Quick Answer
GPU prices are rising sharply in August 2026 and are unlikely to fall before late 2027 at the earliest. The RTX 5090 now lists for $4,929.99 on Nvidia’s own store, and AMD confirmed at least a 10% Radeon price increase effective this month. If you need a card now, buy before more hikes land. If you can wait 12 months or more, last-generation cards offer better value today.
Key Takeaways
- The RTX 5090 now sells for $4,929.99 on Nvidia’s official store, up from its $1,999 MSRP at launch in January 2025, according to Tom’s Hardware price data from August 5, 2026.
- MSI told investors its gaming hardware prices will rise 15–30% across 2026, citing a roughly 20% GPU supply shortfall from Nvidia and a severe DRAM shortage driven by AI demand.
- AMD confirmed at least a 10% price increase on Radeon GPU kits effective August 2026, ending any meaningful price gap between Radeon and GeForce at equivalent performance tiers.
- The root cause is structural: Samsung, SK Hynix, and Micron are allocating wafer capacity to high-bandwidth memory for AI data centers, leaving less supply for the GDDR6 and GDDR7 used in consumer graphics cards.
- Analysts at TrendForce, IDC, and Intel point to 2027 or 2028 before memory supply meaningfully normalizes, meaning further GPU price increases before Q4 2026 are more likely than not.
What is actually happening to GPU prices right now?
GPU prices across both Nvidia and AMD product lines are increasing at a rate not seen in a single consumer hardware generation, and the increases accelerated in late July and early August 2026. The Asus ROG Astral RTX 5090 is now listed on Nvidia’s website for $4,929.99 before tax, an increase of $500 (11.28%) applied in early August, compared to the original $1,999 MSRP. That is not an isolated premium model anomaly. Through the first half of 2026, an artificial-intelligence buildout of unprecedented scale has swallowed the world’s memory supply, and GPU prices are climbing across every tier, with memory chips now accounting for more than 80% of a GPU’s bill of materials.
Major end-of-July price increases rolled through China first, and by early August, ASUS and Gigabyte officially announced further price bumps to their distributing partners, with MSI following to confirm its graphics products will see up to 20%+ increases due to memory shortages. Based on information from Board Channels, all models from the RTX 50 series down to the RTX 30 series are affected. This is not a shortage affecting only flagship cards. The price pressure spans the full consumer lineup from Nvidia’s entry-level Blackwell cards through its top-tier models.
AMD notified partners of at least a 10% increase on Radeon GPU and memory kits effective August 2026, following Nvidia’s recent hikes. The RX 9050, which listed at $279, is now selling above $450, while midrange models have moved closer to elevated Nvidia pricing. The practical result is that neither major GPU vendor currently offers a meaningful price-per-tier advantage over the other at current street prices.
Why are GPU prices rising so fast in 2026?
The cause is a structural reallocation of memory manufacturing capacity, not a temporary supply disruption. Every major cloud provider and AI lab is racing to deploy Nvidia H100, H200, and B200 GPUs for training and serving large language models. Those data-center GPUs use HBM3 and HBM3e memory, manufactured on the same production lines as consumer memory. When Samsung, SK Hynix, and Micron allocate more capacity to HBM3 for data-center orders, less capacity remains for GDDR7 and standard DRAM.
AI data centers need enormous quantities of DRAM and HBM, and that memory carries much higher margins than the GDDR7 used in a gaming card, so memory makers allocate wafers to the AI side first. IDC has estimated AI data centers could absorb around 70% of the world’s memory output in 2026, up from 20–30% a few years ago. Demand for GDDR7, GDDR6, and high-bandwidth memory has surged as AI data-center deployments accelerate, pushing supply well beyond available capacity. Sources cited by TrendForce indicate memory prices have risen by several hundred percent in recent months, with VRAM now accounting for more than 80% of the total bill of materials for some high-end GPUs.
