Earlier rumors suggested the RTX 4070 features a 10GB memory with a 160-bit configuration. However, it looks like NVIDIA may have some other plans.
Twitter user @kopite7kimi revealed new specs for NVIDIA’s GeForce RTX 4070.
NVIDIA’s RTX 4070 will now feature a 12 GB GDDR6 memory with a 192-bit interface.
The model will also have a faster memory speed of 21 Gbps, which along with a wider memory bus, will enable maximum bandwidth of up to 504 GB/s. This new speed is about 36.5% more than the previous specs.
The RTX 4070 will use the AD104 GPU die with 60 Streaming Multiprocessors and the PG141-SKU331 featuring 7680 CUDA cores.
This was previously the configuration for the RTX 4070 Ti variant.
Given the improved specs, the RTX 4070 is expected to score 11K points in the 3DMark TimeSpy Extreme benchmark, outperforming the current gen RTX 3090.
These specs are speculations; nothing can be confirmed until NVIDIA makes an official announcement. The RTX 40 series is expected to launch in early 2023.
Here are the old and updated specs for the RTX 4070:
|Architecture||Ada (TSMC N4)||Ada (TSMC N4)|
|Memory||10 GB G6||12 GB G6|
|Memory Speed||18 Gbps||21 Gbps|
|Bandwidth||360 GB/s||504 GB/s|
NVIDIA Corporation is an American multinational technology company incorporated in Delaware, based in Santa Clara, California. They design graphics processing units (GPUs) for the gaming and professional markets, as well as system on chip units (SoCs) for the mobile computing and automotive market.
Best known for the “GeForce” lines of GPUs, they are a direct competitor to AMD’s “Radeon” series. NVIDIA has also expanded its offerings with its handheld game consoles Shield Portable, Shield Tablet, and Shield Android TV and its cloud gaming service GeForce Now.