NVIDIA Ada Lovelace GPUs To use TSMC’s 4nm Process giving them an advantage over AMD’s RDNA3 GPUs

According to Moore’s Law is Dead, NVIDIA  Ada Lovelace GPUs for next-gen GeForce RTX 40 Gaming graphics cards will have a node advantage over AMD’s RDNA 3.

NVIDIA was expected to use TSMC’s 5nm production node for their Ada Lovelace GPUs, which will power the GeForce RTX 40 series of next-generation gaming graphics cards. Moore’s Law is Dead appears to have exposed the precise node in a recent tweet. The NVIDIA Ada Lovelace GPUs will be based on the TSMC 4N process node, according to the newest rumor.

The Hopper GPUs for the data center HPC industry is powered by the same TSMC 4N process node. The TSMC 4N process node is a refinement of their 5nm process (not to be confused with 4nm/N4, which is a separate node). Compared to the vanilla TSMC 5nm node, the TSMC 4N process node is custom-designed particularly for NVIDIA and features some enhancements that improve power efficiency, performance, and density.

NVIDIA going with TSMC’s 4nm is a wise choice 

The reasons for NVIDIA’s possible selection of TSMC’s 4N as a candidate for its next-generation gaming GPU range are self-evident. The next cards will be extremely power-hungry, and NVIDIA and the firm will optimize them as much as possible using the 4N manufacturing node.

For its future MCM and monolithic GPUs based on the RDNA 3 graphics architecture, AMD will use a mix of TSMC 5nm and 6nm production nodes, and while they won’t have the optimizations that 4N does, they will have an MCM approach that is projected to be highly efficient.

As a result, NVIDIA wins the better node, but AMD wins the superior design approach. End-users, on the other hand, will be unconcerned with these details if they merely want to play their games on the greatest hardware (graphics cards) available.

also read:

NVIDIA AD102 GPUs supporting the Next-Gen GeForce RTX 40 Series will stick with PCIe Gen 4.0 Protocol

source