The partnership integrates Rapt AI’s workload automation capabilities with AMD’s Instinct MI300X, MI325X, and the upcoming MI350 silicon. The big idea is to strip the faff out of managing GPUs while squeezing every last watt of inferencing and training power out of them.

Data Center GPU Business Negin Oliver said: “At AMD, we are committed to delivering high-performance, scalable AI solutions that empower organisations to unlock the full potential of their AI workloads.”

“Our collaboration with Rapt AI combines the cutting-edge capabilities of AMD Instinct GPUs with Rapt’s intelligent workload automation, enabling customers to achieve greater efficiency, flexibility, and cost savings across their AI infrastructure.”

For anyone fed up with GPU bottlenecks, idle silicon, and infrastructure TCO spiralling out of control, the combo promises a bit of respite – if they can get it to go.

Instead of forcing data scientists to play hardware roulette, Rapt’s platform does the heavy lifting—automatically tuning GPU usage across multi-cloud and on-prem setups. The pitch here is that AMD’s kit can finally be run at full speed, and with Raptor’s optimisation, make AMD look like a serious contender in the AI arms race.

While Nvidia is still king of the AI mountain, AMD is quietly bulking up. The Instinct MI325X isn’t a joke, and the MI355X is on the horizon. This deal with Rapt AI throws another log on the fire, providing AMD with a more polished and cost-efficient way to offer scalability to AI shops that are stretched thin on silicon.

Rapt CEO Charlie Leeming said: “Collaboration with AMD allows us to further enhance our platform, optimising it for the powerful AMD Instinct GPUs.” This joint solution is poised to transform AI infrastructure management, delivering improved performance, cost efficiency, and a faster time to value for our mutual customers. We are excited about the impact this will have on accelerating AI innovation across industries.”