_rogame has shared what is reportedly the first public benchmark for NVIDIA’s GeForce RTX 3080 graphics card. The benchmark suggests that the GPU will flaunt a 2.1 GHz max core clock and 10 GB of memory (4750 MHz, 19 Gbps), which is 2 GB less than what the flagship RTX 3090 will have (according to Micron, at least). That core clock is also similar to what overclocked Turing cards (e.g., the RTX 2080 Ti) can manage.
hmmm— _rogame (@_rogame) August 16, 2020
So this RTX 3080 core clock limit is similar to Turing GPUs
A mysterious INF file for an “NVIDIA Graphics Device” has also been spotted. This is reportedly the GeForce RTX 3080.
- NVIDIA_DEV.1FD1 = “NVIDIA Secondary N18P-G0-MP-A1 (TU117-750-A1)”
- NVIDIA_DEV.1FD5 = “NVIDIA Secondary N18P-G62-A1 (TU117-775-A1)”
- NVIDIA_DEV.2206 = “NVIDIA Graphics Device”
- NVIDIA Isn’t Getting Rid of the GeForce RTX 3070’s 12-Pin Power Connector
- GIGABYTE Unlocks AMD Smart Access Memory on 500 Series Motherboards, Improving Radeon Gaming Performance by Up to 17 Percent
- ASRock Announces Taichi Razer Edition AMD X570 and B550 Motherboards
- Some PlayStation 5 Consoles Have a Different, Louder Cooling Fan
- Leaked AMD Radeon RX 6700 BIOSes Reveal Maximum GPU Clocks of 2.85/2.95 GHz