Nvidia’s next batch of mid-range graphics cards are imminent and the 5070 and 5070 Ti’s billed specifications promise to deliver incredible price to performance.
Though the 5080 and 5090 will likely draw much of the attention, and deservedly so, but at $999 and $1,999 respectively, they will inevitably be out of reach for many.
That’s where these two cards come into play, offering a more affordable entry point into the 50 series, while still delivering impressive output to players and creatives alike. The widely reported flaws in the Blackwell architecture at the end of last year have now been fixed and the 2.0 version looks like it’s in good shape to get the most out of these new cards for the next year.
After an announcement from Nvidia at CES we now have all of the details, so we’ve broken down everything you need to know if you’re in the market for an affordable graphics card.
Release window
The exact release date for the 5070 and 5070 Ti has not been confirmed by Nvidia, but they are set to debut at some point in February of this year. The Founders editions of each card will be available to purchase directly from the Nvidia website, while Stock-clocked and factory-overclocked models will also be available from the company’s usual partners.
How much will they cost?
Specifically designed to deliver significant bang for the buck, both cards are considerably cheaper than their more powerful counterparts in the series. The 5070 Ti and 5070 will be available to pick up for $749 and $549 respectively, with both cards able to run even demanding games on reasonable settings.
The company are so confident in these two products that CEO Jen-Hsun Huang described the 5070 as offering “RTX 4090 performance at $549.” Though much of that may come down to some DLSS 4 magic, it’s a bold claim for such an affordable pair of cards.
Subscribe to our newsletter for the latest updates on Esports, Gaming and more.
RTX 5070 & 5070 Ti specifications
RTX 5070
Despite the 5070 being set to be the least powerful card in the series at launch, it still has impressive specs under the hood. Though we don’t have a complete breakdown of the cards capabilities, there’s plenty to get excited about. Here’s what’s on offer:
| Card Feature | Specs |
|---|---|
| GPU Chip | GB205-300-A1 |
| CUDA Cores | 6144 |
| Interface | 16x PCIe 5.0 |
| Memory | 12GB GDDR7 |
| Max Memory Bus | 192-bit |
| Memory Clock | 28 Gbps |
| Bandwidth | 672GB/s |
| Wattage | 250W |
RTX 5070 Ti
The 5070 Ti also has some beefy specs and it offers huge improvements over its equivalent from the last generation in the 4070 Ti. The improved support for DLSS 4 should also appeal, offering more significant future proofing for those willing to make the increased investment.
| Card Feature | Specs |
|---|---|
| GPU Chip | GB203-200-A1 |
| CUDA Cores | 8960 |
| Interface | 16x PCIe 5.0 |
| Memory | 16GB GDDR7 |
| Max Memory Bus | 256-bit |
| Memory Clock | 28 Gbps |
| Bandwidth | 896.3GB/s |
| Wattage | 300W |
RTX 5070 & 5070 Ti features
In a major improvement over the Ada Lovelace cards, both the 5070 & 5070 Ti (as well as the more premium cards in the 50 series) feature support for DisplayPort 2.1a and HDMI 2.1b.
AI performance is also clearly a priority once again, despite the fact that neither of these cards would count among Nvidia’s premium offerings. The RTX 5070 Ti has 1,406 TOPS while the RTX 5070 has 988 TOPS, both of which equal or beat the SUPER TOPS series from the previous gen, which was specifically geared towards AI performance.
Overall, these new cards should prove very tempting to those with demanding workloads, as well as gaming enthusiasts and creative PC users. At the listed price point and performance standard, it’s likely going to be a bit of a kick in the teeth for those who went for the 40 series, but those who are in the market now should find both cards to be impressive options in a new rig.
