GeekBench v6

Background Blur	112.8 frames/s
Face Detection	66.5 frames/s
Horizon Detection	2.36 Gpix/sec
Edge Detection	3 Gpix/sec
Gaussian Blur	3.67 Gpix/sec
Feature Matching	0.85 Gpix/sec
Stereo Matching	287.1 Gpix/sec
Particle Physics	9494.1 fps
API	OpenCL

PassMark

G2D Mark	702
DirectX 11	99 fps
DirectX 12	69 fps
GPU Compute	5058 Ops/s

Geekbench 6 ML

Image Classification (SP)	4878
Image Segmentation (HP)	15282
Horizon Detection	12309
Face Detection (HP)	27767
Pose Estimation (Q)	44436
Text Classification (SP)	2471
Machine Translation (HP)	3958
Object Detection (SP)	6278
Depth Estimation (Q)	23945
Style Transfer (SP)	70513
Framework	ONNX
Backend	DirectML

Manufacturer	Nvidia
Build	Discrete
Announced	March, 18, 2025
Form Factor	Laptop use
Use case	Professional
Segment	Entry-level
Architecture	Blackwell 2.0
Codename	GB207-300-A1

GPU Frequency	1357-1792 MHz
Boost GPU Frequency	1792-2332 MHz
Shader Units	1792
TMUs	56
ROPs	24
Execution Units	14
Ray Tracing Cores	14
Tensor Cores	56
L1 cache	128 KB/core
L2 cache	24000 KB/core
Instructions per Clock	2 IPC

Pixel Fill Rate	43-56 Gpix/sec
Texture Fill Rate	100-131 GTex/s
FLOPS	6.4-8.4 TFLOPS

Interface	PCIe 5.0 x16
TGP	35-75 W
Process	TSMC
Process tech.	5 nm
Transistor Size	149 mm²
Transistor count	16.9 billion
Transistor density	113.42 MTr/mm²

Interface	GDDR7
Memory Frequency	1500 MHz
Bus	96 bit
ECC support	No
Effective Memory Speed	24000 MB/s
Memory Size	6 GB

DirectX	12.2
Vulkan	1.4
OpenGL	4.6
CUDA	12
DisplayPort	2.1b
Ray Tracing	Yes
DLSS	DLSS 4

Сообщить об ошибке / Report an error