Nvidia’s next-generation GeForce RTX 60 series will not arrive in 2027 after all, and as of October 2026, Nvidia has still not announced any RTX 60 model, name, or specification. Multiple hardware outlets, including Notebookcheck, Club386, TweakTown, OC3D, glitched.online and 24/7 Wall St., reported in mid-September 2026 that the Rubin-based RTX 60 lineup is sliding into 2028, mirroring a similar delay hitting most of AMD’s upcoming RDNA 5 graphics cards. The reporting, current as of September 16, 2026 and reinforced by a Tom’s Hardware report on September 21, 2026 pointing to the same 2028 slip, marks the second or third time the RTX 60 launch window has moved in twelve months, and it puts both of the PC industry’s major GPU vendors on the same unusual track: stretching a single graphics architecture across three calendar years instead of the customary two.
The headline reason, according to the leak chain feeding these reports, is not a design problem or a yield issue. It is money. Nvidia and AMD are both reportedly steering scarce TSMC advanced-node wafer capacity toward AI accelerators, where margins dwarf anything a gaming GPU can generate, leaving consumer graphics cards further back in the production queue than gamers have seen in years.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
RTX 60 Series Release Date: From 2026 to 2027 to 2028
The RTX 60 series release date has been a moving target for over a year. Early industry chatter placed the Rubin-based gaming architecture on a 2026 launch, following Nvidia’s usual roughly two-year cadence after RTX 50 (Blackwell) debuted in January 2025. Kopite7kimi himself had projected an H2 2027 window as recently as January 2026, and by July 2026 Verdict Bits was still framing the Rubin architecture as a 2027-to-2028 proposition rather than a settled date. That window shifted first to early 2027, then firmed up into a target of mass production toward the end of 2027, a specific manufacturing schedule that Shattered.io and The FPS Review say Nvidia had been working toward rather than a loose launch guess, and now, per the outlets tracking leaker Moore’s Law Is Dead and known Nvidia-focused leaker kopite7kimi, to 2028. TechPowerUp was among the first to flag the shift, reporting on July 23, 2026 that the Rubin-based GR20x is a 2028 product, and kopite7kimi put it just as bluntly on September 21, 2026, writing that “GR20x is a product from 2028,” a line Tom’s Hardware read as confirmation that the Rubin-based RTX 60 lineup has been pushed to at least 2028.
Notebookcheck’s latest framing is blunt: RTX 60 will effectively skip a 2027 launch entirely, following the same pattern already dogging AMD’s RDNA 5. Club386 has reported the same conclusion, saying the RTX 60 series skips 2027 outright and lands in 2028. That skepticism had been building for months: Gagadget wrote on June 13, 2026 that RTX 60 was expected no earlier than late 2027, a call that looks conservative in hindsight but was already hedging against an on-time 2027 launch well before the 2028 consensus formed. TweakTown’s coverage from earlier in September still floated a first-half-2027 window sourced to Moore’s Law Is Dead, but flagged the same underlying constraints, rising TSMC 2nm production costs and tight memory capacity, that other outlets now cite as the reason the date slipped again. That inconsistency between “early 2027” and “2028” across even a single week of reporting says as much about how fluid these leaks are as it does about Nvidia’s actual internal roadmap.
None of this comes from an official Nvidia statement. Every data point traces back to anonymous sources inside Nvidia’s add-in-board (AIB) partner network, or to leakers with a track record of accurate GPU roadmap calls but no formal confirmation from Santa Clara. Shattered.io’s aggregation of the leak chain makes that explicit: Nvidia has not published any launch window for RTX 60 and has not confirmed the delay in any form. Nvidia has not announced a next-generation gaming architecture, a launch date, or even a codename beyond the informal “Rubin” label that has circulated since 2025, and its current GeForce lineup still identifies the RTX 50 Series as its newest generation.
AMD’s RDNA 5 Faces the Same Wall
AMD’s situation looks nearly identical. Board partner feedback gathered around Computex 2026 indicated that RDNA 5 is “at least a year away,” with the more optimistic partners still hoping for a Q2 or Q3 2027 rollout and others calling that timeline too aggressive, pointing instead to late 2027 or early 2028. Leaker Kepler_L2 made the same call more bluntly on the AnandTech forums, writing that “Everything else from both AMD and Nvidia is 2028,” collapsing that earlier range of estimates into the same single-year target already circulating for Nvidia’s Rubin lineup. That would stretch the gap since RDNA 4’s March 2025 debut, when the Radeon RX 9070 XT and RX 9070 launched in the $550 to $600 range, to roughly three years, well beyond AMD’s typical two-year Radeon refresh rhythm.
