After weeks of rumors and leaks, Nvidia finally revealed three new GPUs in its RTX 30 series, all sporting the new Ampere architecture underneath. If there’s one thing speculation before the event drove home, it was that these cards were going to be expensive. And while that’s certainly true for the flagship, Nvidia uncharacteristically got very aggressive with the pricing for the two cards most people will likely buy.
Nvidia revealed the RTX 3070, the RTX 3080, and the RTX 3090 during its GeForce Special Event, which Nvidia billed as the biggest generational leap forward in the last 20 years of GPUs. On paper, those claims seem sound. The RTX 3070 was described as faster than Turing’s flagship card, the RTX 2080 Ti, while the RTX 3080 boasts twice the performance of the RTX 2080. The RTX 3090, this generation’s Titan card, is in a league of its own–and with a price point of $1500, it’s not for even most enthusiast PC builds.
The pricing on both the RTX 3070 and RTX 3080 came as a bit of a shock though. The least powerful card in Nvidia’s new range boasts better performance that its previous flagship while also costing less than half. Today, a RTX 2080 Ti can still run just over $1000 (not far off what it launched at), while the RTX 3070 launches at just $500 in October.
The RTX 3080 then strikes a good balance between appreciable performance gains while also being reasonably priced compared to the previous generation. The RTX 2080 Super–Nvidia’s mid-generation refresh for the RTX 2080–launched at $700. While Nvidia says the RTX 3080 is twice as fast as the original RTX 2080, it’s priced at the same $700 as its Super variant. That makes it a much sounder purchasing decision, eliminating the RTX 2080 Super’s position in the market without a considerable price cut.
Pricing like this certainly aims at next-generation consoles, too, despite Microsoft and Sony still holding out on specifics for the Xbox Series X and PS5 respectively. While even the RTX 3070 alone might be in the region of a full next-gen console from a price perspective, Nvidia made a compelling case to consider PC. As Digital Foundry’s Alexander Battaglia picked up, Nvidia used Microsoft’s standard of reporting teraflops to paint a clear picture between the Xbox Series X and both its Turning and new Ampere architecture.
Nvidia just referenced Microsoft’s “equivalent teraflop” metric for full turing and full ampere: 34 TFs, and 58 TFs respectively. Those are quite interesting numbers in comparison to what we know about how it works on Xbox Series X (13 equiv teraflops)!
— Alexander Battaglia (@Dachsjaeger) September 1, 2020
While not naming cards specifically, Turing has an expected peak of 34 TFlops performance, while Ampere boasts 58 TFlops. Compared to the Xbox Series X’s 13 TFlops–a figure higher than the PS5 already–it’s a massive leap ahead on metrics alone. It must be stressed that even $500 is still a lot of money for a GPU, especially during this year where the COVID-19 pandemic has stressed many people financially. But within the context of next-generation consoles arriving later this year, it’s a comparison that Nvidia is sensibly making.
AMD, the manufacturer behind the RDNA 2-powered Xbox Series X and PS5, has yet to unveil its own consumer GPUs using its latest architecture. Considering what might be in the pipeline, Nvidia’s aggressive approach might be in anticipation of a performance and market-share fight that the company hasn’t had to pay too much attention too over the past couple of years.
All these figures and estimations need to be tested of course, which will happen as early as September 17 when Nvidia launches the RTX 3080. The monstrous RTX 3090 is expected this September too, while the RTX 3070 launches in October.