The price of 16 GB of GDDR7 reportedly jumped from roughly $65–$80 in mid-2025 to over $200 by the end of the year once long-term supply contracts expired and cards were exposed to spot pricing. That cost increase flows directly into retail pricing, because a graphics card manufacturer cannot absorb a tripling in its single largest component cost.
How far above MSRP are current GPU prices?
The table below shows where key Nvidia and AMD cards stand as of early August 2026 in the US market, compared to their original MSRPs. Street prices vary by retailer and change frequently; treat these figures as directional and verify current pricing before purchasing.
| GPU Model | Launch MSRP | Early Aug 2026 Street Price (US) | Approx. Premium vs. MSRP |
|---|---|---|---|
| RTX 5090 (Founders Edition) | $1,999 | $4,329–$4,929+ | ~117–147% |
| RTX 5080 | $999 | ~$1,400–$2,100 | ~40–110% |
| RTX 5070 Ti | $749 | ~$950–$1,000 | ~27–34% |
| RX 9070 XT (16GB) | $599 | ~$700–$989 | ~17–65% |
| RX 9070 (16GB) | $549 | ~$600–$640 | ~9–17% |
| RTX 5060 Ti (16GB) | ~$429 | Near RTX 5070 pricing | Significant tier compression |
Sources: Tom’s Hardware GPU price tracker (August 7–8, 2026), Best Value GPU price tracker (August 2026), Notebookcheck Hardware Unboxed price analysis, HotHardware. Street prices change daily; always confirm before buying.
Even the budget-focused RTX 5060 Ti 16GB has jumped to near-parity with the RTX 5070 and RTX 5070 Ti, making the entire enthusiast GPU market appear permanently repriced. Founders Edition models remain the closest to MSRP when available, but they sell out within minutes and supply is limited. For most buyers, custom AIB models are the only realistic option, and those carry substantial markups.
What did MSI warn investors, and why does it matter for consumers?
According to comments made during an investor briefing, MSI’s general manager and co-founder Huang Jinqing said the company expects gaming hardware prices to increase by 15% to 30% in 2026. Huang reportedly told investors that 2026 could be the most difficult, challenging, or “severe” year since MSI was founded in 1986. This is significant for consumers not because MSI sets Nvidia’s wholesale pricing, but because MSI is one of Nvidia’s largest add-in-board partners and has direct visibility into supply contracts and component pricing schedules.
As part of its investor meeting, MSI said there is a 20% GPU supply shortfall from Nvidia and that memory issues will necessitate more price increases. The company warned that 2026 will be its toughest year due to DRAM, storage, and GPU supply shortages driven by AI demand, and that the 20% GPU supply gap from Nvidia may shrink the PC market by up to 20%.
MSI’s price increases are expected to primarily affect lower-end and more affordable gaming hardware options, as the company is cutting back on producing these cheaper components. MSI plans to redistribute those resources to mid-range and high-end GPUs and hardware. This matters practically because it means budget-tier GPU buyers face proportionally larger price increases and reduced availability, while the mid-range and high-end markets see more supply concentration at elevated prices. For guidance on what recent price hikes mean for full system builds, see Qualcomm Is Raising Chip Prices September 1. Your Next Phone Will Cost More.
Is AMD Radeon a safer buy right now than Nvidia GeForce?
AMD Radeon cards have offered a partial buffer against the shortage, but that advantage has narrowed significantly in August 2026 and is likely to narrow further. AMD is fighting the same memory war as Nvidia. Its cards rely on GDDR6, which is also caught in the shortage. The AMD Radeon RX 9070 XT launched at a $599 MSRP, but as memory costs climbed, board partners layered their own increases on top.
ASUS, for example, quietly raised US prices on some RX 9070 XT models by up to 17.5%. The Prime RX 9070 XT OC jumped from $799.99 to $939.99, and the TUF Gaming model went from $849.99 to $989.99. The AMD Radeon GPU price increase of at least 10% that takes effect in August 2026 came after AMD deliberately waited for Nvidia to move first, effectively ending any price advantage over GeForce at equivalent performance tiers.