What makes this notable is that it is not an isolated Nvidia decision. When two competitors with almost no coordination between them independently push their next flagship architectures into the same delayed window, it points to a shared external constraint rather than company-specific missteps. In this case, that shared constraint is the foundry itself: Tom’s Hardware and other outlets tracking TSMC’s roadmap have repeatedly noted that both Nvidia and AMD depend on the same limited pool of advanced-node capacity at the same fabs.
Why AI Demand Is Eating the Wafer Queue
The consistent explanation across the leak chain is that Nvidia is deliberately freeing up as much TSMC manufacturing capacity as possible for its AI data center products. Reporting from Tom’s Hardware and TweakTown points specifically to production of Vera Rubin, Nvidia’s next-generation AI superchip platform and the successor to Blackwell, as a priority consuming that capacity. Vera Rubin is a distinct product from the Rubin-codenamed gaming architecture rumored to power RTX 60, but the two reportedly compete for the same limited pool of leading-edge wafers. Nvidia’s AI accelerators, including Vera Rubin and the Blackwell-class B100 and B200 lines already shipping, sell at gross margins that a consumer GPU simply cannot match, and demand from hyperscalers has shown no sign of slowing through 2026. Every wafer TSMC allocates to a Rubin gaming die is a wafer not available for a data center chip that can sell for tens of thousands of dollars per unit.
That tradeoff used to be manageable because gaming GPUs and AI accelerators often ran on different process nodes or different fabs. That gap is narrowing. Rubin is reportedly targeting TSMC’s 2nm (N2) node, the same leading-edge process line that AI accelerators compete for, rather than a slightly older node reserved mostly for consumer parts. Supply chain trackers such as DigiTimes have documented how N2 capacity ramps have lagged the surge in AI chip orders, leaving less room for anything that is not a top-margin data center product.
Wccftech and other outlets covering the leak ecosystem have framed this the same way: 2027 was never truly abandoned as a target so much as deprioritized every time AI order books grew. Each quarter that Nvidia’s data center revenue outpaces expectations reportedly makes it easier, internally, to justify pushing the gaming lineup back another few months.
RTX 50 (Blackwell) Becomes a Three-Year-Long Generation
The practical result is that RTX 50, unveiled at CES 2025 with the RTX 5090 at $1,999 and the RTX 5080 at $999 (both launched on January 30, 2025, with the RTX 5060 following in May 2025), is on track to remain Nvidia’s current gaming flagship lineup for close to three years rather than the usual two. As iTech Guides noted in August 2026, RTX 60 is still expected to simply follow on from the RTX 50 generation whenever it finally lands, underscoring just how long Blackwell has had to carry Nvidia’s gaming lineup on its own. Reports circulating over the summer suggested Nvidia may fill part of that gap with an RTX 50 “Super” refresh in 2027 rather than a full next-generation architecture, similar to how the company extended the RTX 40 series with Super variants late in that generation’s life.
That approach lets Nvidia keep new SKUs on shelves without committing scarce leading-edge wafer capacity to an all-new die. It also means gamers hunting for a meaningful generational leap, rather than a modestly higher-clocked or memory-bumped refresh, are now looking at a multi-year wait. Tech Insider has previously covered how RTX 5090 pricing has climbed well past its original MSRP amid persistent supply constraints, and an extended Blackwell shelf life does nothing to relieve that pressure.
GPU Launch Cadence: Then vs. Now
To put the scale of this delay in context, it helps to look at how consistently both companies have shipped new architectures roughly every two years for the past decade. The table below lays out the recent cadence for both Nvidia and AMD’s consumer GPU lines.
| Generation | Vendor | Architecture | Launch Year | Gap From Prior Gen |
|---|---|---|---|---|
| GTX 10 series | Nvidia | Pascal | 2016 | ~2 years |
| RTX 20 series | Nvidia | Turing | 2018 | ~2 years |
| RTX 30 series | Nvidia | Ampere | 2020 | ~2 years |
| RTX 40 series | Nvidia | Ada Lovelace | 2022 | ~2 years |
| RTX 50 series | Nvidia | Blackwell | 2025 | ~2.3 years |
| RTX 60 series | Nvidia | Rubin | 2028 (reported) | ~3 years |
| RX 5000 series | AMD | RDNA | 2019 | — |
| RX 6000 series | AMD | RDNA 2 | 2020 | ~1 year |
| RX 7000 series | AMD | RDNA 3 | 2022 | ~2 years |
| RX 9000 series | AMD | RDNA 4 | 2025 | ~3 years |
| Next Radeon | AMD | RDNA 5 | 2028 (reported) | ~3 years |
Two three-year gaps back to back, for RDNA 4 to RDNA 5 and now RTX 50 to RTX 60, would represent the longest stretch without a full architectural refresh either company has had in the GPU market this decade. For a market that has grown used to a reliable two-year drumbeat, that is a meaningful change in how these companies plan product cycles.
