Khaberni - Google took a significant step in the global AI race by deciding to promote Amin Wahdat to the position of Senior Expert in AI Architecture, a position newly created for the first time within the company, elevating him directly to the level of CEO Sundar Pichai.
The decision, revealed by media reports and later confirmed by the company, reflects the magnitude of the bet Google is placing on developing its infrastructure, especially as it prepares to spend up to $93 billion on capital investments by the end of 2025, a figure that is likely to rise significantly next year.
Wahdat is not a new name within the company. The scientist who began his research career at Xerox in the early nineties and holds a PhD from the University of California, Berkeley, has spent 15 years at Google building what resembles the "backbone" that the company’s AI relies on, according to a report published by TechCrunch and seen by Al Arabiya Business.
Before joining Google in 2010, he held a position as a professor at Duke University, and then as a professor and chair of the SAIC at the University of California San Diego, with an extensive research record that includes about 395 scientific papers, most of which revolve around improving computing efficiency on massive scales.
Wahdat was highlighted in April when he took to the stage at the Google Cloud Next conference to unveil the company’s seventh generation of AI processors (TPU), known as Ironwood.
The numbers he presented were astounding: more than 9,000 chips inside each "pod," with a capacity of up to 42.5 exaflops of compute power, surpassing the world's most powerful supercomputer by more than 24 times at that time.
He said in his speech: "The demand for AI computing capabilities has increased by 100 million times in just eight years."
But what is more important is what happens behind the scenes. Wahdat led the development of TPU chips specialized for training and running AI models, which gives Google a competitive edge against players like "OpenAI".
He also oversees the "Jupiter" internal ultra-high-speed network, which transfers massive amounts of data between the company's servers, and currently has a capability of 13 petabits per second, theoretically enough to handle a video call for every inhabitant on Earth at once.
Wahdat is also one of the engineers of the Borg system, a data center management platform that coordinates work between thousands of servers as if it were "one brain," and has participated in supervising the development of Axion processors, the first "Arm" processors specialized for servers designed by "Google".
In short, Wahdat represents the cornerstone in Google's AI story.
Industry sources believe that promoting an engineer of this stature also sends a clear message about retaining talent, as when it takes 15 years to build such an impactful expert, you do not let him leave easily.
