Elon Musk Praises AMD AI Hardware for Small-to-Medium Models, Notes NVIDIA Still Dominates Large-Scale Training

Elon Musk has publicly acknowledged the growing strength of AMD’s AI accelerators, stating that AMD hardware performs “pretty well” for small-to-medium AI models. However, Musk was quick to note that NVIDIA still holds a commanding lead when it comes to large-scale model training. His comments were made in a recent post on X.

Musk’s endorsement highlights the progress AMD has made in AI compute, especially with its Instinct MI300 and MI300X accelerators, which have already been adopted by his AI company xAI. Tesla has also previously partnered with AMD for its hardware needs, showcasing Musk’s willingness to diversify beyond NVIDIA in certain workloads.

While NVIDIA’s CUDA ecosystem remains the industry standard for training massive foundation models, AMD’s GPUs are increasingly recognized as viable alternatives for:

  • Inference workloads

  • Fine-tuning tasks

  • Small-to-medium foundation models

This distinction is important as the AI industry shifts toward efficiency and cost-effectiveness in addition to raw training performance.

Despite the recognition, AMD’s market penetration remains limited compared to NVIDIA’s near-monopoly among Big Tech firms like Microsoft, Meta, and Google. NVIDIA’s early AI momentum created a powerful lock-in through CUDA and proprietary software frameworks, making it harder for rivals to break through.

AMD, however, has been making strides with its ROCm software stack, improving compatibility with popular machine learning frameworks to reduce adoption friction. Combined with more competitive hardware, this positions AMD to compete more aggressively moving forward.

Musk’s statement serves as a valuable endorsement for AMD at a time when hyperscalers and enterprises are actively seeking alternatives to NVIDIA’s costly and supply-constrained GPUs. Even if NVIDIA remains the leader for cutting-edge training, AMD’s growing relevance in mid-tier AI models and inference workloads could help them carve out a larger slice of the market.


Do you think AMD can turn Musk’s endorsement into broader Big Tech adoption, or will NVIDIA’s CUDA ecosystem keep dominating AI for the foreseeable future?

Share
Angel Morales

Founder and lead writer at Duck-IT Tech News, and dedicated to delivering the latest news, reviews, and insights in the world of technology, gaming, and AI. With experience in the tech and business sectors, combining a deep passion for technology with a talent for clear and engaging writing

Previous
Previous

Squadron 42 Content Director Admits 2026 Launch Isn’t Certain, Fans Worry About Possible Delay

Next
Next

AMD Aims for “Milan Moment” With Instinct MI450, Strengthening Software Stack to Compete With NVIDIA