NVIDIA GeForce RTX 2050
Specification
- Manufacturer
- NVIDIA
- Architecture
- Ampere
- GPU chip
- GA107
- Chip variant
- GN20-S7
- Foundry
- Samsung
- Process node
- 8 nm
- Transistors (bn)
- 8.7 mld
- Die size
- 200 mm²
- Announcement date
- December 17, 2021
- Series
- RTX 20 Mobile
- Short name
- RTX 2050
- TGP
- 30–45 W
- Dynamic Boost 2.0
- No data
- CUDA cores
- 2048
- TMUs
- 64
- ROPs
- 32
- Ray tracing cores
- 16
- Tensor cores
- 64
- SM blocks
- 16
- L2 cache
- 1 MB
- L1 cache
- 128 KB
- Boost clock
- 1155–1477 MHz
- VRAM
- 4 GB
- Memory type
- GDDR6
- Memory speed
- 14 Gb/s
- Memory bus
- 64 bit
- Bandwidth
- 112 GB/s
- Interface
- PCIe 3.0 x8
- AI performance
- No data
Features
DLSS 2, DLSS Ray Reconstruction, DLSS Super Resolution, DLAA, NVIDIA Reflex, NVIDIA Broadcast, G-SYNC, Optimus, DirectX 12 Ultimate, Vulkan RT, OpenGL 4.6