How the Rumor Timeline Has Shifted
Tracking the reported RTX 60 launch window across the past year and a half shows just how much it has moved, and how quickly outlets have had to revise their own coverage. The earliest calls were far more optimistic: PC Guide wrote in August 2025 that RTX 60 would arrive no earlier than 2026, a window that looks almost quaint next to the 2028 target now circulating.
| Reported Date | Outlet / Source | Claimed RTX 60 Window at the Time |
|---|---|---|
| Nov 2025 | Notebookcheck, citing leaker RGT | Q1/early Q2 2027, AIBs already skeptical |
| Jan 2026 | Notebookcheck, citing kopite7kimi | Second half of 2027 |
| Jul 2026 | Notebookcheck, citing Moore’s Law Is Dead | No full RTX 60 launch in 2027; RTX 50 Super instead |
| Sep 14, 2026 | TweakTown, citing Moore’s Law Is Dead | First half of 2027, capacity concerns flagged |
| Sep 16, 2026 | Notebookcheck | 2027 launch skipped outright; 2028 target |
| Sep 2026 | Club386 | RTX 60 series skips 2027 outright, lands in 2028 |
| Sep 2026 | Shattered.io and The FPS Review | RTX 60 had targeted mass production toward the end of 2027; now slipping to 2028 |
| Sep 2026 | Tom’s Hardware (kopite7kimi) and AnandTech forums (Kepler_L2) | “GR20x is a product from 2028”; “Everything else from both AMD and Nvidia is 2028” |
The pattern is not random noise. Each revision has pushed the date later, never earlier, and each one has cited the same root cause: TSMC capacity being redirected toward AI silicon. That consistency across eight separate reporting cycles, spanning nearly a year, is what gives this particular delay story more weight than a single unverified leak would carry on its own.
Where the RTX 60 Timeline and the RTX 6060 Rumor Stand
Nvidia’s current GeForce lineup still identifies the RTX 50 Series as its newest generation, and no RTX 60 model has been confirmed by Nvidia. Nvidia has announced no RTX 60 Series prices, and there is no official RTX 60 release date. Reports that began on October 20, 2026 claim the mainstream RTX 6060 could be delayed until 2029, but neither the name nor the date is official.
The 2029 claim is attributed to leaker Kepler_L2, who says the RTX 6060 could slip to 2029 because of semiconductor and high-density GDDR-memory shortages. TechSpot reported in October 2026 that RTX 60 might arrive in 2028, nearly 4 years after RTX 50. TechSpot also reported that the RTX 6060 could trail the RTX 5060 by at least 3 years 9 months. The RTX 5060 launched in May 2025, per PC Guide, which anchors that comparison.
For readers, the safest reading is that 2028 and 2029 are leaks rather than Nvidia plans. Names such as RTX 6090, RTX 6080 and RTX 6060 are unofficial and speculative. Another report in September 2026 suggested the first half of 2027, while later reporting cited conflicting claims pointing to 2028, so any specific date should stay provisional until Nvidia publishes a schedule of its own. Anyone deciding whether to buy now can weigh current RTX 50 pricing against a replacement that, by every current report, is still years away.
Memory Supply Adds a Second Bottleneck
Wafer capacity is not the only constraint being cited. TweakTown’s coverage specifically flagged constrained memory capacity as a factor pushing RTX 60 pricing and availability in a difficult direction, on top of the higher baseline cost of manufacturing on a 2nm node. The broader memory market has been under pressure through 2026 as AI accelerator demand for high-bandwidth memory has pulled DRAM production capacity away from consumer-facing parts, a dynamic Tech Insider has tracked in the context of VRAM-heavy modified GPUs commanding a premium in gray-market channels.
Even if Nvidia and AMD resolve their wafer allocation questions in time for a 2028 launch, memory suppliers face their own capacity fights between AI-grade high-bandwidth memory and the GDDR modules gaming cards depend on. A GPU architecture ready to tape out does not help much if the memory to pair with it is backordered or priced out of a reasonable consumer product.
