Nvidia's Ampere Leaks Have PC Gamers Counting Down to September
Leaked details on RTX 3080 and 3070 point to a big generational leap for Nvidia's Ampere GPUs, with AMD's RDNA2 waiting in the wings.
If you spend any time in PC gaming forums or hardware Twitter right now, you’ve noticed the temperature rising. Nvidia hasn’t officially said much yet, but the drip of leaks around its next-generation Ampere architecture has become a flood, and an official reveal is widely expected in September.
The two cards everyone’s fixated on are the GeForce RTX 3080 and RTX 3070. Leakers and industry watchers have been piecing together specs, board shots, and cooler designs for weeks, and the picture that’s emerging is a genuinely big jump rather than the incremental bump we sometimes get between generations. The numbers being thrown around suggest up to 50% more performance than the current Turing-based cards, paired with roughly half the power draw for a given performance level. If that holds up, it’s the kind of efficiency gain that changes what’s possible in a GPU, not just how fast it runs the same workload.
Why this generation feels different
Turing, which gave us the RTX 20-series back in 2018, was notable mostly for introducing real-time ray tracing and DLSS to consumer graphics cards. It was a foundational shift, but the raw rasterization performance gains over the previous Pascal generation were modest by historical standards. Ampere sounds like it’s aiming to fix that complaint directly: more traditional performance headroom, plus presumably a second generation of the ray tracing and tensor cores that power DLSS. Nvidia has had two years to refine both of those technologies, and a lot of the excitement stems from wanting to see how much better ray tracing and upscaling can get with mature hardware and drivers behind them.
Power efficiency doesn’t always get gamers’ hearts racing the way a big frame-rate number does, but it matters more than people give it credit for. Lower power draw at a given performance tier means quieter cards, cooler cases, and more headroom for factory overclocks or dense builds. If the leaks are accurate, it also hints that Nvidia made real architectural gains rather than just cranking up clock speeds and calling it a day.
The AMD wildcard
None of this is happening in a vacuum. AMD’s RDNA2 architecture is coming later this year too, and after the genuinely competitive RDNA cards earlier in 2020, there’s real anticipation that AMD could challenge Nvidia at the high end for the first time in years. RDNA2 is also expected to show up in the next-gen PlayStation and Xbox consoles, which gives it a different kind of pedigree and scale than past AMD GPU launches.
For gamers, a real fight between Nvidia and AMD at the top of the market is the best possible outcome. Competition tends to mean better pricing and faster iteration, and it’s been a while since the high end felt genuinely contested.
Nothing here is official yet, so treat the specific numbers as rumors until Nvidia actually gets on stage. But the shape of the story is clear: a September announcement is coming, the leaks suggest a meaningful leap in both speed and efficiency, and AMD is positioned to make this the most interesting GPU generation in years. I’ll be watching September closely, and I’d bet a lot of you will be too.