Reading mode:
Key Takeaways
  • The flagship GeForce RTX 5090 has been reclassified as “Not Carried” by major retailers like Micro Center, sparking rumors of an abrupt production halt.
  • Nvidia Senior VP Jeff Fisher publicly denied the cancellation, stating that the card remains in production despite severe supply chain constraints.
  • Street prices for standalone GPUs have climbed between $7,300 and $8,500, leaving prebuilt gaming desktops as the only viable acquisition route.

The consumer graphics card market has been plunged into another wave of uncertainty. Nvidia’s flagship GeForce RTX 5090, which had already been notoriously difficult to buy at its $1,999 MSRP, has begun disappearing from major retail shelves across the United States. The abrupt drop in availability led to industry-wide speculation that Nvidia had quietly halted consumer production to redirect silicon toward high-margin enterprise AI accelerators.

Retail Exodus: Micro Center Reclassifies Flagship GPU

Signs of severe distress surfaced when leading U.S. electronics retailer Micro Center updated its inventory listings, designating custom GeForce RTX 5090 models as “Not Carried.” Unlike standard out-of-stock notices that signal pending restocks, this status typically indicates that a product is no longer part of the merchant’s active catalog. Physical stores reported completely bare shelves, leaving only the ultra-premium Asus ROG 20th Anniversary Bundle priced at nearly $11,000.

Reports on social platforms such as Reddit corroborated the retailer shift, with customers alleging that remaining floor units were abruptly recalled from display. Meanwhile, major online storefronts show a similar vacuum: Newegg and Amazon currently list units via third-party sellers at inflated price points ranging from $7,300 to $8,500, while Best Buy and Nvidia’s direct marketplace show zero available stock.

Interactive Hardware & AI BenchmarkFull VRAM & ROI Calculator →
Can Your GPU / AI Station Run 2026 Local AI Models?
✓ Optimal (100% GPU VRAM, ~55-85 tok/s)ollama run qwen3:14b

Hardware Specifications and Market Valuation

The market panic stems largely from the unprecedented computing power that the flagship delivers over lower-tier models, rendering it irreplaceable for extreme enthusiasts and professional creators:

Specification / Model GeForce RTX 5090 GeForce RTX 5080
CUDA Core Count 21,760 10,752
VRAM Capacity & Type 32GB GDDR7 16GB GDDR7
Memory Bus Width 512-bit 256-bit
Memory Bandwidth Up to 1,792 GB/s —
Official MSRP $1,999 —
Third-Party Market Price $7,300 – $8,500 —

Nvidia Denies Discontinuation Claims

The supply void spurred theories that Nvidia is reallocating the underlying GB202 silicon to its workstation lineup, including the RTX Pro 6000 Blackwell, RTX Pro 5500, and RTX Pro 5000 series, where profit margins are substantially larger. Addressing these rumors during a post-event interview reported by Taiwanese outlets including 4Gamers, Nvidia Senior Vice President Jeff Fisher firmly denied that the card has been discontinued.

Fisher clarified that while supply chain capacity remains severely constrained, Nvidia is actively working to deliver products to the consumer market. He emphasized that the GeForce gaming foundation remains vital to Nvidia’s identity, noting that enterprise AI capabilities exist today largely because of long-term investments in consumer gaming hardware.

The current hardware squeeze comes as enterprise demands command unprecedented attention, mirroring moves where Nvidia commits $250M to AI safety research and warns of chip supply bans across the broader technology ecosystem.

What This Means in Practice

For PC builders, acquiring a standalone GeForce RTX 5090 at reasonable pricing has become practically impossible in the near term. The most cost-effective alternative for securing the flagship silicon is purchasing prebuilt systems from OEMs like Dell, MSI, or Skytech, where the total desktop price often equals or undercuts scalper rates for the GPU alone.