Nvidia’s reported RTX 5000 Super plans may be facing an unexpected obstacle: memory costs rather than GPU performance. A single 3GB GDDR7 chip is said to cost around USD 60 to USD 70, putting major pressure on the production cost of cards built around the new memory configuration.
The reported pricing could make it difficult to launch the new cards at or near Nvidia’s suggested retail price. Nvidia has reportedly asked its board partners to hold back product launches while the cost situation remains unresolved.
Why the Memory Upgrade Matters
A 3GB GDDR7 chip would allow Nvidia to increase VRAM capacity without adding more memory modules or changing the card’s memory bus. That could simplify board design, but the cost of the higher-capacity chips may outweigh that advantage.
VideoCardz reported that standard 2GB GDDR7 chips are priced at roughly USD 20 each. The jump from 2GB to 3GB represents a 50% capacity increase, yet the reported chip price is more than three times higher.
| Memory Type | Capacity | Reported Price |
|---|---|---|
| Standard GDDR7 | 2GB | About USD 20 |
| Higher-capacity GDDR7 | 3GB | About USD 60 to USD 70 |
That difference can significantly raise the bill of materials for every graphics card using the larger chips. The financial strain is especially important for board partners, which have reportedly handled memory procurement independently since 2025.
If Nvidia expects the cards to stay close to MSRP, partners could be left with very thin margins. Some manufacturers may decide that producing those models is not worthwhile if every unit carries a risk of loss.
Several Models Are Reportedly Affected
The delay is said to affect multiple Super models that have appeared in leaks. GeForce RTX 5080 Super, GeForce RTX 5070 Ti Super, and GeForce RTX 5070 Super are reportedly among the products being prepared.
| Model | Reported Detail | Launch Status |
|---|---|---|
| GeForce RTX 5080 Super | Listed as a model in development | On hold |
| GeForce RTX 5070 Ti Super | Listed as a model in development | On hold |
| GeForce RTX 5070 Super | Listed as a model in development | On hold |
| GeForce RTX 5050 9GB | Reportedly linked to 3GB GDDR7 chips | Undecided |
The GeForce RTX 5050 9GB is also reportedly affected because it may use the same 3GB GDDR7 memory chips. Its release status has not been decided, according to the report.
Seasonic’s PSU calculator has added further weight to the reports. RTX 5070 Super and RTX 5070 Ti Super were added to the tool in September, while the RTX 5080 Super name has also appeared there.
AI Demand Adds to the Supply Pressure
High demand from AI computing companies is reportedly absorbing a large share of DRAM and NAND manufacturing capacity. That competition for supply may make memory-heavy consumer graphics cards harder to price competitively.
The issue may also persist beyond a short-term launch delay. SK Hynix previously projected that 2027 could be the industry’s worst period for memory availability.
The company expects market demand to exceed available supply even beyond 2030. Until Nvidia makes an official announcement, the RTX 5000 Super lineup and RTX 5050 9GB remain subject to reports and leaks rather than confirmed launch plans.
