Self-driving startup Turing adds AMD Ventures and moves to AMD AI GPUs

Turing Inc., a self-driving technology startup based in Japan, has added AMD Ventures to its list of backers and begun adopting Advanced Micro Devices (AMD) AI accelerators in its systems, according to a report carried by Yahoo Finance citing Bloomberg.

The change centers on the compute hardware used to train and run the machine-learning models that underpin perception, prediction, and planning in autonomous driving software. GPUs are specialized processors designed to run large batches of parallel computations efficiently, and AI accelerators are hardware optimized to speed up neural-network workloads.

While details were limited in the report, the adoption of AMD GPUs suggests Turing is aligning more of its AI pipeline with AMD’s data-center and AI-oriented silicon strategy rather than using only GPUs from other chip vendors. For startups, that kind of shift can affect development timelines, model performance, software integration work, and access to available compute resources.

The reports also framed the AMD Ventures backing as part of Turing’s broader push toward commercialization. Japan-based outlets characterized the company as about five years old and said the new support is intended to bolster its capabilities as it builds toward a commercial launch.

AMD Ventures is described in the reporting as joining Turing’s investor base, a announcement that larger chip suppliers and their venture arms are increasingly looking beyond chip sales alone and toward upstream participation in the companies writing the software that will later demand those chips at scale.

From AMD’s perspective, involvement with an autonomous-driving developer can help validate the practical readiness of its AI platform in real-world perception workloads, where latency and accuracy requirements are strict. For the broader self-driving sector, hardware selection is becoming a strategic decision because machine-learning systems can require sustained compute access, both during training and during ongoing refinement.

Still, the reporting did not disclose which specific AMD GPU models or accelerator families Turing is using, what proportion of its workloads have moved, or whether the company has signed any exclusive supply or integration agreements. It also did not provide performance results, deployment timelines, or customer validation details.

What to watch next is whether Turing provides more specifics on its hardware roadmap, including the types of accelerators and software stacks used, and whether AMD’s role evolves from venture backing into deeper collaboration around developer tooling, reference platforms, or other support for automotive AI workloads.