According to renowned leaker @kopite7kimi, Nvidia's rumored GeForce RTX 40-series 'Super' graphics cards will see a performance boost, without an increased power consumption when compared to regular models.
Since the information is from an unofficial rumor we must take it with a pinch of salt, as it may well prove to be inaccurate.
Nvidia's RTX 40-series Super family of graphics cards is expected to consist of three models: Nvidia RTX 4080 Super based on anAD103 GPU with 10240 CUDA cores, Nvidia RTX 4070 Ti Super featuring an AD103 GPU with 8448 CUDA cores, and Nvidia RTX 4070 Super powered by an AD103 or AD104 processor with 7168 CUDA cores. If kopite7kimi's information is correct, then these boards will consume 320W, 285W, and 220W of power, respectively. As a result, only the RTX 4070 Super will gain a 20W power draw.
Alleged Nvidia RTX 40-Series Super Specifications
| Header Cell - Column 0 | GPU | FP32 CUDA Cores | Memory Configuration | L2 Cache | TBP | MSRP |
|---|---|---|---|---|---|---|
| *GeForce RTX 4090 Ti | AD102 | 18176 (?) | 24GB 384-bit 24 GT/s GDDR6X (?) | ? | 600W (?) | ? |
| GeForce RTX 4090 | AD102 | 16384 | 24GB 384-bit 21 GT/s GDDR6X | 72 MB | 450W | $1,599 |
| *GeForce RTX 4080 Super | AD103 | 10240 | 16GB 256-bit 22.4 GT/s GDDR6X | ? | 320W | ? |
| GeForce RTX 4080 | AD103 | 9728 | 16GB 256-bit 22.4 GT/s GDDR6X | 64 MB | 320W | $1,199 |
| *GeForce RTX 4070 Ti Super | AD103 | 8448 | 16GB 256-bit 22.4 GT/s GDDR6X | 48 MB | 285W | ? |
| GeForce RTX 4070 Ti | AD104 | 7680 | 12GB 192-bit 21 GT/s GDDR6X | 48 MB | 285W | $799 |
| *GeForce RTX 4070 Super | AD104 or AD103 | 7168 | ? | 48 MB | 220W | ? |
| GeForce RTX 4070 | AD104 | 5888 | 12GB 192-bit 21 GT/s GDDR6X | 36 MB | 200W | $599 |
| GeForce RTX 4060 Ti | AD106 | 4352 | 8GB or 16GB 128-bit 18 GT/s GDDR6 | 32 MB | 160W | $399/$499 |
| GeForce RTX 4060 | AD106 | 3072 | 8GB 128-bit 17 GT/s GDDR6 | 24 MB | 115W | $999 |
*Specifications are unconfirmed.
Considering that Nvidia's RTX 40-series 'Super' products will likely offer noticeable performance gains compared to non-Super offerings, maintaining power consumption of the latter will be a major achievement of the GPU designer. We could easily see a Super card making it to our list of the best graphics cards.
Improving performance while keeping power consumption low is a key goal of chip design. Chipmakers and chip designers tend to engage in continuous process improvements (CPI) to increase production yields and lower performance variations by using statistical process control (SPC). Nvidia's AD103 and AD104 graphics processors have been in production for well over a year now. Both Nvidia and TSMC have likely discovered multiple ways to enhance Nvidia's custom 4N process technology to boost production yields, increase performance, and lower power consumption. Furthermore, Nvidia has also learnt better GPU power management, making it possible that it has also managed to increase performance of the new products and maintain power consumption of its old ones.
Keep in mind that the information comes from an unofficial source, so take it with a grain of salt. After all, Nvidia has yet to confirm any GeForce RTX 40-series 'Super'-badged products coming.
