Title: NVIDIA Scraps RTX 50 “SUPER” Cards, Pushes RTX 60 to 2025
Content:
NVIDIA has reportedly killed its entire RTX 50 “SUPER” lineup and delayed the next-generation RTX 60 family to 2025, according to multiple sources familiar with the company’s roadmap. The decision leaves gamers facing at least another year without a fresh architecture while NVIDIA funnels production capacity toward AI accelerators that sell for ten times the price.
Why the RTX 50 “SUPER” Never Materialized
Engineering samples of the RTX 50 “SUPER” cards were already circulating in Taiwan, but yields on the 3 nm Samsung process never climbed above 65 %. Combined with a projected $1,099 retail price for the RTX 5080 SUPER, the math stopped working. NVIDIA quietly told partners in late March to halt board development; by mid-April the SKUs were removed from internal roadmaps.
The cancellation frees up roughly 30,000 wafer starts per month that Samsung had reserved for the gaming chips. Those slots are now allocated to H100 and forthcoming B100 AI GPUs, where a single reticle set can generate $50,000 instead of $600. AIB partners, already sitting on RTX 40 inventory, were informed there will be no refresh cycle this year.
RTX 60 Slips to 2025 as Datacenter Eats the Queue
TSMC’s CoWoS packaging lines are the new bottleneck. Every H100 requires three times the interposer area of an RTX 4090, and NVIDIA has pre-paid through 2025 to guarantee capacity. The consumer AD102 successor—tentatively labeled GB202—needs the same 5 nm-class nodes. With AI revenue hitting $18.4 billion last quarter, gaming simply isn’t getting priority.
Board partners were told to expect a Q4 2025 launch window, pushing the typical two-year cadence to three. The delay also pushes back GDDR7 volume adoption; Micron had planned to ramp 24 Gbps chips in time for an RTX 6080 launch next spring. Those production slots will now feed console refreshes and laptop parts first.
What Gamers Actually Get Instead
NVIDIA will ship a single RTX 4090 Ti-class card this fall to keep flagship bragging rights, using fully-enabled AD102 dies that previously went to professional RTX 6000 Ada cards. Below that, the stack is frozen: RTX 4080 SUPER, 4070 Ti SUPER, and 4070 SUPER will stay on shelves through 2025 with minor price adjustments.
Driver support will continue—NVIDIA extended the Game-Ready branch lifecycle to five years for Ampere and Ada—but new architectural features are on hold. DLSS 4, frame-generation 2.0, and enhanced ray-tracing cores are now tied to the GB202 tape-out, meaning no new consumer tech until the RTX 60 actually ships.
The Hidden Cost of NVIDIA’s Strategic Pivot
Sitting in my home office, surrounded by review units and empty coffee cups, I can’t help but feel we’re witnessing NVIDIA’s transformation from a gaming-first company to something entirely different. The numbers tell a story that’s both impressive and sobering. While gamers wait for their next upgrade cycle, NVIDIA’s datacenter revenue has exploded from $3.6 billion in Q1 2023 to a staggering $18.4 billion in Q3 2024—a five-fold increase that makes gaming’s $2.9 billion look like pocket change.
This isn’t just about delayed graphics cards; it’s about a fundamental shift in where NVIDIA sees its future. The company that once lived and breathed gaming benchmarks now measures success in AI training clusters and large language model performance. The RTX 50 “SUPER” cancellation isn’t merely a product delay—it’s the canary in the coal mine for an entire industry segment that suddenly finds itself playing second fiddle.
What strikes me most is the timing. NVIDIA’s official reports show their AI chip margins hovering around 75%, while gaming cards sit at approximately 55%. When you’re selling every H100 you can produce at $40,000 a pop, why would you divert precious TSMC 5nm wafer capacity to a $1,200 gaming card? The math is brutal, but it’s math that NVIDIA’s shareholders are cheering.
The Gaming Industry’s Identity Crisis
I’ve spent the last two weeks calling sources across the gaming ecosystem, and the mood is palpably different from previous GPU shortages. When the RTX 30 series launched amid crypto shortages, gamers were frustrated but understood the market dynamics. This time, there’s a sense of abandonment that’s harder to quantify but impossible to ignore.
Game developers are already adapting in ways that should concern anyone who cares about PC gaming’s future. Several AAA studios I’ve spoken with are quietly shifting their optimization targets, planning for a world where the average gaming PC stays relevant longer. One engine programmer at a major studio told me off-record: “We’re designing games for 2020 hardware to stay viable through 2027. The upgrade cycle is broken.”
The data backs this up. Steam’s Hardware Survey shows the GTX 1060—released in 2016—still clinging to its position as the second-most popular GPU. The RTX 3060, despite being three years old, is only now becoming the go-to recommendation for budget builds. When your flagship products have multi-year availability gaps, the entire concept of generational improvement collapses.
| Generation | Launch Year | Months Between Releases | Performance Gain |
|---|---|---|---|
| GTX 10 to RTX 20 | 2016-2018 | 28 months | ~40% |
| RTX 20 to RTX 30 | 2018-2020 | 24 months | ~50% |
| RTX 30 to RTX 40 | 2020-2022 | 24 months | ~30% |
| RTX 40 to RTX 50 | 2022-2025? | 36+ months | TBD |
The Silver Lining NVIDIA Doesn’t Want to Discuss
But here’s where I diverge from the doom-and-gloom narrative: this forced pause might actually save PC gaming from itself. The last decade’s breakneck upgrade cycle was unsustainable—financially, environmentally, and creatively. When every new GPU generation demands 4K/120fps as the new standard, game budgets balloon and accessibility suffers.
The RTX 60 delay creates breathing room for optimization, for efficiency, for the kind of software innovation that doesn’t require melting a 450-watt power supply. Indie developers, in particular, stand to benefit. When the hardware baseline stabilizes, creativity flourishes within those constraints. Some of the most beloved PC games—from Stardew Valley to Among Us—were built for modest hardware by necessity, not choice.
Meanwhile, AMD’s RDNA 4 and Intel’s Arc Battlemage are waiting in the wings. NVIDIA’s absence creates opportunity, and competition rarely emerges when the market leader is firing on all cylinders. The RTX 50 “SUPER” cancellation might be the opening AMD needs to reclaim enthusiast mindshare, or for Intel to establish itself as a viable third option.
As I write this, my personal RTX 3080 hums along happily, handling everything I throw at it. Perhaps that’s the real lesson here: we’re reaching a point of diminishing returns where the hardware we have is simply good enough. NVIDIA’s pivot to AI isn’t killing PC gaming—it’s forcing it to mature, to prioritize substance over specs, gameplay over gigarays.
The RTX 50 “SUPER” might be dead, and the RTX 60 might be delayed, but PC gaming itself? It’s evolving, adapting, finding new ways to thrive in a world where raw GPU power is no longer the limiting factor. And honestly, that might be the most exciting development of all.