Nvidia’s next generation of gaming graphics cards is not arriving on the timeline anyone expected a year ago. Multiple outlets, including Notebookcheck, Club386, and Wccftech, reported this week that the GeForce RTX 60 series will now skip a 2027 launch window entirely and land sometime in 2028. That is a full generation cycle later than the roughly two-year cadence gamers have grown used to since the Turing era.
The RTX 60 series was originally scheduled to begin mass production toward the end of 2027, according to reporting cited across the industry. That plan has now slipped again, pushing the earliest realistic retail availability into 2028. Layered on top of that delay is a second piece of bad news for buyers hoping for a mid-cycle refresh: the rumored RTX 50 “Super” lineup, including a 24GB RTX 5080 Super that reportedly reached a completed design stage, has been shelved. Put together, the picture is a GeForce lineup that stays frozen at Blackwell for at least two more years, while prices on existing RTX 50 cards keep climbing.
What the Reports Actually Say About the RTX 60 Delay
The RTX 60 series is Nvidia’s next gaming architecture after Blackwell, and it is built on the company’s Rubin platform, the same silicon family Nvidia has already detailed for data-center AI accelerators. Several outlets, including Notebookcheck and Wccftech, report that Rubin-based gaming GPUs were expected to reach mass production in the second half of 2027 based on earlier leaks. That window has now closed, with the newest wave of reporting pointing to a 2028 launch instead.
None of this comes from an official Nvidia roadmap. Every date tied to the RTX 60 series, from the original 2027 target to the revised 2028 estimate, traces back to industry leakers and supply-chain sourcing relayed by tech outlets. Nvidia has not published a launch window for Rubin-based gaming cards, and the company has not confirmed a specific delay. What has changed is the tone of the coverage: outlets that once framed a 2027 launch as a working assumption are now treating 2028 as the more likely floor, not the ceiling.
That shift matters because it stacks on top of an already unusual year. Nvidia has not shipped a new consumer gaming GPU generation at all in 2026, and reports from outlets including PC Gamer have consistently pointed to memory supply, not design readiness, as the constraint. For the rest of shattered.io’s chip and GPU coverage, see the hardware hub.
The RTX 50 Super Refresh Is Dead, at Least for Now
Before the RTX 60 delay became the headline, the first casualty of Nvidia’s 2026 gaming pause was the rumored RTX 50 Super refresh. The lineup was expected to bump VRAM on existing Blackwell cards, led by a 24GB RTX 5080 Super, giving buyers a mid-cycle reason to upgrade without waiting for a full new architecture. Reports say the design work on that card was finished. It simply never shipped.
Nvidia has not announced the RTX 50 Super series and has not confirmed it was ever planned as a public product, so treat “shelved” as the current status rather than a permanent cancellation. Still, the practical effect for buyers is the same either way: the RTX 5080 remains capped at 16GB, the RTX 5070 Ti and RTX 5070 see no memory bump, and nobody has an updated Blackwell option to consider while waiting on Rubin. The gap between “what’s on shelves” and “what was designed but never released” is now one of the more concrete data points in this story.
Why Memory, Not Chip Design, Is the Bottleneck
The recurring explanation across outlets covering this story is a global memory shortage, not a problem with Rubin’s silicon. GDDR7, the memory standard used on current RTX 50 cards, and HBM, used in AI accelerators, are both in short supply as memory makers redirect capacity toward data-center customers. AI buyers pay more per wafer and buy in volume that consumer GPU makers cannot match, so when memory suppliers have to choose where to allocate output, gaming cards lose.
This is not a new theme for anyone who has followed hardware coverage this year. Shattered.io reported in August that Intel’s CEO warned memory costs would not ease until 2028, and separate coverage tracked chip stockpiles falling below 10 days of supply industry-wide. The RTX 60 delay is arguably the clearest downstream consequence of that shortage yet: an entire GPU generation pushed back because the memory that would go inside it is being bought by someone else first.
Nvidia’s own commentary on the shortage has focused on demand rather than the delay itself. In prior statements relayed by Tom’s Hardware, the company has pointed to strong demand for GeForce GPUs and acknowledged that memory supply is constrained, without directly addressing the 2028 timeline for Rubin-based gaming cards.
RTX 50 Pricing Right Now: The Cost of Waiting
With no Super refresh and no RTX 60 series until 2028, current RTX 50 cards are the only game in town for at least two more years. That has consequences for pricing that are already visible. According to Tom’s Hardware’s GPU price tracker, street prices on nearly every RTX 50 card sit well above the launch MSRP, and the gap widens the closer you get to the top of the stack. Shattered.io tracked the early stages of that climb back when RTX 5090 prices first doubled toward $5,000, and the trend has not reversed since.
| GPU | Launch MSRP | Street Price (Sept 2026) | Premium Over MSRP |
|---|---|---|---|
| RTX 5090 | $1,999 | $5,000+ | 150%+ |
| RTX 5080 | $999 | $1,579 | ~58% |
| RTX 5070 Ti | $749 | $1,149 | ~53% |
| RTX 5070 | $549 | $799 | ~45% |
| RTX 5060 | $299 | Close to MSRP | Minimal |
The RTX 5090 is the clearest example of how far this has gone. Tom’s Hardware reported the card crossing the $5,000 mark, more than double its $1,999 launch price, driven largely by GDDR7 scarcity and AI buyers scooping up high-VRAM cards for local inference work. Lower-tier cards like the RTX 5060 have stayed closer to their original pricing, since they use less memory and draw less interest from AI workloads, but the overall trend across the stack points the same direction: up, and likely to stay there until Rubin actually ships.
