Nvidia RTX 60 and AMD RDNA 5 GPUs Rumored for 2028
The next generation of gaming GPUs could reportedly take longer to arrive than expected, with hardware leaker Kepler_L2 claiming that Nvidia’s GeForce RTX 60 series and most of AMD’s RDNA 5 gaming lineup may not launch until 2028.
The reported delay is being attributed to a familiar pressure point in the semiconductor industry: rapidly expanding demand for AI accelerators.
According to the rumor, data center products are receiving priority because of their significantly higher revenue and margins, potentially diverting manufacturing capacity and engineering resources away from consumer graphics cards.
The claim would represent a major shift from earlier rumors that pointed toward 2027 launches for both companies’ next-generation gaming architectures.
However, there is an important caveat: neither Nvidia nor AMD has officially confirmed the rumored 2028 timeline, and the product names, specifications, launch dates, and supply-chain explanations discussed here remain based on third-party information.
⏳ RTX 60 and Radeon 10000 Series Could Slip to 2028 #
Kepler_L2 recently responded to questions on the AnandTech forums by claiming that Nvidia’s Rubin-based GeForce RTX 60 family and most of AMD’s RDNA 5-based Radeon 10000 series could arrive in 2028.
The claim reportedly covers a large portion of both companies’ future consumer GPU lineups rather than a single delayed product.
If accurate, the delay would place the next major Nvidia and AMD gaming GPU generations considerably later than some previous leaks suggested.
The rumored explanation centers on production priorities.
AI data center hardware has become one of the industry’s most important sources of demand, while high-end accelerators command considerably greater selling prices than conventional gaming GPUs.
That creates a potential allocation problem when advanced manufacturing capacity, packaging resources, memory, and engineering teams are already under pressure.
In such an environment, consumer GPUs could become less urgent from a business perspective.
The result would not necessarily be that gaming GPU development stops. Instead, product launches could be scheduled around the availability of manufacturing and packaging capacity that remains after higher-priority data center products are allocated.
🤖 AI Demand Could Be Reshaping GPU Production Priorities #
The broader industry trend provides context for the rumor.
Modern AI accelerators require advanced process nodes, high-bandwidth memory, advanced packaging, large substrates, and substantial manufacturing capacity. At the same time, cloud providers and AI companies continue expanding data center infrastructure.
Gaming GPUs use many of the same underlying semiconductor resources, even though their designs and target markets differ.
This creates an increasingly important question for GPU roadmaps:
When manufacturing capacity is constrained, which products receive priority?
From a business standpoint, data center accelerators can generate substantially more revenue per unit than consumer graphics cards.
That does not prove that gaming GPUs will be delayed, but it provides a plausible economic mechanism behind the current rumor.
For Nvidia in particular, the Rubin generation is being developed first and foremost around its data center AI platform. Public demonstrations have so far focused on the Vera Rubin platform and related AI infrastructure rather than consumer GeForce products.
AMD has similarly been expanding its data center accelerator portfolio while continuing development of future Radeon architectures.
Neither company has publicly provided a finalized consumer roadmap that confirms the rumored dates.
🔵 Nvidia Rubin Gaming GPUs Remain Unconfirmed #
One of the biggest uncertainties concerns the consumer version of Nvidia’s Rubin architecture.
Nvidia has publicly discussed Rubin in the context of data center computing, including its Vera Rubin platform.
However, a consumer GeForce product based on Rubin has not been officially announced.
That leaves considerable room for speculation over how Nvidia might adapt the architecture for gaming.
A consumer GPU could differ substantially from its data center counterparts in:
- GPU configuration
- Memory subsystem
- Packaging
- Power limits
- Die configuration
- Ray-tracing hardware
- AI acceleration
- Product segmentation
Consequently, the existence of a Rubin data center architecture does not by itself establish the specifications or launch schedule of a future GeForce RTX 60-series product.
The same applies to the rumored 2028 release date.
Until Nvidia publishes a consumer GPU roadmap, the timing remains uncertain.
🟠 AMD’s AT2 Could Be the 2027 Exception #
The current rumor contains one notable exception on the AMD side.
An AMD GPU known by the codename AT2 is still rumored to launch sometime in 2027, potentially ahead of the rest of the RDNA 5 lineup.
According to earlier third-party reports, AT2 could be an RDNA 5-based GPU with:
- 8 Shader Engines
- 5 Compute Units per Shader Engine
- 40 Compute Units in total
- 12GB of VRAM
- 192-bit memory interface
The chip is reportedly positioned somewhere in the mid-to-high-end segment.
Other rumors have suggested performance targeting around Nvidia’s current GeForce RTX 5080, potentially placing it above AMD’s existing Radeon RX 9070 XT in the performance hierarchy.
However, these specifications and positioning claims have not been confirmed by AMD.
The AT2 situation therefore illustrates the broader uncertainty surrounding RDNA 5.
