ASUS ROG Strix NVIDIA GeForce RTX 3070 Gaming
Excellent aesthetic design
Great cooling solution
High boost clock speed
On the higher end of the price spectrum
Battle of the mid range
WePC is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more
Taking a look at the mid-range selection we have the 6700XT vs 3070 battle. A likely good compromise of price and performance, fighting to be the best graphics card. With plenty of pros and cons for each of the choices.
The AMD vs Nvidia battle is a close one, so either of the two is a possible option. We will take a look at what separates team red and team green in this selection. Especially at the lower price points, we may hope for at this level.
Excellent aesthetic design
Great cooling solution
High boost clock speed
On the higher end of the price spectrum
Great cooling solution
Best performing option
More expensive option
Firstly we take a look at the heart of the card with what the GPU die and what it is made of. Being from two different manufacturers the architectural differences stem right from the heart and work differently.
AMD’s cards are based on RDNA 2.0 architecture, with the Nvidia cards based on the Ampere architecture. More specifically these GPUs are built with Navi 22 XT and GA104 300 dies respectively.
The Navi chip uses a 7nm TSMC process to create a 335mm2 die size with 17.2 billion transistors. Compared to the 8nm Samsung die in the GA chip in a 392mm2 size with 17.4 billion transistors.
RX 6700 XT | RTX 3070 | |
---|---|---|
GPU | Navi 22 XT(215-127000006) | GA104-300-A1 |
Shaders | 2560 | 5888 |
RT Cores | 40 | 46 |
TMUs | 160 | 184 |
Memory | 12GB GDDR6 | 8GB GDDR6 |
Base clock | 2321 MHz | 1500 MHz |
Boost clock | 2581 MHz | 1725 MHz |
Memory clock | 16 Gbps | 14 Gbps |
Card bus | PCIe 4.0 x16 | PCIe 4.0 x16 |
Memory bandwidth | 384.0 GB/s | 448.0 GB/s |
TDP | 230W | 220W |
Not technically a one-to-one comparison with the cards being built differently but the specs still give us a good basis. There are a few options we will take a closer look at to see what they mean and how they differ.
The video memory of a card is used as a frame buffer and a store for texture maps. It’s more important in higher resolution applications. Since they require more pixels, they take up more space in those buffers.
There is a trade-off between speed and capacity with both sides taking different approaches. AMD takes the capacity route with 12GB of VRAM but with a lower throughput of 384.0GB/s. Compared to the 8GB found in Nvidia’s card, but running at 448.0 GB/s. The difference may have a different impact on performance
The TDP of the card informs us of how much energy the card will consume at max power. This also corresponds to the heat produced by the GPU. Between the 6700XT vs 3070, the AMD card just about has a higher rating. It is 10W higher than the 3070 and so will have a higher consumption and heating effect.
Clock speed determines how quickly the GPU processes instructions and spews out those graphics. Based on how the cycle rate of the silicon crystals can determine the performance of the card. But also has the opportunity for overclocking with good cooling.
Between the two cards, there is a significant difference in the factory clock. AMD manages a base clock of 2,321MHz with a boost of 2,581MHz. Compared to just a 1,500MHz base and 1,725MHz boost of the Nvidia GPU. Giving AMD a large advantage in processing speed.
Cores are the sections that make up the die sections, They each have a specialization that allows them to focus on just one task. With the different setups between the two, they are likely to behave and set up differently.
Shaders are focused on the shadows and shading of the scene and graphics. Nvidia takes advantage of numbers on these cores. It features 5,888 units compared to just 2,560 of the 6700XT.
There are also the RT cores, these are the ray tracing focused units. Here the 3070 just about also has the advantage, with 46 compared to 40. And the same case is in TMUs, the texture mapping units that focus on 3d mapping from bitmaps. It sports 184 compared to AMD’s 160.
Taking a look at 3dmarks benchmarks they show pure graphics performance of the GPUs. In the scenario of the 6700XT vs 3070, it’s the RTX card that takes the slight lead. It achieves a score of 13,679 compared to the RXs of 12,604. Showing a significant advantage even with the slower clock.
We look to TechSpot’s comparison for a good look at the performance of the cards. Its benchmarks show the RX 6700 XT falls behind on average performance. With the RTX 3070 manages a 4-8% higher framerate than the 6700 XT. It will also especially lead in raytracing scenarios with its better optimizations and cores.
It’s at MSRP that the AMD card tries to claw back its disadvantage. The launch price for the card was $479, whilst the Nvidia card launched for $499. So there is only $20 in it and so the performance benefit is likely worth that extra cost.
Although, at the current time, GPUs are still expensive in the current market. And even with the prices still falling it is hard to find them at MSRP. So at the moment, the cheapest 6700XT starts at $800, with the 3070 at $1195. A rather significant difference between their MSRP and between each other.
Smaller in size than some optioons
Cheaper yet still powerful
More basic design
Fantastic build quality
EVGA iCX3 Cooling
Adjustable ARGB LED
Design can be a little basic for some
Both of these options are excellent possibilities. You should, however, choose the GPU that is the closest to MSRP that you can find, given current GPU costs. You should select AMD over Nvidia if you can get an RX 6700 XT that is significantly less expensive than the RTX 3070.
Overall, looking at the 6700XT vs 3070 battle it’d be hard not to recommend the Nvidia card. With such a bigger performance at only a slight extra cost on MSRP, it is an easy choice. You can always wait for the next generation of RTX 4000 series or RDNA 3 cards for more choice and performance, or even Intel ARC Alchemist at this performance level.
As standard in the current generation, it’s in fact the 6700 XT that’s faster than the 3070. It has a 55% faster base clock, with a 50% faster boost clock. Even in terms of memory clock, it has a 14% lead, but it has a slower bandwidth by 17%.
From the 3DMarks benchmarks, we see the graphics performance the two cards achieve. The 3070 takes the lead with 13,679 compared to the 6700 XTs 12,604. This means the 3070 is better by 9%. Whilst in gaming it also leads with an average of 4-8% more FPS.