Google and Intel deepen AI infrastructure partnership

Editor's take

Google and Intel are expanding their collaboration to jointly design specialized AI hardware, aiming to address the escalating demand for processing power. This move signifies a strategic effort by two leading technology companies to secure their supply chains and accelerate AI development amidst a persistent global chip shortage that has impacted various industries.

The partnership is particularly significant as it involves co-designing custom silicon, a complex undertaking that could yield bespoke solutions tailored for Google's extensive AI workloads, such as those powering its Bard chatbot or internal research initiatives. By integrating Intel's manufacturing capabilities with Google's AI expertise, the companies aim to optimize performance and efficiency beyond off-the-shelf solutions, potentially impacting the competitive landscape of AI infrastructure providers like NVIDIA.

Future developments to monitor include the specific performance benchmarks of these co-designed chips compared to existing offerings from competitors. Crucially, the success of this venture will hinge on whether it can demonstrably reduce Google's reliance on external chip suppliers and offer a tangible advantage in AI model training and inference speeds, thereby influencing other hyperscalers' infrastructure strategies.