Competitive Landscape: Nvidia, AMD and Intel Arc
Nvidia currently holds the dominant share of discrete GPU shipments, a lead Tech Insider reported had climbed to roughly 90 percent of discrete GPU shipments earlier this year. That dominance gives Nvidia more room to delay a consumer refresh without immediately losing ground, since AMD is delayed on a similar timeline and Intel’s Arc lineup has not scaled to compete at the high end.
AMD, by contrast, has less cushion. The company has leaned on RDNA 4’s price-to-performance positioning, with the RX 9070 XT trading blows with Nvidia’s midrange RTX 5070 lineup, a comparison Tech Insider examined in detail earlier this year. If RDNA 5 slips to 2028 while Nvidia manages to squeeze out an RTX 50 Super refresh in 2027, AMD risks facing a refreshed competitor with its own aging architecture for another full year, a tougher spot than facing an equally delayed Rubin lineup.
Intel’s Arc GPUs remain a value-tier alternative rather than a genuine flagship contender, and nothing in current reporting suggests Intel’s roadmap changes that dynamic before 2028. That leaves the practical high-end and midrange GPU competition as effectively a two-horse race for at least two more years, with both horses now running on delayed schedules.
What This Means for GPU Prices Through 2027
A longer gap before the next architecture typically means fewer reasons for retailers and board partners to discount the current generation. Historically, GPU prices soften in the months before a new architecture launches, as retailers clear inventory. If RTX 60 and RDNA 5 both push to 2028, that discounting window likely doesn’t open until sometime in 2027 at the earliest, and possibly not until early 2028 itself.
Combine that with the memory supply pressure already pushing DRAM and GDDR module prices upward through 2026, and the near-term outlook for anyone shopping for a new GPU in 2027 looks less favorable than it would in a normal generational transition year. Nvidia’s professional Blackwell-based cards have already shown how far pricing can climb when a product sits at the intersection of gaming and AI workloads, and an extended Blackwell generation gives that dynamic more time to play out at the consumer level too.
Historical Precedent: This Isn’t the First Multi-Year Gap
Long gaps between GPU generations aren’t unprecedented, though they are rare in the RTX and RDNA eras specifically. Nvidia’s Kepler-to-Maxwell transition and the extended life of the GTX 900 series before Pascal both saw the company stretch a single node’s product stack further than usual, generally driven by manufacturing node availability rather than demand from an adjacent business line. What’s different this time is the cause. Previous multi-year gaps were largely about foundry process nodes not being ready. This one is about a healthy, ready process being allocated somewhere more profitable, which is a structurally different problem that will not resolve itself just because a node matures.
AMD’s own RDNA 3 to RDNA 4 transition already ran closer to three years than its usual two, foreshadowing exactly this pattern before RDNA 5 reporting made it explicit. In hindsight, RDNA 4’s stretched timeline in 2025 looks less like an isolated scheduling issue and more like the first sign of the AI-driven capacity squeeze that is now shaping RDNA 5 and RTX 60 alike.
What Gamers and PC Builders Should Actually Do
For anyone planning a new build or an upgrade, the practical takeaway is straightforward: RTX 50 and RDNA 4 are not stopgap products waiting to be replaced next year. They are the products both companies plan to sell, refresh and support through at least 2027, and in Nvidia’s case, potentially with a Super refresh layered on top rather than a clean-sheet replacement.
- Anyone needing a GPU now should buy based on current-generation value, not on the assumption that RTX 60 or RDNA 5 arrives soon enough to wait for.
- Watch for an RTX 50 Super refresh in 2027 as the more realistic near-term upgrade path rather than a full Rubin launch.
- Expect used and current-generation card prices to stay firmer for longer than a typical pre-launch period, since there’s no imminent replacement pulling secondhand values down.
- Memory-heavy workloads, including AI inference on local hardware, may face indirect price pressure as GDDR and HBM allocation continues favoring data center customers.
Predictions: What Happens Between Now and 2028
Based on the current reporting pattern and the underlying capacity dynamics, a few outcomes look more likely than others heading into 2027 and 2028.
- An RTX 50 Super refresh ships in 2027. This looks like the most probable near-term move for Nvidia, giving retailers new SKUs without committing 2nm capacity to a full Rubin ramp.
- RDNA 5 launches in a limited capacity before RTX 60 does. AMD has less AI-revenue cushion than Nvidia, giving it more incentive to prioritize at least a partial gaming lineup refresh ahead of a full Rubin response.
