Alex Medick

Discover how AMD GPUs and open-source AI software are competing with CUDA. Insights from Anush Elangovan on Beyond CUDA Podcast.

The AMD Advantage: Anush Elangovan on Building AI Software to Compete with CUDA

What does it take to challenge CUDA’s dominance in AI? According to Anush Elangovan, Corporate VP of AI Software at AMD, the answer lies in relentless innovation, an open-source mindset, and building software that just works. On this episode of Beyond CUDA, host Jeff Tatarchuk sits down with Anush to unpack his journey - from early hacking days in India to founding Nod.AI, leading it through a decade-long grind, and ultimately joining AMD to spearhead its AI software strategy.

This conversation

What does it take to challenge CUDA’s dominance in AI? According to Anush Elangovan, Corporate VP of AI Software at AMD, the answer lies in relentless innovation, an open-source mindset, and building software that just works. On this episode of Beyond CUDA, host Jeff Tatarchuk sits down with Anush to unpack his journey - from early hacking days in India to founding Nod.AI, leading it through a decade-long grind, and ultimately joining AMD to spearhead its AI software strategy.

This conversation dives deep into the role of open ecosystems, how Nod.AI helped AMD achieve critical breakthroughs, and what the future looks like for developers building beyond CUDA.









Key Takeaways

1. Open Source as a Foundation, Not a Risk

From the start, Nod.AI was fully open source, a decision many questioned. Anush’s philosophy: speed is the moat. If competitors can copy your code and catch up, you’re not moving fast enough. That mindset has carried into AMD’s AI software strategy.

2. The Long Grind of Innovation

Nod.AI’s journey wasn’t a single “pivot” but a decade of daily adjustments, constantly refining direction while building toward hardware-agnostic AI tools. This persistence positioned the company for acquisition by AMD after proving real-world wins against CUDA.

3. Building Software That Just Works

Anush’s vision for ROCm is simple: developers shouldn’t think about it. Just like no one worries what version of Chrome they’re running, AMD’s goal is to make ROCm a frictionless, trusted foundation for AI development.

4. Developers Are the North Star

While hyperscalers drive GPU demand, Anush is clear that developers matter most. AMD’s success depends on making it seamless for anyone - from students to startups to enterprises - to build directly on its platform without barriers.

5. The AMD Roadmap is Clear and Bold

With Instinct MI325 and MI355 GPUs powering large-scale clusters, AMD is proving its hardware advantage. But the real differentiation comes from pairing powerful silicon with a robust open-source software stack designed for long-term trust.


Wrapping Up

The future of AI won’t be defined by a single ecosystem. As Anush says, Beyond CUDA is about unlocking innovation everywhere: on open platforms, with open research, for all developers.

Ready to see how AMD Instinct GPUs can power your workloads at scale? Get started with TensorWave today and experience the AMD advantage for yourself.


About TensorWave

TensorWave is the AI cloud purpose-built for performance. Powered exclusively by Instinct™ Series GPUs, we deliver high-bandwidth, memory-optimized infrastructure that scales with your most demanding models—training or inference.

Ready to get started? Connect with a Sales Engineer.

AMD Instinct™ MI355X GPUs Now Available

The AMD Advantage: Anush Elangovan on Building AI Software to Compete with CUDA

Subscribe to our Blog

Stay ahead of the curve with the latest in AI, AMD accelerators, and all things TensorWave.

Product

Resources

Company

© 2025 TensorWave Inc. - All rights reserved.