Nvidia’s Quiet Price Hike Lands First on RTX 5090 Partners
Nvidia has reportedly increased the price it charges board partners for the RTX 5090 and the China-focused RTX 5090D V2, without changing official MSRPs. According to industry channel reports, the company informed partners on May 13 that higher GDDR7 costs necessitated an adjustment of approximately USD 300 (approx. RM1,380) per card. Because the RTX 5090 carries 32 GB of GDDR7—twice the memory of the RTX 5080—it is uniquely exposed to VRAM cost inflation, making it the first Blackwell GPU to feel the squeeze. Retail pricing is already far detached from Nvidia’s original guidance; one retailer lists its cheapest RTX 5090 at a level that sits 83% above the stated MSRP. While Nvidia has not changed public pricing, partners buying chips and memory at elevated rates are under pressure to pass these increases to consumers, setting the stage for further retail hikes.
From USD 1,999 to High-End Luxury: How Far RTX 5090 Pricing Has Climbed
The GeForce RTX 5090 debuted at USD 1,999 (approx. RM9,200), already one of the steepest launch prices for a consumer GPU. Even before the recent VRAM crunch, strong demand and opportunistic reselling saw cards frequently selling closer to USD 3,000 (approx. RM13,800). Now, with Nvidia reportedly adding another USD 300 (approx. RM1,380) hit to board partners due to VRAM sourcing costs, market-watchers expect typical street prices to move beyond USD 4,000 (approx. RM18,400). Some listings have already hovered around that level or higher, prompting speculation that fully custom flagship models could approach the USD 4,500–5,000 (approx. RM20,700–23,000) bracket. This escalation erodes the RTX 5090’s frames-per-dollar value, turning what was once a halo product for enthusiasts into an ultra-luxury component accessible only to a shrinking slice of buyers, while leaving the broader gaming audience effectively priced out.
VRAM Cost Increases and Supply Constraints Behind the Surge
At the core of the RTX 5090 pricing spiral is a GPU memory shortage that has made GDDR7 both scarce and expensive. Channel reports indicate that VRAM procurement costs have surged enough that Nvidia is simply passing the bill to add-in-card manufacturers, rather than absorbing it. Because the RTX 5090 uses more GDDR7 than any other RTX 50-series GPU, even a moderate per-chip increase translates into a substantial per-card hit. For now, only the RTX 5090 and RTX 5090D V2 are confirmed to be affected, but the underlying problem—a tight GDDR7 supply chain—is broader. That raises the risk that other RTX 50-series models could eventually see similar adjustments if memory markets remain constrained. Each incremental increase compresses partner margins and incentivizes them to ratchet up retail prices, magnifying the impact of the VRAM cost increase across the product stack.
Board Partners, Retailers, and the Rise of a Parallel Market
Board partners now have to factor an extra USD 300 (approx. RM1,380) into every RTX 5090 or RTX 5090D V2 they ship, on top of already inflated component and logistics costs. Retailers, in turn, are buying inventory far above Nvidia’s official MSRP, which explains why some storefronts list their “cheapest” RTX 5090 at huge premiums rather than acting as the primary source of price inflation. As supply tightens and official channels struggle to keep cards in stock at predictable prices, a parallel market thrives: resellers and grey-market operators can exploit both the scarcity of GDDR7-based cards and the willingness of deep-pocketed buyers to pay whatever it takes. With flagship GPU prices sliding into luxury territory, mainstream consumers are pushed down the product stack or out of the market entirely, while board partners walk a tightrope between volume, margin, and brand damage.