Nvidia partners are not interested in RTX 4060 Ti 16GB

In May of this year, NVIDIA introduced a fresh generation of mainstream GPUs based on the Ada Lovelace architecture, inclusive of the GeForce RTX 4060, RTX 4060 Ti 8GB, and RTX 4060 Ti 16GB models. The Manufacturer’s Suggested Retail Price (MSRP) for these products is, correspondingly, $299, $399, and $499. The GeForce RTX 4060 and RTX 4060 Ti 8GB have already entered the market, with rumours suggesting the RTX 4060 Ti 16GB is slated for a July 18th release.

According to a report by HardwareLuxx.de, NVIDIA’s partners have privately intimated their lack of enthusiasm for the GeForce RTX 4060 Ti 16GB, expressing limited interest in customising designs. The underlying reason is its price point, set at $499.

Comparatively, the RTX 4060 Ti 16GB costs an additional $100 over the GeForce RTX 4060 Ti 8GB, and apart from doubling the memory capacity, there is no discernible difference between the two. According to standard business practices, manufacturers might release overclocked premium versions at prices significantly above the MSRP. In the end, the price difference may not be significant when compared to the higher-end GeForce RTX 4070. Our ladder data indicates that the performance of the RTX 4070 surpasses the RTX 4060 Ti 8GB by more than 25%. Even with larger memory, it would be theoretically difficult to reduce the performance gap significantly.

The GeForce RTX 4060 Ti 16GB is equipped with the AD106-351 GPU, adopting a PCIe 4.0 x8 interface. The complete chip version possesses 36 Streaming Multiprocessors (SM), 4608 CUDA cores, 36 RT cores, and 144 Tensor cores. However, the RTX 4060 Ti has only 34 enabled SM groups, or 4352 CUDA cores, 34 RT cores, and 136 Tensor cores. Furthermore, it features an 8th-generation NVENC encoder and a 5th-generation NVDEC decoder, supporting AV1 hardware encoding and decoding.

The new model has a base clock of 2310 MHz, a boost clock of 2535 MHz, is paired with 16GB of GDDR6 memory, has a memory bus width of 128-bit, a memory speed of 18Gbps, and a memory bandwidth of 288 GB/s. The entire card’s power consumption is 165W, an increase of 5W compared to the 8GB version. NVIDIA has stated that through the 32MB of L2 cache, the effective memory bandwidth increases to 554 GB/s.