The memory crisis continues to worsen, with no immediate signs of easing. Growing demand for memory from AI data centres has driven up the cost of memory chips, and Zoho founder Sridhar Vembu has now acknowledged that the price surge is making business increasingly difficult even for his company.
Vembu recently shared a post on X citing a report that said memory prices had risen by 500% over the past 12 months. He said the combination of higher memory costs and increasing AI token prices was creating significant challenges. “Memory prices, along with AI token prices, have made business very difficult. We have held back from raising prices but it is becoming hard,” he wrote.
According to a Tom’s Hardware report, memory prices have indeed climbed sharply. Average prices for certain DDR5 memory kits increased from around $90, or approximately Rs 8,600, in August 2025 to $425, or roughly Rs 40,000, by August 2026.
The increase has been even more dramatic for high-capacity kits. A 128GB DDR5-6400 kit is now selling for around $3,399, equivalent to roughly Rs 3.25 lakh. That is nearly 10 times its lowest recorded price of $329, or about Rs 31,490.
Vembu’s comments come as rising component costs continue to put pressure on the broader technology industry. Smartphone manufacturers such as Oppo, OnePlus, Vivo and Nothing have increased prices in recent months, while Apple is also rumoured to be considering an iPhone price hike. The impact of rising costs therefore appears to be extending beyond hardware manufacturers to software companies such as Zoho.
SRIDHAR VEMBU SAYS PROGRAMMING LANGUAGES MUST EVOLVE
Vembu believes the memory shortage could force a fundamental rethink of how modern programming languages are designed.
“For a long time, programming languages were designed with the assumption that memory is ‘free.’ That era has now ended,” he said.
According to the Zoho founder, developers now need to focus on creating more memory-efficient programming languages and smarter compilers, including those used for AI applications.
He said programmers should work towards reducing memory consumption and argued that code safety and developer productivity should not come at the cost of excessive memory usage. Vembu added that improving memory efficiency in programming languages would be an area of research for him going forward.
Meanwhile, AI companies continue to expand their global data-centre infrastructure to keep pace with growing demand.
OpenAI recently signed an agreement with SoftBank’s SB Energy for a 10-gigawatt data centre in Ohio, one of the largest such facilities announced so far. The project is partly supported by a commitment from Nvidia.
Indian technology companies are also increasing their investments in data-centre infrastructure. HCLTech has announced plans to invest Rs 3,500 crore in AI-focused data centres, while TCS is reportedly planning to spend up to $7 billion on a 1-gigawatt data-centre business.
