AI Data Centres Are Making Laptops Costlier for Bangladeshi Buyers

Walk into a computer shop in Dhaka or Chattogram today and the price tags on ordinary laptops tell a different story from a year ago. Machines that once sat comfortably in the mid-forty-thousand-taka range now start closer to the mid-sixties or higher. Mid-range models have climbed as well, and buyers looking for better processors or more storage are being asked to stretch their budgets further still. The shift is not the result of a sudden local tax change or a one-off import hiccup. It is the downstream effect of a global scramble for the same basic components that power artificial-intelligence systems.

Major technology firms are racing to expand data centres so they can train and run larger AI models. Those facilities require enormous quantities of memory chips, high-speed storage, and related silicon. Chip makers, facing strong demand and attractive margins from these orders, have redirected a larger share of their production lines toward the specialised parts used in servers. The result is a tighter supply of the conventional memory modules, solid-state drives, and processors that go into ordinary personal computers. When the volume of standard components shrinks while demand for finished devices remains steady, prices rise. Bangladesh, which imports nearly all of its computer hardware, feels the movement almost immediately.

Retailers describe a market in which costs have moved higher across the board. Entry-level laptops that previously sold for around Tk 48,000 now commonly list between Tk 65,000 and Tk 70,000. Configurations that used to fall in the Tk 65,000–70,000 band have shifted upward, and higher-specification machines have seen still larger percentage increases. Memory sticks, storage drives, and graphics cards have all become more expensive as well. Some sellers note that the present tightness feels more persistent than the disruptions of the pandemic years, when factories paused and then resumed. This time the constraint is not a temporary shutdown but a deliberate reallocation of limited factory capacity toward higher-value AI orders.

The practical effects land unevenly. Students who need a basic machine for online classes or assignments face a higher entry cost. Small businesses replacing aging office computers find their upgrade budgets stretched. Households that once planned to buy a second device for shared use are postponing the purchase. Some buyers settle for lower specifications than they intended; others wait, hoping for a correction that industry observers say is unlikely in the near term. Because Bangladesh's currency and shipping costs add another layer on top of already elevated global prices, local buyers absorb the full force of the squeeze.

Global forecasts offer little immediate comfort. Analysts tracking memory-chip contracts expect conventional DRAM prices to continue rising through the final months of 2026. Production capacity is not being expanded overnight, and the demand from AI infrastructure projects shows no sign of easing. Chip makers have limited incentive to divert lines back to lower-margin consumer parts while large orders from data-centre operators remain on the table. Most estimates point to continued pressure into 2027 before any meaningful easing appears.

For Bangladesh the episode is a reminder of how tightly local technology markets are linked to distant infrastructure decisions. A factory decision made to serve AI training clusters in other regions quickly changes the cost of a laptop sold in a neighbourhood shop. Until the balance between AI-related demand and ordinary consumer supply improves, buyers will continue to face higher prices for the same computing capability. In the meantime, careful shopping, longer upgrade cycles, and greater attention to value for money remain the main practical responses available to ordinary users.