sbc.compare

ArmSoM Sige5 vs Radxa Cubie A7A

Benchmarks, specifications, and real-world performance data

Specification
Basic Specifications
Price
RAM8GB6GB
SoC
Mali-G52 MC3
Imagination BXM-4-64 MC1
ArchitectureARMARM
CPU Cores
8 cores
4× Cortex-A72 @ 2.2GHz
4× Cortex-A53 @ 1.8GHz
8 cores
2× Cortex-A76 @ 2.0GHz
6× Cortex-A55 @ 1.8GHz

Legend

Best value in category
Key performance metric
*Manufacturer's RRP
**Affiliate link that may result in commissions if you purchase a qualifying item
ArmSoM Sige5 (8GB)
Radxa Cubie A7A (6GB)

CPU Performance

Processor performance across single and multi-core workloads

Geekbench 6

Higher Scores are Better

Single-Core
ArmSoM Sige5 (8GB)
332
Radxa Cubie A7A (6GB)
641
Multi-Core
ArmSoM Sige5 (8GB)
1,440
Radxa Cubie A7A (6GB)
1,545

7-Zip Benchmark

Higher Scores are Better

Single-Core Combined
ArmSoM Sige5 (8GB)
1,796 MIPS
Radxa Cubie A7A (6GB)
2,347 MIPS
Multi-Core Combined
ArmSoM Sige5 (8GB)
11,391 MIPS
Radxa Cubie A7A (6GB)
10,509 MIPS

High Performance Linpack

Higher Scores are Better

Performance
ArmSoM Sige5 (8GB)
22.61 GFLOPS
Radxa Cubie A7A (6GB)
18.77 GFLOPS
Performance/Watt
ArmSoM Sige5 (8GB)
2.61 GFLOPS/W
Radxa Cubie A7A (6GB)
2.04 GFLOPS/W

UnixBench

Higher Scores are Better

Single Core
ArmSoM Sige5 (8GB)
492
Radxa Cubie A7A (6GB)
1,125
Multi Core
ArmSoM Sige5 (8GB)
3,252
Radxa Cubie A7A (6GB)
4,001

PassMark CPU

Higher Scores are Better

CPU Score
ArmSoM Sige5 (8GB)
2,189
Radxa Cubie A7A (6GB)
2,042

CPU Mining (cpuminer-multi)

Higher Hashrates are Better

Single Core
ArmSoM Sige5 (8GB)
2.7 kH/s
Radxa Cubie A7A (6GB)
3.59 kH/s
Multi Core
ArmSoM Sige5 (8GB)
17.73 kH/s
Radxa Cubie A7A (6GB)
18.24 kH/s

Memory Performance

Memory bandwidth and throughput benchmarks

TinyMemBench

Higher Bandwidth is Better

memcpy
ArmSoM Sige5 (8GB)
5,661.6 MB/s
Radxa Cubie A7A (6GB)
No Data
memset
ArmSoM Sige5 (8GB)
16,885.8 MB/s
Radxa Cubie A7A (6GB)
No Data

PassMark RAM

Higher Scores are Better

RAM Score
ArmSoM Sige5 (8GB)
1,004
Radxa Cubie A7A (6GB)
978

AI & LLM Performance

Large Language Model inference performance

Ollama LLM Benchmark

Higher Scores are Better

Tokens/sec
ArmSoM Sige5 (8GB)
8.22 tok/s
Radxa Cubie A7A (6GB)
0.63 tok/s
Tokens/Watt
ArmSoM Sige5 (8GB)
1.01 tok/W
Radxa Cubie A7A (6GB)
0.09 tok/W
Tokens/Dollar
ArmSoM Sige5 (8GB)
0.07 tok/$
Radxa Cubie A7A (6GB)
0.01 tok/$

Storage Performance

Disk I/O throughput measured with FIO

fio Storage Performance - microSD

Higher Speeds are Better • Raspberry Pi 64GB (64GB microSD)

4K Random Read
ArmSoM Sige5 (8GB)
No Data
Radxa Cubie A7A (6GB)
9.1 MB/s
4K Random Write
ArmSoM Sige5 (8GB)
No Data
Radxa Cubie A7A (6GB)
15.3 MB/s

fio Storage Performance - eMMC

Higher Speeds are Better • ArmSoM Sige5 (64GB eMMC)

4K Random Read
ArmSoM Sige5 (8GB)
50.5 MB/s
Radxa Cubie A7A (6GB)
No Data
4K Random Write
ArmSoM Sige5 (8GB)
76.8 MB/s
Radxa Cubie A7A (6GB)
No Data

fio Storage Performance - M.2 NVMe

Higher Speeds are Better • Crucial P5 Plus (1024GB M.2 NVMe)

4K Random Read
ArmSoM Sige5 (8GB)
375.5 MB/s
Radxa Cubie A7A (6GB)
No Data
4K Random Write
ArmSoM Sige5 (8GB)
307.6 MB/s
Radxa Cubie A7A (6GB)
No Data

fio Storage Performance - UFS

Higher Speeds are Better • Radxa UFS (128GB UFS)

4K Random Read
ArmSoM Sige5 (8GB)
No Data
Radxa Cubie A7A (6GB)
509.6 MB/s
4K Random Write
ArmSoM Sige5 (8GB)
No Data
Radxa Cubie A7A (6GB)
235.9 MB/s

Network Performance

Ethernet and WiFi throughput measured with iPerf3

iPerf3 Network Performance

Higher Speeds are Better

Ethernet Up
ArmSoM Sige5 (8GB)
938 Mbps
Radxa Cubie A7A (6GB)
941 Mbps
Ethernet Down
ArmSoM Sige5 (8GB)
939 Mbps
Radxa Cubie A7A (6GB)
939 Mbps

Power & Thermal

Power consumption at idle and under various workloads

Power Consumption

Lower is "Better"

Idle Power
ArmSoM Sige5 (8GB)
3W
Radxa Cubie A7A (6GB)
2.4W
Load Power
ArmSoM Sige5 (8GB)
6.1W
Radxa Cubie A7A (6GB)
5.9W
Ollama Max
ArmSoM Sige5 (8GB)
8.6W
Radxa Cubie A7A (6GB)
7.1W
Linpack Max
ArmSoM Sige5 (8GB)
8.7W
Radxa Cubie A7A (6GB)
9.2W

Add One More Board

You can compare up to 3 boards at once. Add one more to get the complete comparison.