What We Know About Nvidia's Incoming RTX 30 Series
Nvidia's Ampere-based RTX 30 series promises double the performance-per-watt of Turing, with the $699 RTX 3080 leading the charge on September 17.
Nvidia finally pulled back the curtain on its next GPU generation last Tuesday, and the numbers being thrown around are hard to ignore. At the “GeForce Special Event” livestream on September 1, the company introduced three new cards built on its Ampere architecture: the RTX 3070, RTX 3080, and RTX 3090. The headline claim is roughly double the performance-per-watt compared to the outgoing Turing generation, which if it holds up in real-world testing would be one of the bigger generational leaps we’ve seen from Nvidia in years.
The lineup, briefly
The RTX 3080 is clearly the card Nvidia wants everyone talking about. It’s priced at $699 and set to launch September 17, which is a surprisingly aggressive price point given how the RTX 2080 launched at a noticeably higher cost two years ago. The 3070 and 3090 round out the stack, with the 3090 presumably aimed at the enthusiast/creator crowd who want the absolute ceiling of performance, though pricing and firm release dates for those two haven’t been fully nailed down in what’s been shared so far.
Why “double the efficiency” matters
Performance-per-watt is the metric that actually tells you something useful about a new architecture, more so than raw frame rates in a single cherry-picked benchmark. If Ampere really does deliver twice the efficiency of Turing, that headroom can go one of two ways: dramatically higher performance at similar power draw, or similar performance at a fraction of the power draw. Given Nvidia’s marketing angle here, it sounds like they’re leaning into the former — pushing clock speeds and core counts up while letting power consumption climb too, betting that gamers care more about raw output than a quiet, cool case.
It’s worth staying a little skeptical about vendor-supplied efficiency claims until independent reviewers get cards in hand. “Double” is a big, round, marketing-friendly number, and how Nvidia is measuring it (which games, which resolutions, which settings) hasn’t been spelled out yet.
What to watch for between now and launch
With the 3080 landing September 17, we’re less than two weeks out from the first real benchmarks. A few things I’ll be watching closely: how AMD responds with its own next-gen RDNA2 cards, whether ray tracing performance sees the same kind of jump as raster performance, and whether $699 street pricing actually holds once retailers get their hands on limited launch-day stock. GPU launches have a way of turning into refresh-button chaos on release day, and a genuinely well-priced flagship card tends to sell out instantly.
For now, this is shaping up to be the most interesting GPU generation in a while. If the efficiency claims hold up even partially, it’s good news for anyone building a new rig this fall — or anyone still gaming on a card from three or four generations back who’s been waiting for a real reason to upgrade.