growing emphasis on uptime and reliability as key performance metrics.

The declining urgency for immediate platform changes reflects a strategic pivot among enterprises. The share of respondents planning to switch platforms within zero to three months decreased significantly, while those anticipating changes within three to six months increased. This shift suggests that enterprises are prioritizing integration with existing cloud and data stacks over immediate performance improvements, indicating a more cautious and calculated approach to technology adoption.

Interest in non-Nvidia alternatives is particularly pronounced among decision-makers and C-suite executives, with the likelihood of evaluating these options rising significantly. This trend underscores a strategic shift in how organizations view accelerator diversity, moving it from a technical concern to a broader strategic imperative. Smaller enterprises are also increasingly considering non-Nvidia options, indicating that this trend is not limited to larger firms.

The survey findings also point to a growing interest in neoclouds—specialized cloud providers focused on AI infrastructure. The share of respondents planning to engage more with neoclouds rose from 33% to 38%, particularly among technology and software sectors. This suggests that neoclouds are gaining traction as viable alternatives to traditional hyperscale providers, offering organizations additional leverage in negotiations and access to diverse accelerator options.

Open-source AI infrastructure usage is also on the rise, with the share of respondents utilizing self-managed open-source stacks increasing from 3.7% to 12.9%. This trend indicates that enterprises are seeking greater control and flexibility in their AI deployments, although it also brings the responsibility of managing upgrades and security.

The combination of increased infrastructure activity, diminished urgency for platform changes, and a growing appetite for diverse accelerator strategies suggests that enterprises are positioning themselves for a more competitive future. As organizations continue to refine their AI infrastructure and explore alternatives, the next wave of platform changes will likely be driven by strategic considerations rather than reactive decisions. This evolving landscape will challenge established players like Nvidia to innovate and adapt, as enterprises prioritize flexibility, reliability, and performance in their AI strategies.

## Entities

- **Companies**: Nvidia, AWS, Google, AMD, Intel, Microsoft, OpenAI, Anthropic, CoreWeave, Lambda, Crusoe, Nebius
- **Products**: AWS Trainium, Google TPU, AMD Instinct, Intel Gaudi, Nvidia Blackwell, Gemini, AI harness, LMCache, vLLM
- **Technologies**: PyTorch, Triton, Ray, Kubernetes

## Key Concepts

AI infrastructure, accelerators, Nvidia alternatives, neoclouds, open-source AI, context layer, production adoption, platform change urgency

## Definitions

- **AI infrastructure**: The underlying systems and components that support the development and deployment of AI applications.
- **neoclouds**: Specialized cloud providers focused on AI infrastructure, particularly access to accelerators and supporting services.
- **context layer**: The operational layer that connects AI models to enterprise data and tools, determining access and actions for agents.
- **open-source AI**: AI technologies and frameworks that are publicly available for use, modification, and distribution.
- **accelerators**: Hardware components designed to speed up the processing of AI workloads, such as GPUs and specialized chips.

## Use Cases

- Evaluating non-Nvidia accelerators for AI workloads
- Implementing open-source production stacks
- Using neoclouds for AI infrastructure
- Building custom AI harnesses
- Optimizing existing AI infrastructure
- Integrating multiple cloud and data stacks

## Frequently Asked Questions

**What are the main alternatives to Nvidia chips for AI acceleration?**

The main alternatives include AWS Trainium, Google TPU, AMD Instinct, Intel Gaudi, and in-house ASICs. Enterprises are increasingly considering these options to diversify their AI infrastructure.

**How is the urgency for platform changes among enterprises shifting?**

The urgency for immediate platform changes is declining, with more respondents planning changes in the three to six-month and six to twelve-month windows. This suggests a more deliberate approach to evaluating and adopting new technologies.

**What role do neoclouds play in enterprise AI strategies?**

Neoclouds are becoming a credible part of enterprise strategies by providing specialized AI infrastructure and access to accelerators. Their adoption is expected to grow as organizations seek more options for workload placement.

**What is the significance of the context layer in AI infrastructure?**

The context layer connects AI models to enterprise data and tools, ensuring that agents can access the right information and take appropriate actions. Its governance is crucial for operational effectiveness.

**How is open-source AI infrastructure being adopted by enterprises?**

Open-source AI infrastructure usage is growing, with an increase in organizations reporting custom, self-managed stacks. This trend reflects a desire for greater control and flexibility in AI deployments.

## Links

- [Read on Welcome.AI](https://welcome.ai/content/enterprises-favor-non-nvidia-chips-in-ai-accelerator-evaluations)
- [Original source](https://venturebeat.com/data/enterprises-put-non-nvidia-chips-14-points-ahead-of-nvidias-next-gen-gpus-on-their-evaluation-lists)
- [OpenAI](https://welcome.ai/company/openai): Featured company

---

Source: Welcome.AI | https://welcome.ai/content/enterprises-favor-non-nvidia-chips-in-ai-accelerator-evaluations