Nvidia partners with Run: ai and Weights & Biases for MLops Stack
We’re excited to bring Transform 2022 live back on July 19 and virtually from July 20 – 28. Join AI and data leaders for insightful conversations and Interesting networking opportunities. Sign up today!
Running the entire machine learning process lifecycle can often be a complex operation, involving many disconnected components.
Users need hardware optimized for machine learning, the ability to orchestrate workloads on that hardware, and then also some form of machine learning operations technology (MLops) to manage models. In an effort to make it easier for data scientists, artificial intelligence (AI) computing orchestration provider Run: ai, raised $75 million in March, as well as platform provider MLops Weights & Bias (W & B)), is partnering with Nvidia.
“With this tripartite partnership, data scientists can use Weights & Biases to plan and execute their models,” said Omri Geller, CEO and co-founder of Weights & Biases. Run: AI told VentureBeat. “On top of that, Run: who organizes all the workload efficiently on GPU Resources from Nvidia, so you get the complete solution from hardware to data scientist. “
Run: ai is designed to help organizations use Nvidia hardware for machine learning workloads in a cloud-native environment – a deployment method that uses containers and microservices managed by the platform Kubernetes container orchestration platform.
Among the most popular ways for organizations to run machine learning on Kubernetes are Kubeflow Open Source Project. Run:who is integrated with Kubeflow can help users optimize Nvidia GPU usage for machine learning, explains Geller.
Omri added that Run: who has been designed as a plug-in for Kubernetes that allows virtualization of Nvidia GPU resources. By GPU virtualization, resources can be split up so that multiple containers can access the same GPU. Run: anyone enables virtual GPU instance quota management to help ensure that workloads always have access to needed resources.
Geller says that the goal of the partnership is to make the entire machine learning workflow more consumable for business users. To that end, Run: ai and Weights & Biases are integrating to make it easier to run the two technologies together. Before partnering, Omri said, organizations that want to use Run:ai and Weights & Biases must go through a manual process for the two technologies to work together.
Seann Gardiner, vice president of business development for Weights & Biases, commented that the partnership allows users to take advantage of the training automation provided by Weights & Biases with GPU resources provided by Run:ai coordinator.
Nvidia is not faithful monogamous and cooperates with everyone
Nvidia is partnering with both Run: ai and Weights & Biases, as part of the company’s larger partnership strategy of collaborating within the machine learning ecosystem of vendors and technologies.
“Our strategy is to cooperate fairly and equally with the overarching goal of ensuring that WHO becoming popular,” Scott McClellan, senior director of product management at Nvidia, told VentureBeat.
McClellan says the partnership with Run: ai and Weights & Biases is particularly interesting because, in his view, the two vendors offer complementary technologies. Both providers can now also connect to Nvidia AI Enterprise Platformprovides software and tools that help make AI usable for businesses.
With the three vendors working together, McClellan said that if a data scientist is trying to use Nvidia’s AI enterprise containers, they don’t have to find a way to implement orchestrated implementation frameworks by Nvidia. their own or make their own schedule.
“These two partners complete our stack – or we complete theirs and we complete each other – so the whole is greater than the sum of the parts,” he said.
Avoid MLops’ “Bermuda Triangle”
For Nvidia, partnering with vendors like Run:ai and Weights & Biases are both intended to help solve a key challenge many businesses face when embarking on an AI project for the first time.
“The moment when a data science or AI project tries to go from test to production, is sometimes like the Bermuda Triangle, where a lot of projects die,” McClellan said. “I mean, they just disappeared in the Bermuda Triangle – how do I get this into production?”
Using Kubernetes and the cloud-native technologies commonly used by businesses today, McClellan hopes that developing and operating machine learning workflows has become easier than before.
“MLops is for ML — literally, how do these things not die in transition to production and continue to live a full and healthy life,” McClellan said.