"During an internal meeting, Amin Vahdat, responsible for AI infrastructure at Google Cloud, emphasized that the company must double its computing and delivery capacity every six months. According to him, Google is facing a technological and financial race in which infrastructure is the determining factor. In his presentation, Vahdat stated that the company must achieve a thousandfold increase in capacity over the next four to five years to remain competitive."
"The AI market is developing at an unprecedented pace, and all major hyperscalers are investing heavily. Alphabet has again raised its investment forecast this year to $91 to $93 billion, with another significant increase in 2026. Microsoft, Amazon, and Meta are doing the same. Nevertheless, Vahdat emphasized that for Google, it is not simply a matter of spending more than the competition. According to him, the goal is to build infrastructure that is more reliable and scalable than what is available elsewhere."
Google must double computing and delivery capacity every six months and achieve a thousandfold increase over the next four to five years to remain competitive. All major hyperscalers are investing heavily, with Alphabet raising its investment forecast to $91–$93 billion and further increases planned for 2026 while Microsoft, Amazon, and Meta expand spending. Google prioritizes building more reliable, scalable infrastructure rather than merely outspending rivals, emphasizing more efficient AI models and use of in-house chips like the Ironwood TPU for energy gains. DeepMind is positioned as a strategic asset. Massive expansions in compute, storage, and network capacity are required without increasing costs or energy, and 2026 is expected to be a challenging year due to cloud infrastructure pressure and concerns about an AI bubble.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]