China-Exclusive RTX 4090D Features Higher Base Clocks than the RTX 4090

NVIDIA plans to launch a cut-down RTX 4090, specific for the Chinese market, in light of the recent US sanctions. This cut-down RTX 4090 has been assigned the suffix ‘D‘, which apparently stands for Dragon. Anyhow, leaker MEGAsizeGPU asserts that the GPU is on track without any delays and actually features higher base clocks than the RTX 4090.

RTX 4090D Base and Boost Clocks Confirmed

For the unaware, it has been reported by various outlets that NVIDIA will launch a US-compliant version of the RTX 4090 for China. This RTX 4090, will of course have lower specifications than the base model. As such, NVIDIA has resorted to AD102-250, which is a toned-down version of AD102-300, used on the RTX 4090.

We are not sure of the CUDA core count, memory size, bus width, and expected performance yet. However, MEGAsizeGPU reports that the RTX 4090D has a base clock of 2280MHz which is higher than the RTX 4090. Likewise, its boost clocks match the RTX 4090, at 2520MHz.

While frequencies do play a key role in performance, they are not independent of other factors such as CUDA cores, memory size, and whatnot. NVIDIA will have to mix and match the configuration, such that it falls under 4800 TPP (Total Performance Power). For context, the RTX 4090 currently sits at 5285 TPP, courtesy of semianalysis.

An interesting point is that NVIDIA was initially rumored to circumvent the US sanctions, with an RTX 4080 Super. Both variants likely launch in China, since the RTX 4080 Super is most probably a bit slower than the RTX 4090D, though nothing is confirmed.

Release Date

The RTX 4090D will debut next year, probably during CES 2024, or after the event. Importantly, the SUPER refresh arrives at the same time, alongside other Hopper-based AI GPUs for China. Do you think the RTX 4090D can match the RTX 4090 with some overclocking? Tell us in the comments below.

SKUChipFP32/CUDASMsMax ClockCacheMemory BusVRAMMemory SpecSpeed (Gbps)TDP
RTX 4090AD102-300-A1163841282.52GHz96MB384-bit24GBGDDR6X21450W (TGP)
RTX 4090DAD102-2502.52GHz
RTX 4080AD103-300-A19728762.505GHz64MB256-bit16GBGDDR6X22.5320W(TGP)
RTX 4070 TiAD104-400-A17680602.61GHz48MB192-bit12GBGDDR6X21285W (TGP)
RTX 4070AD104-250/251-A15888462.475GHz36MB192-bit12GBGDDR6X21250W
RTX 4060 Ti 8GBAD106-350-A14352342.535GHz32MB128-bit8GBGDDR618160W
RTX 4060 Ti 16GBAD106-350-A14352342.535GHz32MB128-bit16GBGDDR618160W
RTX 4060AD107-400-A1307224~2.5GHz24MB128-bit8GBGDDR617115W

Source: MEGAsizeGPU


Abdullah Faisal

With a love for computers since the age of five, Abdullah has always sought to delve into the depths of information, and uses it as his guiding light. He believes success is of utmost importance as history is written by the victor.