ASUS TUF GeForce RTX 4070 Ti SUPER Review: More Cores, Memory, Performance

	ASUS TUF GeForce RTX 4070 Ti Super: MSRP $799 More cores, more memory, and higher clocks make the new ASUS TUF GeForce RTX 4070 Ti SUPER one heck of a performer in its price category.
	More Cores 16GB Video Memory Excellent Performance Relatively Cool & Quiet Highly Efficienty	Oversized For No Reason Only DP 1.4

---

NVIDIA is launching its second salvo of GeForce RTX 40 SUPER series GPUs today, with the middle-child in the line-up, the GeForce RTX 4070 Ti SUPER. Unlike its higher and lower end siblings, the GeForce RTX 4080 SUPER and RTX 4070 SUPER, however, the RTX 4070 Ti SUPER won’t be offered in Founders Edition trim. Instead, NVIDIA is relying on its board partners to bring the GeForce RTX 4070 Ti SUPER to market, which means there will be quite a bit of variation from one manufacturer to the next, at least in terms of features and pricing. Performance of all GeForce RTX 4070 Ti SUPER cards will be roughly similar, give or take a few percentage points depending on peak boost clocks and power levels, and the robustness of the card’s cooler.

We have what will likely be one of the more popular GeForce RTX 4070 Ti SUPER cards to show you today, an ASUS TUF branded model that ticks virtually all of the right boxes. The ASUS TUF GeForce RTX 4070 Ti SUPER is a beefy, triple-slot, triple-fan card that offers strong performance, while remaining relatively cool and quiet. Let’s dive in and check it out...

NVIDIA GeForce RTX 4070 Ti Super Specifications

---

As we’ve mentioned in multiple articles in the past, all GeForce RTX 40 Series cards – including these latest SUPERs -- are based on the same Ada GPU architecture, but the cards are not built around the same chips. The current flagship RTX 4090 is built around the largest and most powerful AD102 GPU. The GeForce RTX 4080 is powered by the scaled-down AD103. GeForce RTX 4070 cards use the smaller AD104. And GeForce RTX 4060 series cards use the AD106, which is smaller still.

The new GeForce RTX 4070 Ti SUPER is still based on the AD104, but the chips has a different configuration than the original Ti.

In terms of feature support, the AD104 is identical to the larger AD102 and AD103. It offers the entirety of features inherent to the Ada Lovelace architecture, like DLSS 3, and other NVIDIA technologies like NVIDIA Reflex latency reduction, NVIDIA Broadcast, Ansel, G-SYNC, Omniverse, optional Studio drivers, AV1 encoding, and GeForce Experience, to name just a few. For a more comprehensive explanation of Ada Lovelace, we suggest reading our launch coverage of the GeForce RTX 4090 and Ada Architecture Overview, which goes into much more detail than we will here.


In comparison to its larger siblings, the AD104 on the RTX 4070 Ti SUPER is scaled down in a few ways.

The AD102 on the RTX 4090 has 11 GPCs, 64 TPCs, 16,384 CUDA cores, 512 Tensor cores, and 128 RT cores, with a 384-bit memory interface. The smaller AD103 on the RTX 4080 has 7 GPCs, 38 TPCs, 9,728 CUDA cores, 304 Tensor cores, and 76 RT cores, with a 256-bit memory interface. As configured on the GeForce RTX 4070 Ti SUPER, the AD104 has 6 GPCs, 33 TPCs, 8,448 CUDA cores, 264 Tensor cores, and 66 RT cores, with a 192-bit memory interface. The AD104 is effectively scaled down in every way that affects compute, ray tracing and rasterization performance versus higher-end card. The RTX 4070 Ti SUPER, however, is more powerful than the original 4070 Ti, which is equipped with 5 GPCs, 30 TPCs, 7,680 CUDA cores, 240 Tensor cores, and 60 RT cores. Both cards retain similar 48MB L2 cache configurations, but the memory interface on the SUPER is increased to 256-bits, which necessitates that addition of a couple of memory chips. As such, the total size of video memory is increased to 16GB, from 12GB on the original RTX 4070 Ti. Peak bandwidth is also increased, up to 672GB/s.

Meet The ASUS TUF GeForce RTX 4070 Ti SUPER

The ASUS TUF GeForce RTX 4070 Ti SUPER is a rather beefy graphics card.
The ASUS TUF GeForce RTX 4070 Ti SUPER is over a 11.5” long, over 5 inches high, and 2.7 slots wide. The cooler features triple axial fan, situated atop an array of heatsinks, linked together via heat-pipes. As you’ll see in some of the images below, though the heatsinks are ultimately wider than the entire PCB, there are some gaps in the assembly and they don’t completely cover the front side of the board.

In terms of its specifications, the card has 16GB of GDDR6X RAM clocked at an effective data rate of 21Gbps per NVIDIA’s reference specs. This ASUS TUF card has goosed up GPU boost clocks, however. NVIDIA’s specs call for a 2,610MHz GPU boost clock (with a 2,340MHz base clock), but we routinely saw this card boosting to nearly 2,800MHz in its stock configuration. And, of course, higher-clocks are possible with overclocking.
There is a dual-BIOS switch on the ASUS TUF GeForce RTX 4070 Ti SUPER, which gives users the ability to switch between normal / performance mode and a quiet fan mode. This card is effectively silent at idle though, because its fans spin down completely when the GPU isn’t being stressed. And under full load, even when overclocked, GPU temps remained under 70°C, and fans speeds only ramped up to about 45%, which didn’t produce much noise at all.

There is a metal heat-plate on the backside of the card, with cut-outs to allow for pass-through cooling, along with an array of graphics and branding. The front fan shroud has numerous angles and cut-outs as well. The card also features RGB lighting, which can be managed with ASUS GPU Tweak III utility.
In terms of connectors, the ASUS TUF GeForce RTX 4070 Ti SUPER feature a single 12VHPWR power connector (adapter dongle included), and five display outputs – 3 full-sized DisplayPorts (DP 1.4) and dual full-sized HDMI ports (2.1a).

With its huge cooler, lighting, and aggressive design, the ASUS TUF GeForce RTX 4070 Ti SUPER is obviously a GPU targeted at gamers. And now it’s time to find out how it performs...