Microsoft's LinkedIn plans to keep its investment in GPU, compute, and storage capacity flat for fiscal year 2027, according to a report from Wired. This move counters the broader industry trend of escalating spending on AI infrastructure amid a technological arms race.
The decision is reportedly driven by significant internal efficiency gains. LinkedIn is said to have doubled the efficiency of its existing GPUs within the past six months, enabling it to support growing AI demands without increasing its hardware and compute footprint.