Rather than a clearly established product roadmap, the current picture consists of multiple leaks with different assumptions about launch timing and market positioning.
📅 Earlier Reports Pointed to a 2027 Launch Window #
The latest 2028 rumor also conflicts with information reported earlier this year.
In June, Dutch technology outlet Tweakers reported comments from several AMD board partners concerning the company’s next-generation GPU schedule.
The responses reportedly varied considerably.
Some manufacturers expected shipments to begin around the middle or latter part of 2027, while others suggested that availability could move into early 2028.
That disagreement is significant because board partners generally need advance information about production schedules and product launches.
However, varying expectations also indicate that the roadmap may not yet be firmly locked down—or that different partners may have been referring to different products.
The resulting range of possibilities currently spans much of 2027 and 2028.
⚔️ RTX 60 Rumors Are Already Contradicting Each Other #
Nvidia’s next-generation gaming GPU schedule is similarly uncertain.
A previous report from YouTube channel Moore’s Law Is Dead claimed that Nvidia was preparing to launch the RTX 60 series, including a flagship RTX 6090, during the first half of 2027.
Kepler_L2 subsequently disputed that timeline.
The conflicting claims demonstrate why early GPU roadmaps should be treated cautiously.
At this stage, there are effectively competing scenarios:
| Rumored Timeline | Current Claim |
|---|---|
| First half of 2027 | Previous RTX 60 launch rumor |
| Mid-to-late 2027 | Some AMD board-partner expectations |
| Early 2028 | Some earlier AMD partner expectations |
| 2028 | Kepler_L2’s latest claim for most RTX 60/RDNA 5 products |
None of these dates currently represents an official Nvidia or AMD consumer-GPU roadmap.
🧠 Why 2028 Would Matter for Gamers #
If the 2028 timeline eventually proves accurate, the delay could have several consequences for the consumer GPU market.
First, current-generation products would remain on the market for longer.
That could extend the lifespan of existing architectures and potentially give manufacturers more time to refine drivers, firmware, board designs, and memory configurations.
Second, the performance gap between successive generations could become more complicated.
GPU progress is not determined solely by architecture. Process technology, memory bandwidth, cache capacity, power limits, software optimization, and packaging all influence real-world performance.
A longer development cycle could therefore allow manufacturers to pursue larger architectural changes, but it could also leave consumers waiting longer for new performance tiers.
Third, the continued expansion of AI infrastructure could reinforce a structural divide between consumer and data center GPU development.
The same companies increasingly operate in both markets, but the economics of those markets are very different.
💾 Memory and Packaging Could Also Become Constraints #
The potential delay should not be viewed purely as a GPU wafer-capacity issue.
Next-generation graphics cards depend on an entire supply chain.
Advanced GPUs require:
- Leading-edge semiconductor wafers
- High-performance memory
- Advanced packaging
- Substrates
- Power-delivery components
- Cooling hardware
- Board manufacturing
- Testing and validation capacity
AI accelerators place pressure on several of these areas simultaneously.
High-end AI hardware, in particular, increasingly relies on advanced packaging and high-bandwidth memory technologies that are also experiencing strong demand.
This means that even if GPU dies themselves can be manufactured, other parts of the supply chain can become bottlenecks.
For gaming GPUs, the situation is therefore more complicated than simply asking whether Nvidia or AMD has enough wafer capacity.
🔍 Official Roadmaps Are Still the Missing Piece #
At present, the most important fact is that neither Nvidia nor AMD has officially announced a complete next-generation gaming GPU roadmap matching these rumors.
For Nvidia, consumer Rubin-based GeForce products remain unconfirmed.
For AMD, detailed RDNA 5 gaming specifications and launch dates are likewise unannounced.
The conflicting third-party reports also make it difficult to establish a single reliable timeline.
The 2028 scenario should therefore be treated as a rumor rather than a confirmed postponement.
Likewise, the rumored AT2 specifications should not be considered final hardware specifications until AMD provides official information.
🚀 The GPU Market Is Entering a Different Kind of Competition #
The most interesting aspect of these rumors may be less about whether a particular GPU launches in 2027 or 2028 and more about what is driving the uncertainty.
The gaming GPU market is increasingly connected to the much larger AI hardware ecosystem.
Advanced manufacturing capacity, packaging technologies, memory supply, and engineering resources are being pulled toward AI infrastructure at unprecedented scale.
That creates a different competitive environment from previous GPU generations.
Instead of consumer graphics cards existing largely within their own product cycle, future gaming GPUs may increasingly be scheduled around the requirements of a much larger AI hardware market.
For gamers, the practical takeaway is straightforward: 2027 and 2028 remain competing rumor windows, not confirmed launch dates.
Until Nvidia and AMD publish official roadmaps, reports of RTX 60 and RDNA 5 delays should be treated as provisional. The eventual schedule will depend not only on GPU architecture development, but also on manufacturing capacity, memory availability, packaging, and the continuing demand for AI accelerators.