24x faster than supercomputers: Google unleashes new AI chip to chop competition – Ironwood
Ironwood’s 42.5 exaflops of computing power make it the foundation for Google’s advanced AI models, including Gemini 2.5..
AI is the latest hot technology taking over the world with a storm. That being said, it’s natural to see the largest of tech giants tinkering around with this technology and uncovering new inventions as the AI era takes shape.
The latest amongst these tech giants is Google, who unveiled its most efficient Tensor Processing Unit(TPU) called Ironwood, designed specifically for AI models.
This TPU is built to help AI models work faster and smarter, especially when it’s working on tasks that require reasoning or making predictions. This particular activity is known as ‘inferencing.’ Ironwood is specifically built for such tasks, unlike its predecessors that were designed for training AI models from scratch.
Discussing the capabilities
Announced at Google Cloud Next ’25, this seventh-generation TPU delivers 24 times the computing power than any other supercomputer in the world when deployed at scale.
Amin Vahdat, Google’s Vice President and General Manager of ML, Systems, and Cloud AI, in a virtual press conference ahead of the event, said:
Ironwood is built to support this next phase of generative AI and its tremendous computational and communication requirements,
He said,
This is what we call the ‘age of inference’ where AI agents will proactively retrieve and generate data to collaboratively deliver insights and answers, not just data
Ironwood comes with a set of impressive tech specifications. It can deliver 42.5 exaflops of computing power when scaled to 9,216 chips per pod. These massive statistics dwarf the El Capitan’s 1.7 exaflops – the world’s fastest supercomputer.
Ironwood also comes with 192 GB of High Bandwidth Memory (HBM). It’s six times more than Trillium, Google’s previous generation TPU, announced last year. Ironwood can also reach 7.2 terabits per second per chip in terms of memory bandwidth, a 4.5x improvement over Trillium.
Vahdat explained:
At a time when available power is one of the constraints for delivering AI capabilities, we deliver significantly more capacity per watt for customer workloads,
Gemini’s brain?
Google has also labeled Ironwood as the foundation for all of its advanced AI models, including Gemini 2.5.
At the conference, Google introduced Gemini 2.5 Flash—a more affordable version of its flagship model that adapts its reasoning depth based on the complexity of the prompt. While Gemini 2.5 Pro targets advanced tasks like drug discovery and financial modeling, Flash is optimized for fast, everyday use where speed matters most.
The company also showcased its complete lineup of generative media models—spanning text-to-image, text-to-video, and the newly introduced text-to-music tool, Lyria. In a live demo, the models were used together to generate a full promotional video for a concert.
The larger vision
While Ironwood is rightfully the talk of the town at the moment, it still represents just one part of Google’s larger AI infrastructure strategy. The company also announced Cloud WAN – a managed wide area network that can give businesses direct access to Google’s network infrastructure.
Vahdat said:
Cloud WAN is a fully managed, viable and secure enterprise networking backbone that provides up to 40% improved network performance, while also reducing total cost of ownership by that same 40%,
This demonstration highlights the growing power of multimodal AI and its potential to transform creative workflows across industries.
READ the latest news shaping the AI Chips market at AI Chips News
24x faster than supercomputers: Google unleashes new AI chip to chop competition – Ironwood, source
Add comment