That said, the RX 9070 (non-XT) remains the closest current-generation card to its MSRP in the US market. The Radeon RX 9070 GRE is available for around $549, while the RX 9070 retails for around $640, and the RX 9070 XT sits near $689 at some retailers. These prices are likely to increase as pre-hike channel inventory clears. A Radeon RX 9070 currently available near $600 could cost $660–$700 or more once AIBs are working exclusively through August-priced inventory.
For a broader look at how AI competition between vendors is reshaping the technology landscape, see Alibaba’s Qwen3.8-Max Just Launched: What the New Open-Source AI Model Means for You and Google DeepMind’s Demis Hassabis Steps Down as CEO: What It Means for Gemini and Your AI Tools.
When will GPU prices come down?
GPU price relief is not a near-term prospect for most buyers. Intel has pointed to 2028 before conditions normalize, and Silicon Motion has warned the shortage across DRAM, NAND, and HBM will persist into 2028. Gartner expects the storage crunch to run into 2027. New memory fabs take years to build, so meaningful relief is a multi-year horizon.
Research analysts have already warned of 50% price increases in DRAM prices in Q3 2026, followed by another 40% increase in Q4 2026. Analysts point to structural forces that will not ease soon. Global memory production capacity remains constrained. AI server build-outs show no sign of slowing. Forecasts suggest the squeeze could persist into 2027 or even 2028.
TrendForce’s baseline projection is for GDDR7 supply to normalize by Q4 2026, though this depends on data-center demand not accelerating further. Street prices for Blackwell GPUs should begin falling toward, but not reaching, MSRP in that scenario. The TrendForce baseline is the most optimistic credible projection currently available; the Intel and Silicon Motion timelines represent the more conservative end of analyst consensus. This is the first time in modern memory that a gaming generation is getting more expensive as it matures, and the suppliers themselves say relief is years away.
Should you buy a GPU now or wait?
The decision depends primarily on your timeline, performance needs, and budget. There is no universally correct answer, but the data points clearly in two directions depending on your situation.
Buy now if you need a card in the next 6 months. If you need a card, buying now is reasonable, because analysts expect prices to climb further before they fall, with relief unlikely before late 2027. A buyer locking in current pricing on a $400 card avoids approximately $40 in increases; on a $700 card, approximately $70. The window to purchase pre-August-hike inventory is narrow and closing as channel stock turns over.
Wait if you can defer the purchase by 12 months or more. For buyers who are not ready to purchase immediately, the most defensible option in the current market is a last-generation card, such as the Radeon RX 7800 XT, RX 7900 GRE, or Nvidia GeForce RTX 4070, built with memory that predates the current acute shortage. These cards use GDDR6X memory that is less directly constrained, and their prices reflect a market that was inflating more slowly earlier in the product cycle. If you can wait, watch for major sales events and consider a used previous-generation card, which often delivers better value during a shortage.
One scenario to monitor carefully is a rumored RTX 50 Super refresh. Nvidia has never officially confirmed an RTX 50 Super refresh, and reporting through 2026 has swung between “canceled” and “delayed,” with CES 2027 floated as the earliest realistic window. The reason is that the Super cards were rumored to use GDDR7 modules, which are the single most supply-constrained memory type, so the product intended to offer better value needs the exact component the shortage made scarce. Do not plan your GPU purchase around an unconfirmed refresh that faces a supply constraint on its most critical component. For context on how Windows platform changes can affect upgrade decisions, see Windows 11 Is Finally Getting Lighter on 8GB PCs: What It Means and Windows 11 August 2026 Update (KB5101684): New Features and How to Get It.
What practical steps should a US buyer take right now?
- Check the official Nvidia and AMD stores first. Founders Edition models remain closest to MSRP when available, but they sell out within minutes. Set up in-stock alerts using a reputable price tracker such as NowInStock or CamelCamelCamel for Amazon listings to catch Founders Edition drops before scalpers do.
