The first review of NVIDIA's China-exclusive GeForce RTX 4090D GPU has been published & it is around the same performance as normal 4090.
NVIDIA GeForce RTX 4090 & RTX 4090D GPUs Offer Very Similar Performance Despite The Latter Being Cut-Down and TDP-Constrained
Expreview managed to get hold of the GALAX GeForce RTX 4090D Metal Master graphics card which is based on the new AD102-250 GPU silicon and is designed to comply with US policies since the original RTX 4090 had been restricted due to its TPP (Total Processing Performance) exceeding the 4800 points limit. As a result, NVIDIA had to cut back the specs of its flagship GPU and offer a new chip to China which we now know as the RTX 4090D.
The GALAX GeForce RTX 4090D Metal Master GPU features the reference specifications with 14,592 cores, 456 tensor cores, 114 RT cores, a base clock of 2280 MHz, a boost clock of 2520 MHz, 24 GB of GDDR6X memory running across a 384-bit bus interface at 21 Gbps for an effective bandwidth of 1008 GB/s. The graphics card has a TGP of 425W and is powered by the same 16-pin 12V-2x6 plug as the rest of the lineup. Despite being a custom model featuring a triple-fan cooling solution, the company hasn't shipped the card with factory overclocks as doing so can once again breach the 4800 TPP limit.
So with the RTX 4090 D, you are getting:
- Similar Gaming Performance As RTX 4090
- -11% Lower CUDA Cores & Tensor Cores
- Same Memory Configuration (24 GB G6X)
- Slightly Lower TGP (425W)
- No Founders Edition Model
- Only Custom Models
- No Overclocking Support
- Same Price As RTX 4090 (12,999 RMB / 1599 USD)
However, overclocking is possible but only within the given TGP limit of 425W. The card was able to overclock by 200 MHz and ended up delivering 2.3% higher performance in synthetic workloads so we can expect a 1-2% gain in games.
The tech outlet tested the NVIDIA GeForce RTX 4090D graphics card against the RTX 4090 in various games at 4K resolution and found it to be around 5.8% slower and with frame generation enabled, the difference falls to just 5%. The NVIDIA GeForce RTX 4090D was also compared against the 4090 & 4080 GPUs in AI and productivity benchmarks and once again ended up 6% slower.
- RTX 4090D vs RTX 4090 in Gaming (Rasterized): -5.5% slower
- RTX 4090D vs RTX 4090 in Gaming (Ray Tracing): -5.8% slower
- RTX 4090D vs RTX 4090 in Gaming (Frame-Gen): -5.0% slower
- RTX 4090D vs RTX 4090 in Productivity, AI, Content: -6.0-% slower
Following is the performance chart which shows the RTX 4090D easily exceeding the AMD Radeon RX 7900 XTX in rasterized performance:
These results are decent enough given the number of cores that have been cut down but it should be pointed out that the NVIDIA GeForce RTX 4090D GPU has a retail price of 12,999 RMB in China, the same as the standard RTX 4090 MSRP when it was available in the country. There will be certain cases where the cut-down specs end up with more reduced performance. The flip side is that the GPU consumes slightly lower power and also runs cool, hardly exceeding 70C (GPU temp) and 76C (Hotspot Temp). The peak power consumption of the card reached 461W.
With the GeForce RTX 4090D, NVIDIA once again has the flagship gaming graphics card in China. The graphics card offers the best AI performance of all the consumer-tier graphics cards available on the market. AMD's Radeon RX 7900 XTX and XT models did start to sell well after the ban on the RTX 4090 but given the strong AI software and support for NVIDIA's GPUs, they are once again expected to regain their popularity in the enthusiast segment.
NVIDIA GeForce RTX 40 Desktop GPU Family:
| Graphics Card | GPU SKU | GPU Cores | TMUs / ROPs | Clock Speed (Boost) | VRAM / Bus | Bandwidth | TGP | Price (MSRP) | Launch |
|---|---|---|---|---|---|---|---|---|---|
| GeForce RTX 4090 | AD102-300 | 16384 | 512 / 176 | 2520 MHz | 24 GB GDDR6X / 384-bit | 1008 GB/s | 450W | $1599 US | October 2022 |
| GeForce RTX 4090D | AD102-250 | 14592 | 456 / 176 | 2520 MHz | 24 GB GDDR6X / 384-bit | 1008 GB/s | 425W | 12999 RMB | January 2024 |
| GeForce RTX 4080 SUPER | AD103-400 | 10240 | 320 / 112 | 2550 MHz | 16 GB GDDR6X / 256-bit | 736 GB/s | 320W | $999 US | January 2024 |
| GeForce RTX 4080 | AD103-300 | 9728 | 320 / 112 | 2510 MHz | 16 GB GDDR6X / 256-bit | 716 GB/s | 320W | $1199 US | November 2022 |
| GeForce RTX 4070 Ti SUPER | AD103-275/AD102-175 | 8448 | 264 / 96 | 2610 MHz | 16 GB GDDR6X / 256-bit | 672 GB/s | 285W | $799 US | January 2024 |
| GeForce RTX 4070 Ti | AD104-400 | 7680 | 240 / 80 | 2610 MHz | 12 GB GDDR6X /192-bit | 504 GB/s | 285W | $799 US | January 2023 |
| GeForce RTX 4070 SUPER | AD104-350/AD103-175 | 7168 | 224 / 80 | 2475 MHz | 12 GB GDDR6X / 192-bit | 504 GB/s | 220W | $599 US | January 2024 |
| GeForce RTX 4070 | AD104-250 | 5888 | 184 / 64 | 2475 MHz | 12 GB GDDR6X / 192-bit | 504 GB/s | 200W | $549 US | April 2023 |
| GeForce RTX 4060 Ti 16 GB | AD106-350 | 4352 | 136 / 48 | 2535 MHz | 16 GB GDDR6 /128-bit | 288 GB/s | 165W | $449 US | July 2023 |
| GeForce RTX 4060 Ti 8 GB | AD106-350 | 4352 | 136 / 48 | 2535 MHz | 8 GB GDDR6 / 128-bit | 288 GB/s | 160W | $399 US | May 2023 |
| GeForce RTX 4060 | AD107-400 | 3072 | 96 / 48 | 2460 MHz | 8 GB GDDR6 / 128-bit | 272 GB/s | 115W | $299 US | June 2023 |
News Source: Expreview
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
