Introduction

ASUS GeForce RTX 4070 Dual is the company's most affordable custom-design RTX 4070 offering. The DUAL brand of ASUS has a lot in common with the company's PRIME brand of motherboards—they are both under the company's main marquee, unlike the ROG Strix or TUF Gaming lines; and ASUS ensures they directly represent the company well—no corners are cut with product design or quality, but all the essentials are covered. The GeForce RTX 4070 debuting today brings the new "Ada" graphics architecture to an even wider audience in the performance segment, with a starting price of $600, which is exactly what the ASUS RTX 4070 DUAL from this review is priced at. NVIDIA ensured that every board partner has products at this price, and all the RTX 4070 reviews you'll see today (April 12) can be had at MSRP. We'll review premium, factory-overclocked variants, including those from ASUS, tomorrow (April 13). Both kinds of cards release to the market on April 13.



The GeForce RTX 4070 is intended for maxed out 1440p gaming, including real time ray tracing; as well as high refresh-rate e-sports gaming at 1440p and 1080p. Gaming at 4K Ultra HD is very much possible, if you can dial down some settings, or take advantage of DLSS, or the new force-multiplier that is DLSS 3. The GeForce "Ada" graphics architecture debuts the third generation of NVIDIA's groundbreaking RTX real time ray tracing technology, and the fourth generation of its on-die AI acceleration in the form of Tensor cores. These GPUs are built on the new TSMC 5 nm process, and NVIDIA has refined its architecture at various levels, to bring down graphics card PCB complexity and power draw. Take for instance, the memory sub-system, which has been generationally narrowed, but uses faster GDDR6X memory, and is accelerated by larger on-die caches on the GPU.

The RTX 4070 is based on the same "AD104" silicon that the RTX 4070 Ti maxes out, but is heavily cut down. It features 5,888 CUDA cores, 46 RT cores, 184 Tensor cores, 64 ROPs, and 184 TMUs. The memory setup is unchanged from the RTX 4070 Ti—you get 12 GB of 21 Gbps GDDR6X memory across a 192-bit wide memory bus, yielding 504 GB/s of memory bandwidth, which is still higher than the 448 GB/s the previous-generation RTX 3070 manages with its 256-bit memory bus. The core-configuration, particularly the shader count, is identical to that of the RTX 3070, but over a brand new architecture that promises generational performance uplifts.

The ASUS GeForce RTX 4070 Dual draws its name from the dual-fan custom-design cooling solution. Its design is minimalist, with a focus on low noise. If all you care about is the RTX 4070 at MSRP, and you don't need too many "custom-design" bells and whistles, this is the product ASUS wants you to buy. Its cooler features a pair of Axial-Tech fans with webbed impellers that guide all their airflow axially. Perhaps the biggest design choice ASUS made in favor of end-users, has to be the power connector. This card uses a single 8-pin PCIe power connector, sparing you of the need for a power-adapter to convert PCIe power connectors to 12VHPWR.