AMD is reportedly preparing to increase the supply prices for its graphics card and VRAM bundles to partners, with a hike of at least 10% expected to be implemented uniformly across brands starting next month. This move follows a similar, and reportedly more substantial, price increase by Nvidia, which industry analysts claim has gone as high as an extra 30%.
According to industry analyst Harukaze5719, citing information confirmed by AMD's internal supply chain, the company had finalized plans for this price adjustment. However, the implementation was initially deferred due to a weak consumer graphics card market, end-user price sensitivity, the need to clear existing inventory, and Nvidia's pricing remaining unadjusted at the time.
With Nvidia's subsequent price increase, AMD appears poised to proceed with its own adjustment. The pricing strategy is expected to place AMD series graphics cards in a similar position to their Nvidia counterparts. While a price increase to partners does not automatically translate to higher consumer prices, it is unlikely that companies will absorb the additional costs themselves, especially given the reportedly slim margins for AIB partners.
This potential price increase could also influence how partners manage their current inventory. As observed with Nvidia's pricing changes, some partners have expressed reluctance to sell at existing price points. The current market conditions, characterized by a memory crisis driven by the high demand for AI training, are contributing to component scarcity and price inflation.
Samsung's recent earnings call indicated that supply constraints are expected to intensify through 2027 and 2028, suggesting the memory crisis is unlikely to abate soon. This situation makes it a challenging time for PC builders, with the prospect of further price increases and an uncertain recovery timeline.
Graphics card prices have seen relative stability in recent months, according to market trackers. However, the escalating demand for components driven by the artificial intelligence sector is now exacerbating supply issues for both GPUs and VRAM, components essential for AI model training. This escalating demand is directly impacting the availability and cost of these critical PC hardware parts.