Historical Context: How This Compares to Past Nvidia Generations
Nvidia’s gaming GPU cadence has been remarkably consistent for most of the past decade. Turing (RTX 20) launched in 2018. Ampere (RTX 30) followed in September 2020. Ada Lovelace (RTX 40) arrived in September 2022. Blackwell (RTX 50) was announced at CES in January 2025. Each gap ran close to two years, occasionally stretching slightly longer.
| Architecture | Consumer Series | Launch | Gap From Prior Gen |
|---|---|---|---|
| Turing | RTX 20 | 2018 | – |
| Ampere | RTX 30 | September 2020 | ~2 years |
| Ada Lovelace | RTX 40 | September 2022 | ~2 years |
| Blackwell | RTX 50 | January 2025 | ~2.3 years |
| Rubin | RTX 60 | 2028 (projected) | ~3 years or more |
If the 2028 timeline holds, the gap between RTX 50 and RTX 60 will be the longest stretch between GeForce generations in the modern RTX era, and by a wide margin. A three-year gap does not sound dramatic in isolation, but it means gamers who bought an RTX 4090 in 2022 expecting a two-generation upgrade path by 2026 are instead looking at the same architecture on the market for six years running, from Ada’s 2022 debut through a 2028 Rubin launch.
What We Actually Know About Rubin
Rubin is confirmed as Nvidia’s architecture following Blackwell, and it already exists in a data-center form factor aimed at AI inference workloads. The gaming variant that will eventually become the RTX 60 series is the part still shrouded in leaks rather than official specs. No exact model names, no confirmed die configurations, and no verified performance targets exist for RTX 60 cards at this point. Names like RTX 6090 or RTX 6080 circulate in leaker communities, but Nvidia has not confirmed any of them.
The safest way to read the current state of Rubin gaming GPUs is this: the architecture is real, the timing keeps slipping, and everything else, including specs, names, and exact release dates, remains speculation dressed up as reporting. Readers should treat any specific performance percentage or exact release date attached to the RTX 60 series with real skepticism until Nvidia says something directly.
Market Impact: Who Gains and Who Loses From a Longer Cycle
A frozen GPU generation redistributes winners and losers across the PC hardware market. Retailers and board partners selling current RTX 50 inventory benefit from sustained demand without the usual pressure to discount ahead of a new launch. Secondary markets for used RTX 40 and even RTX 30 cards stay stronger for longer, since buyers priced out of RTX 50’s inflated cards have fewer new alternatives to wait for.
Gamers and system builders are the clearest losers. Anyone planning to skip RTX 50 and wait for RTX 60 is now looking at a wait that could stretch past two years from today, with no guarantee prices ease in the meantime. Game developers targeting future hardware baselines also lose a data point, since a Rubin-class jump in ray tracing or AI-driven rendering performance that might have shaped 2027 engine roadmaps is no longer something they can plan around with any confidence.
Nvidia itself is a more complicated case. Gaming revenue is a smaller share of the company’s total business than it was five years ago, with AI data-center chips now dominating the income statement. A delayed RTX 60 series costs Nvidia less in relative terms than it would have a decade ago, even if it frustrates a loyal gaming customer base that helped build the brand in the first place.
Competitive Landscape: Where AMD and Intel Stand
Nvidia is not alone in pushing its next generation back. The same Wccftech report citing the 2028 RTX 60 timeline also states that most of AMD’s RDNA 5 lineup will not ship until 2028, with one exception: a chip codenamed AT2, reportedly built on RDNA 5, expected to reach market in 2027. Shattered.io covered that leak in detail separately, including what is known about AMD’s AT2 chip and the broader RDNA 5 delay.
The practical effect is that the two biggest players in discrete gaming GPUs are converging on the same delayed timeline, for the same underlying reason: memory supply. That removes the usual competitive pressure that might otherwise push one company to rush a product to market ahead of the other. Intel’s discrete GPU roadmap, further behind both companies in market share, is not directly implicated in the memory-driven delays reported this week, which could theoretically give Intel more runway to close the gap if its own Arc successors stay on schedule. Nothing in current reporting confirms that Intel will actually capitalize on that opening.
Nvidia’s Business Calculus: AI Chips First, Gaming Second
The choice Nvidia is reportedly making comes down to where scarce memory delivers the most revenue per unit. AI accelerators sell at prices and margins that consumer gaming cards cannot match, and hyperscale customers buy in volumes that dwarf gaming demand. When GDDR7 and HBM supply is tight, directing that memory toward AI chips produces a better outcome for Nvidia’s overall financial results, even if it means gamers wait longer for new hardware.
