Quobyte is strengthening its capabilities to provide AI inferencing and agentic context to GPUs and building up its sales side. CEO and co-founder Bjorn Kolbek said: “A year ago it was all about training and refining or fine tuning models. Suddenly it has completely moved away [towards] inference, inference, inference.” He explained that multi-turn agentic AI involves an ongoing conversation where context, such as loaded documents, grows over time and must be swapped out to persistent storage and loaded into a GPU as fast as possible. This historical context can be substantial; “We're talking 100 gigabytes. 100 gigabytes or more.” Kolbek noted a bottleneck in GPU node architecture, where a “100 gig front-end for everything else in the storage” exists. Consequently, Quobyte is getting involved in the KV caching area, having rolled out GPU Converged into production. Kolbek said: “If my context is distributed across them, because it needs to be redundant, then, if I need to load it, we can use all the backend nodes. So let's say eight times 200 gigabits to pull your context from the other nodes. So the bottleneck is gone.” Quobyte initially envisioned GPU Converged as a tool to save cost and power, but Kolbeck said: “Now it turns out to be an accelerator for this context because we load over this massive back-end network.” Regarding software to store context data, Kolbeck said: “We're just a