GPUDesktopReleased Dec 7, 2021

NVIDIA GeForce RTX 2060 12GB

192-bit · 2,176 cores

The GeForce RTX 2060 12GB is a late addition to NVIDIA's Turing lineup, released on December 7, 2021 during a period of severe graphics card shortages. Compared with the original 6 GB RTX 2060 it doubles the memory and enables more of the TU106 GPU, bringing its shader count level with the RTX 2060 SUPER. NVIDIA sold it only through board partners, with no Founders Edition and no official launch price.

General

Manufacturer
NVIDIA
Model
GeForce RTX 2060 12GB
Segment
Desktop
GPU chip
TU106
Architecture
Turing
Release date
Dec 7, 2021

Memory

Memory size
12 GB
Memory type
GDDR6
Bus width
192-bit
Bandwidth
336 GB/s
Memory clock
1,750 MHz

Clocks & silicon

GPU clock
1,470 MHz
Boost clock
1,650 MHz
Shader units
2,176 CUDA Cores
Tensor / AI cores
272
RT cores
34
Transistors
10.8 billion
Technology (nm)
12 nm
TSMC 12 nm FFN
TDP / TBP
185 W

Interface & APIs

Bus interface
PCIe 3.0 x16
SLI / multi-GPU
None (no NVLink / SLI)
DirectX
12 Ultimate (12_2)
OpenGL
4.6
Vulkan
1.4
Other APIs
OpenCL 3.0, CUDA 7.5, DLSS 4.5 (Super Resolution), NVENC (7th Gen)

AI support

Precisions
FP8FP16BF16INT8INT4TF32FP64
AI accelerator
2nd Gen Tensor Cores (272)
Peak AI TOPS
—
Notes
Turing 2nd-generation Tensor Cores accelerate FP16 (with FP16 or FP32 accumulate) plus INT8, INT4 and INT1 inference modes. BF16, TF32 and FP8 are not supported - BF16 and TF32 arrived with Ampere and FP8 with Ada/Hopper. FP64 runs on the CUDA cores at 1/32 the FP32 rate rather than on the Tensor Cores. NVIDIA published no AI TOPS rating for this card.
All NVIDIA GPUs →

Desktop gpus from NVIDIA announced around the same time as the NVIDIA GeForce RTX 2060 12GB. Add any of them to your compare list to see the specs side by side.

NVIDIAgpu

NVIDIA CMP 30HX

6 GB GDDR6 · 192-bit · 1,408 cores

Released
Feb 18, 2021