- Compare AIB models across multiple US retailers. Prices at Best Buy, Newegg, Amazon, and Microcenter often differ by $50–$200 on the same AIB card on the same day. Check all four before purchasing. Prices are volatile and can change overnight without announcement, as the Asus RX 9070 XT price history demonstrates.
- Factor in the full system cost. The same memory shortage affects everything with RAM. System RAM (DDR5) and SSDs have also risen because they draw on the same DRAM and NAND supply. Budget for higher-than-expected RAM and storage pricing when planning a new build alongside a GPU purchase.
- Evaluate last-generation alternatives seriously. RTX 4070 and RX 7900 GRE cards offer strong 1440p performance at prices that, while elevated versus their own MSRPs, remain substantially below current-generation equivalents. Performance per dollar currently favors previous-generation cards for most gaming workloads below 4K at maximum settings.
- Avoid unverified third-party sellers. GPU shortage periods historically see counterfeit, refurbished-as-new, and return-fraud listings increase on marketplace platforms. Purchase from established, authorized US retailers and keep your receipt. If a listing looks significantly cheaper than current market rates, treat it with skepticism. For related guidance on spotting counterfeit or fraudulent app listings, see That ‘Free ChatGPT’ App Might Be a Scam: How to Spot Fakes.
FAQ
Why is the RTX 5090 so expensive right now?
The RTX 5090 is selling well above its $1,999 launch MSRP because it uses 32GB of GDDR7, the most supply-constrained consumer memory type available. TrendForce-cited sources indicate memory prices have risen by several hundred percent in recent months, with VRAM now accounting for more than 80% of the total bill of materials for some high-end GPUs. The Founders Edition, when it appears in stock, sells within minutes; custom AIB models currently list at $4,329–$4,929 at major US retailers.
Is AMD Radeon cheaper than Nvidia GeForce right now?
AMD confirmed at least a 10% Radeon price increase effective August 2026, ending the price gap over GeForce at equivalent performance tiers. The RX 9070 and RX 9070 XT do still undercut comparable GeForce cards in absolute dollar terms, but the gap is smaller than it was in early 2026 and is likely to narrow further as pre-hike channel inventory clears in coming weeks.
Will GPU prices go back down in 2026?
A significant price reduction in 2026 is unlikely. TrendForce’s baseline projection is for GDDR7 supply to normalize by Q4 2026, though this depends on data-center demand not accelerating further. The more conservative analyst positions from Intel and Silicon Motion point to 2027 or 2028 before the DRAM and GDDR shortage meaningfully eases. Expect prices to remain elevated or increase further before any sustained relief arrives.
Does buying a used GPU make sense right now?
Buying a used previous-generation GPU is one of the more practical strategies in the current market, provided the card comes from a verifiable source and has not been used heavily for cryptocurrency mining. Cards such as the Radeon RX 7800 XT, RX 7900 GRE, or GeForce RTX 4070 are built with memory that predates the current acute shortage, and their used market prices on platforms such as eBay tend to lag the new-card price surge by several weeks, providing a temporary window of relative value. Inspect listings carefully and ask for purchase proof or original packaging where possible.
What caused the 2026 GPU shortage, and is it different from 2021?
The 2026 shortage is structurally different from the 2020–2021 GPU shortage, which was driven primarily by cryptocurrency mining demand and pandemic supply-chain disruption. Nvidia’s GeForce RTX 50 series is the clearest barometer of the current crisis because it uses next-generation GDDR7, the exact memory class in shortest supply. AI data centers are consuming the majority of the world’s memory supply, driving up the cost of the DRAM and GDDR7 that graphics cards depend on. Memory has become the dominant cost in a modern GPU, so a memory price spike is a card price spike. Unlike the 2021 shortage, this one is driven by sustained, long-term capital expenditure by hyperscalers, which makes a rapid reversal significantly less likely. For more on how AI investment is reshaping consumer technology, see ChatGPT Is Closing In on 1 Billion Weekly Users: What That Number Means for You.