glmark2 (OpenGL) Benchmark Results

Compare OpenGL ES 2.0 GPU performance across Single Board Computers using glmark2 at 1080p and 4K resolutions

24 boards tested · scores from 19 to 1,717 (1080p Score)

24 results

Full Results

Compare#
Board
1080p Score
4K Score
1
Radxa
Orion O6N
32GB LPDDR5
1,717632
2
Radxa
Dragon Q8B
32GB LPDDR4
1,457480
3
Radxa
NX5
8GB LPDDR4X
1,152535
4
Radxa
Fogwise Airbox Q900
36GB LPDDR5
973324
5
Radxa
ROCK 5B+
16GB LPDDR5
947279
6
Libre Computer
Alta
4GB LPDDR4
731217
7
Banana Pi
M2S (S922X)
4GB LPDDR4
693211
8
HardKernel
ODROID N2+
2GB DDR4
681200
9
HardKernel
ODROID N2+
4GB DDR4
678198
10
Raspberry Pi
5
8GB LPDDR4X
657150
11
Raspberry Pi
Compute Module 5
4GB LPDDR4X
648157
12
ASUS
Tinker Board S
2GB LPDDR3
351109
13
Raspberry Pi
4
2GB LPDDR4
19052
14
HardKernel
ODROID C2
2GB DDR3
12822
15
Banana Pi
M5
4GB LPDDR4
11524
16
Raspberry Pi
3 (Model A+)
512MB LPDDR2
106N/A
17
Raspberry Pi
Compute Module 3+ Lite (CM3+)
1GB LPDDR2
106N/A
18
Raspberry Pi
Compute Module 0 (CM0)
512MB LPDDR2
105N/A
19
Orange Pi
6
16GB LPDDR5
9757
20
Orange Pi
PC Plus
1GB LPDDR3
7817
21
PINE64
PINE A64
2GB DDR3
629
22
Orange Pi
Lite
512MB
59N/A
23
Raspberry Pi
1 (Model B+ v1.2)
512MB LPDDR2
25N/A
24
SolidRun
HummingBoard I2
1GB DDR3
19N/A
Showing 24 results

Frequently Asked Questions

What is the fastest single board computer in glmark2 (OpenGL)?

The Radxa Orion O6N 32GB currently tops our glmark2 (OpenGL) ranking with 1,717 (1080p Score), measured on real hardware with our standardised test setup.

What is a good glmark2 (OpenGL) score for an SBC?

Across the 24 boards we have tested, glmark2 (OpenGL) results range from 19 to 1,717 (1080p Score), with a median of 351. Anything above the median puts a board in the faster half of the market.

About glmark2 (OpenGL) testing

glmark2 is an OpenGL ES benchmark that renders a series of increasingly demanding 3D scenes and reports a combined score. It is our headline GPU comparison for boards with working open source or vendor graphics drivers. Driver quality matters as much as silicon here: two boards with the same GPU can score very differently depending on the state of their graphics stack, which is exactly the kind of thing these results surface.