In 30 seconds
In this January 2026 short, Matthew Berman connects rising memory prices with pressure on gaming GPU availability and prices. He discusses Nvidia CEO Jensen Huang’s response to a CES question about using older GPUs and argues that demand for AI infrastructure may disadvantage gamers. Huang entertained the idea without announcing a production plan, and Berman’s $5,000 RTX 5090 figure was a prediction, not a confirmed price.
- 1A price forecastBerman predicts steep GPU price increases and raises $5,000 as a possible RTX 5090 price. It is not an announced Nvidia price.
- 2Older GPUs as an optionAt CES, Huang responded to a question about older GPUs and lower-memory products by saying it was possible and worth examining. He did not announce a plan.
- 3AI and gaming demandBerman suspects Nvidia prioritizes memory for data-center AI chips, but provides no evidence of its allocation decisions.
Takeaways
- Memory supply pressures can affect gaming graphics card costs, but the extent of any future price rise is uncertain.
- Berman predicted an RTX 5090 might reach $5,000; Nvidia originally announced a $1,999 starting price in 2025.
- Jensen Huang described older GPU production and newer features on previous generations as possibilities, not confirmed products.
- Gaming GPUs and data-center AI accelerators can use different kinds of memory, so allocation claims need specific evidence.
Read the edited text by chapter
Memory costs and GPU pricing
Berman says memory prices have risen and predicts pressure on gaming GPU prices. His RTX 5090 figure is a forecast rather than a confirmed price.
Computer memory prices have risen sharply, and I worry that graphics cards could become much more expensive as memory supply tightens. I suggest an RTX 5090 could reach $5,000, but that is a January 2026 forecast rather than a confirmed Nvidia price. For context, [Nvidia announced the RTX 5090 at $1,999](https://nvidianews.nvidia.com/news/nvidia-blackwell-geforce-rtx-50-series-opens-new-world-of-ai-computer-graphics) at its 2025 launch; actual retail prices can differ.
Older GPUs and AI demand
He discusses a possible return to older graphics generations and potential upgrades to their features. He also speculates that AI demand could affect memory allocation and gamers.
One possible response is to use older GPU designs or cards with less memory. The RTX 30, 40 and 50 series represent successive generations, but their memory capacities vary by model rather than increasing uniformly. Less memory or an older process could potentially reduce costs, though neither guarantees lower retail prices.
At CES 2026, Nvidia CEO Jensen Huang was asked whether older GPU production or lower-memory products might help. He said that was possible and suggested that bringing newer AI technology to previous generations might also be feasible with substantial engineering. [The question and answer](https://www.tomshardware.com/pc-components/gpus/nvidia-non-committal-on-plans-to-solve-gpu-pricing-squeeze-ceo-jensen-huang-floats-bringing-ai-tech-to-older-models) did not announce a specific production or release plan.
I suspect Nvidia may prioritize memory for high-end AI chips used in hyperscale data centers, while gamers face higher prices. That allocation is my inference, not an established Nvidia decision. Gaming cards such as the RTX 5090 use GDDR7, while major data-center accelerators can use high-bandwidth memory; they are different products even if wider demand and manufacturing constraints affect memory markets.
Nvidia has received criticism from gamers who see the company focusing heavily on AI. The tension illustrates the economics of a market in which gaming and data-center customers compete for computing resources, though the size and duration of any impact on gaming prices remain uncertain.
More from this channel
34:21EN
Matthew Berman· Oct 27, 2026Matthew Berman Examines Anthropic’s IPO Filing, AI Growth and GovernanceMatthew Berman, Sam Altman · Forward FutureThe presenter reviews the S-1’s financial metrics, compute strategy and voting structure while questioning the tension between rapid AI development and safety.
0:26ENMatthew Berman· Oct 25, 2026Sam Altman on Human Values and Problem-Solving in an AI WorldSam Altman · OpenAIIn a short interview excerpt, Altman considers the enjoyment of solving hard problems and the human role in interpreting results.
0:36ENMatthew Berman· Oct 25, 2026Kimi K3.1 and a Rubik’s Cube Solver DemonstrationMatthew Berman · Forward FutureMatthew Berman shows a scrambled virtual cube being solved and gives the Kimi K3.1 test a favorable assessment.
0:21ENMatthew Berman· Oct 24, 2026An AI-Generated Kimi Benchmark Chart Used Invented ScoresMatthew Berman · Forward FutureMatthew Berman asks for a graph of Kimi K3.1 Pro benchmarks, then discovers that the displayed figures are illustrative placeholders.
25:50ENMatthew Berman· Oct 24, 2026Kimi K3.1 Pro Tested: Strong 3D Demos, Uneven Results and Competitive PricingMatthew Berman · Forward FutureMatthew Berman tests the open-weight model on coding, 3D rendering and visual explanations, then compares benchmark results and prices.
32:20ENMatthew Berman· Oct 23, 2026Who Is Local AI For? The Costs, Trade-offs and a Hybrid FutureMatthew Berman · Forward FutureMatthew Berman weighs data control and independence against hardware cost, setup complexity and inference speed.