NVIDIA GeForce RTX 4070 vs NVIDIA GeForce GTX 660 Ti

The NVIDIA GeForce RTX 4070 is a desktop graphics card that launched in Q2 2023 with a MSRP of $599. It is built on the Ada Lovelace GPU microarchitecture (codename AD104) and is manufactured on a 5 nm process.

Memory

The RTX 4070 has 12 GB of GDDR6X memory, with a 1,313 MHz memory clock and a 192 bit interface. This gives it a memory bandwidth of 504.2 Gb/s, which affects how fast it can transfer data to and from memory. GPU memory stores temporary data that helps the GPU with complex math and graphics operations. More memory is generally better, as not having enough can cause performance bottlenecks.

Cores and Clock Speeds

The RTX 4070 includes 5,888 CUDA cores, the processing units for handling parallel computing tasks. The GPU operates at a core clock speed of 1,920 MHz and can dynamically boost its clock speed up to 2,475 MHz. Complementing the processing units are 184 texture mapping units (TMUs) for efficient texture filtering and 64 render output units (ROPs) for pixel processing. Additionally, the GPU features 184 tensor cores optimized for AI-accelerated workloads and 46 RT cores dedicated to real-time ray tracing calculations.

Compatibility & Power Consumption

The RTX 4070 occupies 2 PCIe expansion slots. It supports HDMI 2.1, DisplayPort 1.4a display connections. NVIDIA recommends a power supply of at least 550 W to handle the GPU's thermal design power (TDP) of 200 W.

Benchmark Performance & Value

The RTX 4070 has the 17th best 3DMark 11 Performance GPU score among the 543 benchmarked GPUs in our database. It achieves 57.33% of the performance of the best benchmarked GPU, the NVIDIA GeForce RTX 4090. Its 59,969 score and $599 launch price (MSRP) gives it a performance per dollar of 100.12. This is the 15th best in value for the 3DMark 11 Performance GPU benchmark.