🎉 TensorWave turns 2! 🎉 Two years ago, we set out to prove that an AMD-only cloud could power serious AI at scale. Today, we’re supporting some of the most demanding AI teams with high-performance GPU clusters, open ecosystems, and a focus on real-world results... not hype.

We’re building the next generation of AI infrastructure on AMD GPUs and we’re growing the team. If you’re excited about high-performance compute, open ecosystems, and helping customers ship real AI products, we’d love to hear from you.

Jensen spent 2 min on InferenceMAX at the GTC keynote 😎 InferenceMAX is completely open source and completely impartial, with the support of AMD, OpenAI, Nvidia, Microsoft, Oracle, Pytorch, SGLang, vLLM, Cursoe, Nebius, Tensorwave, etc

Two real-world wins on @AMD with TensorWave: • Modular: ~2× more inference throughput and up to 60% lower costs on MI355s • Telecom giant: 9B+ tokens/day and a 4B Gemma-based model beating larger LLMs on telecom + SQL benchmarks All powered by Supermicro hardware, AMD

Still buzzing from the happy hour we co-hosted with SemiAnalysis. Their team helped bring together an incredible group to talk GPUs, AI infrastructure, and what’s coming next. Really appreciate everyone who showed up and stayed late to keep the conversation going. Catch you at

Funny to watch SemiAnalysis complain about Tensorwave when their own InferenceMAX is totally unreliable. Glass houses and all. Through my own grapevine, I've heard a number of people complain about Tensorwave stability as well. Uptime and reliability is hard. Where you choose to

The biggest token consumers/buyers at OpenAI, from today’s presentation: 1T names (green) • Isaac Andersen — (Duolingo, eng.) • Alex Atallah — co-founder, OpenSea (and OpenRouter) • Chris Colon — (unknown/unverified) • John Emmons — co-founder & president, TensorWave

Building the future of AI is serious work... but it’s even better with this crew. Our holiday event was the perfect way to wrap up the year. Sharing a few of our favorite moments! Onward to an even bigger year ahead.

TensorWave Raises $100M to Build a Liquid-Cooled $AMD GPU Deployment The GPU cluster will consist of 8,192 MI325X GPUs

Grateful for days like yesterday; fast cars, good conversations, and great people. Thank you Reza Zabihi from Trace3 for hosting us at Speed Vegas and to Kabeer Gharzai from Poolside for making it even more fun. Let’s do it again soon.

i'm hiring engineers OBSESSED with enabling AI infra at scale at @tensorwave roles include: - VP Eng - VP Datacenter - Product - SRE - Full Stack Devs - and more.. we are focused and dedicated to a single vision - to empower the next wave of AI 🌊 we are the architects,

The largest $AMD GPU training cluster in North America Built by TensorWave Eager to see MI355X SuperClusters

Running AI workloads or need HPC power? Check out TensorWave Bare Metal on AMD Instinct™ GPUs for dedicated, high-performance infrastructure.

Confirmed and likely buyers of $AMD MI355 and MI450 → Crusoe → Oracle with Zettascale clusters → G42 for their 5GW campus → Tensorwave with its 1GW capacity target → Hot Aisle, best in town → $META, $MSFT, xAI, OpenAI: Will run tests, with final decisions depending on

After building out their cluster of 8,192 $AMD MI325 GPUs, TensorWave will be one of the first companies to deploy the MI355X. Will this trend continue with the MI400X and MI450X?The answer is we don't know. As investors it's our job to discern and try to predict the future.

With so many ppl at the conference and a busy agenda, NeurIPS parties are the best way to meet and connect! From a distillery with @tensorwave to playing arcade games and killing terminators with @vllm_project @AMD @RedHat_AI With @_Alex_Borghi_

NEW: TensorWave raises a $43M SAFE for its cloud GPU platform The platform is powered by AMD hardware only which will help break the supply chain constraints to meet the unmet demand. > 9 slide pitch that @TensorWaveCloud used to raise

Shape the Conversation at Beyond CUDA 2026! Get your work in front of 500+ AI builders, researchers, and decision-makers at Beyond CUDA 2026. Apply to speak and put your ideas at the center of the conversation.

From the TensorWave family to yours...