TECH NEWS – While Nvidia has reduced power consumption on its 4000 series cards (especially the lower-priced ones) compared to the GeForce RTX 3000 series, AMD has done the opposite…
Since we wrote about Nvidia’s GeForce RTX 4060 for the mainstream the other day, it’s a good time to look at AMD’s offering. Among the entry-level RDNA 3 GPUs is the RX 7600, using Navi 33, which we’ve known the specs of for a long time, but Videocardz has learned about an element we haven’t heard much about until now.
The Navi 33 XL GPU could be on the card with 13.3 billion transistors on a 204 square millimeter die, 14% smaller than the Navi 23 GPU used in the Radeon RX 6600 cards. There are 16 WGPs, 32 CUs, 32 ray tracing units, 64 artificial intelligence accelerators, and 2048 stream processors on the RX 7600. In games, it can clock 2250 MHz, which becomes 2625 MHz in Boost mode. The 128-bit bus and 32 MB Infinity cache remain unchanged, but the new card is built with 6-nanometer technology instead of 7, capable of 21.75 TFLOPs of clocking performance in FP32 and 43.5 TFLOPs in FP16.
The AMD Radeon RX 7600 has 8 GB of GDDR6 VRAM and 18 Gbps on the 128-bit bus, giving a bandwidth of 288 GB/s (476.9 effective) and 64 ROPs on the GPU. The Navi 33 GPUs run on PCIe 4.0 x8 interface, as the company has made it entry-level. The problem is that the TBP is 165W through the 8-pin connector, and a 550W power supply is recommended. It is much higher than the consumption of the GeForce RTX 4060 (115W), which is intended to be the direct rival, and is more in competition with the RTX 4060 Ti, even though it is a more powerful card ( in games, the greens’ card consumed 140W on average).
So, the RX 7600 consumes around 165W, which is even higher than its predecessor, as the RX 6600 was around 132W. AMD’s processors have good power consumption (especially compared to Intel!), but Nvidia is a step ahead!