- Further date slips remain likely through 2027. Given that every report over the past year has pushed the date later rather than earlier, a 2028 target should be treated as provisional, not final.
- GPU prices stay elevated relative to a typical pre-launch year. Without a looming replacement generation, neither company has strong pricing pressure to discount current inventory through 2027.
- AI capacity allocation becomes a recurring GPU-launch storyline. Expect this same tension, gaming silicon competing with data center silicon for the same wafers, to resurface with every future Nvidia and AMD roadmap leak until AI demand growth levels off or new fab capacity comes online.
The Bigger Picture: Gaming GPUs as a Secondary Priority
What’s happening with RTX 60 and RDNA 5 is really a symptom of a much larger shift in how chipmakers allocate their most advanced manufacturing capacity. Gaming GPUs used to be the flagship showcase for a new process node, the product that proved a node was ready for high-volume, high-performance consumer silicon. That role has effectively been handed to AI accelerators, and gaming graphics cards are now, by the accounts of multiple outlets, sliding into a lower-priority tier behind chips destined for hyperscale data centers.
That doesn’t mean gaming GPUs are going away. Nvidia and AMD both still generate meaningful, if smaller, revenue from gaming hardware, and neither company has any incentive to walk away from that business. But the era where a new gaming architecture reliably arrived like clockwork every two years appears to be ending, at least for as long as AI accelerator demand keeps outstripping the wafer supply available to make it.
Frequently Asked Questions
When will the Nvidia RTX 60 series actually launch?
Current reporting from Notebookcheck, Club386 and other outlets points to 2028, after the RTX 60 series was reported to skip a full 2027 launch. The series had previously been aiming for mass production toward the end of 2027 before that plan slipped as well. Nvidia has not confirmed a date, and the timeline has shifted multiple times over the past year, so treat any specific date as provisional.
Is AMD’s RDNA 5 delayed for the same reason as RTX 60?
Yes. Reports from AMD board partners around Computex 2026 pointed to a similar late-2027 or early-2028 window for RDNA 5, with the delay attributed to the same TSMC capacity pressures affecting Nvidia’s Rubin architecture.
Will Nvidia release anything new for gamers before RTX 60?
Reports suggest Nvidia may release an RTX 50 “Super” refresh in 2027 rather than a full next-generation architecture, similar to how it extended the RTX 40 series late in its life cycle.
Why is AI demand affecting gaming GPU launch dates?
Nvidia and AMD both manufacture their chips at the same foundries, primarily TSMC, and AI accelerators generate far higher margins per wafer than gaming GPUs. Reports indicate both companies are prioritizing wafer allocation for AI data center chips, pushing consumer gaming GPU production further back in the queue.
What process node will the RTX 60 series use?
Reports point to TSMC’s 2nm (N2) node for the Rubin-based RTX 60 architecture, the same leading-edge process line competing for capacity with AI accelerator production.
Should I wait for RTX 60 or RDNA 5 before buying a GPU?
Given the reported 2028 timeline, waiting is not practical for most buyers needing a GPU in the next year or two. Current-generation RTX 50 and RDNA 4 cards, along with any 2027 RTX 50 Super refresh, are the realistic near-term options.
How reliable are these leaks about the RTX 60 delay?
None of this information comes from an official Nvidia or AMD statement. It originates from leakers such as Moore’s Law Is Dead, kopite7kimi and Kepler_L2, along with anonymous AIB partner sources, as reported by outlets including Notebookcheck, Tom’s Hardware, TweakTown, Club386, Shattered.io and The FPS Review. The consistency of the delay narrative across multiple independent reports over nearly a year adds credibility, but it remains unconfirmed by either company, and Shattered.io notes Nvidia has not published any official launch window.
What exactly did leakers kopite7kimi and Kepler_L2 say?
Kopite7kimi, a leaker with a lengthy track record on Nvidia die codenames, wrote that “GR20x is a product from 2028,” which Tom’s Hardware read as confirmation that the Rubin-based RTX 60 lineup has been pushed to at least 2028. Separately, Kepler_L2 posted on the AnandTech forums that “Everything else from both AMD and Nvidia is 2028,” tying both companies’ next-generation GPU lineups to the same delayed target.
How does this compare to past Nvidia and AMD launch cadences?
Both companies have historically launched new GPU architectures roughly every two years. A 2028 arrival for RTX 60 and RDNA 5 would extend that gap to roughly three years for both companies’ current generations, the longest stretch either has had since the mid-2010s.
