Head to head
Compare Graphics Cards
Pick up to 4. Winners are highlighted in green ▲.
1 selected
Key specs
Memory
4 GB
Memory Type
GDDR6
Boost Clock
1057-1740 MHz (configurable)
TGP
35-95 W (configurable)
Architecture
Ampere
Benchmarks
Rating
27
Gaming
18
Render
27
Energy
60
TFLOPS (FP32)
4.3-7.1
Geekbench OpenCL
48,591
3DMarkTime Spy
4657
Steel Nomad Lite Score
5038
GB6 Compute Score
48,591
G3D Mark Score
9388
Blender GPU
1127.49
GB6 ML Single Precision
9684
GB6 ML Half Precision
18,365
GB6 ML Quantized
6700
FLOPS (FP32)
4.3-7.1 TFLOPS
Memory Bandwidth
176 GB/s
3D Mark
Time Spy
4657
Solar Bay
21,233
Port Royal
577
Fire Strike
12,109
GeekBench 6 OpenCL
Background Blur
80.6 img/sec
Face Detection
47.7 img/sec
Horizon Detection
2.14 Gpixels/sec
Edge Detection
2.82 Gpixels/sec
Gaussian Blur
2 Gpixels/sec
Feature Matching
0.61 Gpixels/sec
Stereo Matching
199.8 Gpixels/sec
Particle Physics
6124.5 FPS
API
OpenCL
Passmark Graphics
G2D Mark
524
DirectX 11
64 FPS
DirectX 12
47 FPS
GPU Compute
3834 Ops/s
GeekBench 6 ML
Image Classification (SP)
4109
Image Segmentation (HP)
10,426
Image Super Resolution (Q)
8368
Face Detection (HP)
18,130
Pose Estimation (Q)
32,595
Text Classification (SP)
2340
Machine Translation (HP)
3225
Object Detection (SP)
4684
Depth Estimation (Q)
16,324
Style Transfer (SP)
70,507
Framework
ONNX
Backend
DirectML
General
Vendor
Nvidia
Build
Discrete
Released
March 30, 2022
Case
Laptop
Purpose
Professional
Segment
Entry-level
Architecture
Ampere
GPU Codename
GA107
Successor
- Nvidia RTX 1000 Laptop (Ada)
Recommended CPU
- Intel Core i7 12800H or above
Graphics Processing Unit
Base Clock
712-1530 MHz (configurable)
Boost Clock
1057-1740 MHz (configurable)
Shading Units
2048
Texture Mapping Units (TMUs)
64
Render Output Units (ROPs)
32
Compute Units (Pipelines)
16
Tensor Cores
64
Ray-tracing Cores
16
L1 Cache
128KB per cluster
L2 Cache
2MB shared
Instructions Per Cycle
2 IPC
Raw Performance
Pixel Fill Rate
34-56 GPixel/s
Texture Fill Rate
68-111 GTexel/s
Physical
Interface
PCIe 4.0 x8
TGP
35-95 W (configurable)
Manufacturing
Samsung
Fabrication Process
8 nm
Die Size
200 mm²
Transistor Count
8.7 billion
Transistor Density
43.5 MTr/mm²
Memory
Memory Type
GDDR6
Memory Size
4 GB
Memory Clock
1375 MHz
Effective Memory Speed
11000 Mbps
Bus
128-bit
ECC
No
API
DirectX
12
Vulkan
1.3
OpenGL
4.6
OpenCL
3.0
CUDA
8.6
Ray Tracing
Yes
DLSS
DLSS 2
DisplayPort
1.4a