Nvidia’s Super refresh line of RTX 40-series graphics cards continues with the RTX 4070 Ti Super, the inelegantly named product designed to replace non-Super cards and offer a host of upgrades: more memory bandwidth, more CUDA cores, and 16GB of frame buffer memory – Same money. On paper, the specs look good, but in practice, Nvidia needs to strike a balance here: we can’t expect this card to rival the existing RTX 4080, as the specs suggest the 4080 Super won’t be far behind. better one. If we push too hard on the 4070 Ti Super, there will be no market for the next generation Super.
The specs do look good, though. The RTX 4070 Ti Super is based on the same AD103 chip as the RTX 4080, which naturally gives it an advantage over the RTX 4070 Ti non-Super, which uses the lower-performance AD104. Ti Super has 8488 CUDA cores, 10% more than its predecessor and 87% of the 4080 cores. At the same time, the new card has a 100MHz boost clock advantage over the RTX 4080.
The AD103 chip also means the 70 Series range has a 256-bit memory bus, up from the 192-bit interface of its non-Super predecessors, and an impressive 33% increase in bandwidth. And there’s more – the AD103’s 256-bit interface not only supports 16GB of memory, it also means users get improved media blocking via dual video encoders. The AD104 only gets one encoder on the non-Super. I actually use the RTX 4080 in my workstation for Adobe Premiere work and a little gaming: the RTX 4070 Ti Super also works fine.
| 4080 super | 4080 | Chapter 4070 Awesome | Chapter 4070 | 4070 super | 4070 | |
|---|---|---|---|---|---|---|
| processor | AD103 | AD103 | AD103 | AD104 | AD104 | AD104 |
| CUDA color | 10240 | 9728 | 8448 | 7680 | 7168 | 5888 |
| boost clock | 2.56GHz | 2.51GHz | 2.61GHz | 2.61GHz | 2.48GHz | 2.48GHz |
| memory allocation | 16 GB | 16 GB | 16 GB | 12GB | 12GB | 12GB |
| memory interface | 256 bit | 256 bit | 256 bit | 192 bits | 192 bits | 192 bits |
| memory bandwidth | 736GB/sec | 717GB/sec | 672GB/sec | 504GB/sec | 504GB/sec | 504GB/sec |
| Three Gorges Project | 320W | 320W | 285W | 285W | 220W | 200W |
| US suggested retail price | $999 | $1199 | $799 | $799 | $599 | $599 |
| UK recommended retail price | £969 | £1199 | £769 | £799 | £579 | £589 |
| release date | 31/1/24 | go out | 24/1/24 | go out | 17/1/24 | go out |
For today’s review, we No There is a Founder’s Edition card, because there isn’t one. Similar to the RTX 4070 Ti, Nvidia leaves it to third-party motherboard manufacturers to produce their own models. The European press received the MSI Ventus 3x model (which at launch did not meet reference specs and required a VBIOS update), and I also received the ASUS TUF Gaming version. Both are big-name, relatively cheap-looking models that lack the aesthetic sophistication of the Founders models.
However, their enormous size and volume (enough to be registered on DRADIS) at least ensure very quiet operation and good temperature management. The 12VHPWR socket is back on both models for power, while the MSI models usually have HDMI 2.1 and DisplayPort 1.4a video outputs on the back, while the Asus (helpfully!) gets an extra HDMI port.
The tests in this review were conducted using Asus’ TUF Gaming version of the RTX 4070 Ti Super, as it’s unclear whether the VBIOS update for the MSI model will work properly until that version arrives. The ASUS and MSI models have an RRP of $799/£769, and both run to Nvidia’s reference specs.
There are no surprises with power analysis. We use Nvidia’s PCAT interposer hardware that sits between the graphics card and its power supply (PCIe slot and auxiliary input). We measure average power consumption and frame rate. When you divide watts by frames per second, you get joules per frame – i.e. lower is better.
I tested here using an Asus motherboard – but I did run MSI Ventus 3x numbers beforehand (before the VBIOS update), and the TUF gaming card seemed to draw more power – slightly above its 285W TGP spec limit. Even so, it’s nothing to write home about just yet, since we all know that Ada Lovelace cards are very efficient. The gains are most obvious in ray tracing applications, where I found the 4070 Ti Super to have a clear advantage over the non-Super and leave the RX 7900 XT (AMD’s closest competitor) well behind.
In Forza Horizon 5, the gap is even wider, with Nvidia tending to have the performance (and therefore efficiency) advantage, while the Ti Super consumes only a little more power than the non-Ti version.Meanwhile, in Hitman 3 with RT disabled, AMD’s performance advantage comes to the fore, so you can see that in the best case, AMD can actually more More efficient than NVIDIA.
| RX7900XT | RTX 4070 Titanium | RTX 4070 Super | |
|---|---|---|---|
| Control, 4K, high RT | 303.2W/31.0fps – 9.8 joules per frame | 277.3W/34.3fps – 8.1 joules per frame | 298.7W/38.7fps – 7.7 joules per frame |
| Forza Horizon 5, 4K, Extreme, RT Off, 4x MSAA | 309.1W/96.1fps – 3.2 Joules per frame | 219.1W/101.4fps – 2.2 Joules per frame | 245.8W/108.0fps – 2.3 Joules per frame |
| Hitman 3, 4K, Highest, RT Closed | 312.1W/165.4fps – 1.9 joules per frame | 280.2W/132.3fps – 2.1 Joules per frame | 298.5W/136.4fps – 2.2 Joules per frame |
With the preliminaries out of the way, it’s time to get into the real benchmarking. As always, we used high-end gear, pairing a Core i9 13900K with 6000MT/s G.Skill DDR5 and an Asus ROG Maximus Hero Z690 motherboard to minimize CPU bottlenecks and thus show the GPU differences more clearly.
Nvidia GeForce RTX 4070 Super Analysis
- introduce [This Page]
- RT Benchmark: Dying Light 2, Cyberpunk 2077, Control
- RT Benchmark: Metro Exodus Enhanced Edition, F1 22
- RT/DLSS/FSR2/DLSS3 Benchmark: “Cyberpunk 2077”, “Dying Light 2”, “Forza Horizon 5”
- Game Benchmarks: Control, Cyberpunk 2077, F1 22, Forza Horizon 5
- Game benchmarks: “Hitman 3”, “A Plague Tale: Requiem”, “Return”
- Conclusion and recommendations
iGamesNews