This is not a new posture for the company. Nvidia’s data-center segment has outgrown gaming by a wide margin over the past two years, a shift visible in the company’s own record quarterly earnings reports, and every public earnings call since has reinforced the same priority order. A delayed RTX 60 series fits a pattern rather than breaking one. What is new is how visible the tradeoff has become to ordinary buyers, who are now watching an entire consumer product category slow down because of decisions made in an entirely different part of the business.
What This Means If You’re Buying a GPU Right Now
For anyone shopping for a graphics card in September 2026, the calculus is straightforward even if the answer is unsatisfying. There is no new Nvidia gaming GPU coming in 2026 or 2027, the RTX 50 Super refresh that might have offered a mid-cycle option appears dead, and RTX 60 is not a realistic purchase target until 2028 at the earliest. Waiting for something better in the next 12 to 18 months is not a viable strategy based on anything currently reported.
That leaves buyers choosing between paying inflated prices for a new RTX 50 card, hunting the used market for RTX 40 or RTX 30 hardware, or delaying a purchase entirely and living with existing hardware longer than planned. None of those options are new advice, but the RTX 60 delay removes the one variable that might have changed the math: a near-term new generation that could pull prices down through normal product-cycle competition.
Predictions: What Happens Between Now and 2028
- RTX 50 prices stay elevated through 2027. Without a Super refresh or new generation to compete against, expect street prices on RTX 5080 and RTX 5090 to remain well above MSRP for at least another year.
- Nvidia issues no official RTX 60 date before mid-2027. Companies rarely confirm delays this far in advance, and Nvidia has stayed silent on specific dates throughout 2026.
- A scaled-down RTX 50 refresh could still surface in 2027. Even if the full Super lineup stays shelved, a narrower update, such as added VRAM on a single SKU, remains plausible if memory supply eases even slightly.
- AMD and Nvidia’s 2028 launches will likely overlap closely. With both companies citing the same memory constraints, expect RDNA 5 and Rubin gaming cards to arrive within months of each other rather than a full generation apart.
- Used GPU prices, especially RTX 4090 and RTX 3090, stay firm. A longer wait for new hardware keeps demand for capable older cards higher than it would be in a normal product cycle.
What Confirmation Would Actually Look Like
It is worth being precise about what would turn this from a leak-driven story into a confirmed one. Nvidia would need to either announce a specific RTX 60 launch window at an event like CES or GTC, or a company executive would need to address the timeline directly on an earnings call, the way the company has already discussed memory constraints affecting gaming revenue in general terms. Neither has happened as of this week. Until it does, treat every date attached to the RTX 60 series, including the 2028 estimate itself, as the current best guess from people outside Nvidia rather than a confirmed roadmap.
The one thing that has moved from rumor to broad consensus across outlets covering this story is the underlying constraint: memory supply, not chip design, is what’s actually limiting the pace of gaming GPU releases in 2026 and 2027. That part of the story is unlikely to change even if the exact 2028 date does.
Frequently Asked Questions
When will the Nvidia RTX 60 series launch?
No official date exists. Reports from Notebookcheck, Wccftech, and Club386 point to 2028, after the series slipped from an earlier late-2027 target. Nvidia has not confirmed a launch window.
Is the RTX 50 Super refresh canceled?
Reports indicate the lineup, including a 24GB RTX 5080 Super, was designed but has been shelved. Nvidia never officially announced the Super series, so there is nothing to formally cancel, but no release appears imminent.
Why is Nvidia delaying new gaming GPUs?
Reporting consistently points to a global shortage of GDDR7 and HBM memory. Memory suppliers are prioritizing AI data-center customers, leaving less capacity for consumer gaming cards.
Will RTX 50 prices drop before the RTX 60 series arrives?
Nothing in current reporting suggests near-term relief. With no new generation launching before 2028, RTX 50 cards face less of the pricing pressure that typically comes from a new product cycle.
What architecture will the RTX 60 series use?
Rubin, Nvidia’s architecture following Blackwell. Rubin already exists in a data-center AI form factor, but the gaming variant remains unconfirmed in terms of specs and naming.
Is this delay officially confirmed by Nvidia?
No. Every date and detail tied to the RTX 60 series comes from industry leakers and outlet reporting, not an Nvidia announcement.
How does this compare to AMD’s RDNA 5 timeline?
Similar reporting suggests most of AMD’s RDNA 5 lineup will also wait until 2028, aside from a single chip codenamed AT2 expected in 2027, tied to the same memory supply constraints affecting Nvidia.
Should I buy an RTX 5090 now or wait for the RTX 60 series?
Waiting means holding out until at least 2028 based on current reporting. Buyers who need a GPU now are better served evaluating current RTX 50 or used-market options rather than delaying for a generation that is more than two years away at best.
