Memory Price Surge and Its Impact on Indian Enterprise IT
Comprint Tech Solutions’ Director Kunal Sancheti notes that DRAM contract prices jumped 90‑95 % in the first quarter of 2026, marking the steepest quarterly increase ever recorded by TrendForce, and rose an additional 58‑63 % in the second quarter. NAND flash prices are now accelerating even faster. Dell’s chief operating officer highlighted that DRAM, which cost roughly US$0.43 per gigabit in mid‑2025, has crossed US$2.39 per gigabit, pushing memory to represent about one‑third of the bill of materials for a new server, up from one‑fifth a year earlier.
Every OEM has re‑priced their offerings, some twice, as high‑bandwidth memory for AI accelerators consumes roughly three wafers of capacity for each wafer of conventional DDR5. The hyperscalers building AI data‑centres are placing open‑ended orders at prevailing market prices, effectively driving the shortage. SK Hynix announced that its 2026 output was essentially sold out before the year began, and new fabrication capacity is not expected in volume until late 2027, meaning the tight market will persist for at least the next eighteen months.
The cloud, built on the same memory, is also feeling the pressure; while it remains valuable for elastic, spiky workloads, the rising memory cost makes the elasticity premium less justifiable for steady, predictable enterprise workloads. Sancheti proposes a “hybrid ownership” model where each workload is evaluated for the most cost‑effective option—buy new, rent, refurbish, or extend existing assets. A five‑year‑old server with memory purchased at 2021 rates can be kept alive for two more years, effectively arbitraging against the current shortage, whereas a certified refurbished machine carries yesterday’s component costs into today’s budget. Renting converts a capital shock into a predictable monthly expense and transfers residual‑value risk to the lessor.
Total cost of ownership—encompassing purchase price, maintenance, downtime risk, financing, upgrade cycles and eventual retirement—has become the primary decision metric. Enterprises are now interrogating IT budgets line‑by‑line rather than approving a single lump sum.
Regulatory developments add another layer of complexity. India’s data‑protection rules were notified in November 2025, with substantive obligations taking effect in May 2027. While the law does not impose blanket localisation, it will require significant fiduciaries to keep certain data categories onshore, and the RBI already mandates localisation for payment data. Companies are therefore designing 2027 architectures amid 2026 uncertainty, favouring control over data location.
Sancheti forecasts three lasting shifts: (1) workload‑level economics will become a standard discipline; (2) refresh‑by‑default will give way to lifecycle extension, certified refurbishment and structured asset retirement as normal boardroom answers; and (3) compliance will move from an after‑thought to a core architectural consideration.
For India’s mid‑market—thousands of firms that keep the everyday economy running—these dynamics translate into survival math: deciding which workloads justify cloud spend at today’s rates, which servers merit replacement, and which can reliably operate for two more years with proper support.
About the author: Kunal Sancheti is a Director at Comprint Tech Solutions, a Mumbai‑headquartered enterprise‑IT firm that has managed procurement, rentals, maintenance, refurbishment and certified recycling for over three decades. Views expressed are personal; the press release is provided under an arrangement with NRDPL, and PTI assumes no editorial responsibility.
---