Thundercomm TurboX C8550 Platform Comparison

From RidgeRun Developer Wiki


Index





The Thundercomm TurboX C8550 powered by the Qualcomm Snapdragon 8, offers several distinct advantages when compared to its counterparts developoed by NVIDIA: the Jetson AGX Orin and Jetson Orin NX.

One of the most significant strengths of the TurboX C8550 lies in its superior mobile and connectivity capabilities. Unlike the Jetson boards, which do not have integrated connectivity solutions, the C8550 comes equipped with cutting-edge 5G, Wi-Fi 7, and Bluetooth 5.3 technologies. This makes it an ideal choice for applications requiring high-speed wireless communication and seamless connectivity, particularly in mobile and IoT scenarios.

The C8550's CPU architecture also stands out. Its octa-core configuration, featuring a high-performance 3.2 GHz Cortex-X3 core, provides a balanced mix of high-performance and energy-efficient cores. This heterogeneous architecture allows for more flexible task allocation and potentially better performance in scenarios that benefit from single-threaded performance.

In terms of graphics capabilities, while the NVIDIA boards excel in CUDA-based computations, the Adreno 740 GPU in the C8550 is highly optimized for mobile graphics for example, the one needed in Autonomous Mobile Robots(AMRs).

The C8550 also boasts UFS 4.0 storage, which offers faster data transfer speeds compared to the eMMC 5.1 storage found in the Jetson boards. This can result in quicker boot times and faster application loading, enhancing overall system responsiveness.

In summary, while the NVIDIA Jetson boards have their strengths in AI and computer vision tasks (the Jetson boards might offer higher raw AI performance in terms of TOPS), the Thundercomm TurboX C8550 shines in scenarios requiring high mobile performance, advanced wireless connectivity, mobile-optimized performance, and a balance between computing power and energy efficiency.

Table 1 compares the features of the Thundercomm TurboX C8550 against two different embedded systems used in the market for similar applications: NVIDIA Jetson AGX Orin 64GB and NVIDIA Jetson Orin NX 16GB.

Table 1: Comparison, Thundercomm TurboX C8550 vs Nvidia Plataforms [1] [2] [3] [4] [5] [6] [7] [8] [9]
Feature Thundercomm TurboX C8550 NVIDIA Jetson AGX Orin 64GB NVIDIA Jetson Orin NX 16GB
Processor Qualcomm® Snapdragon™ QCS8550 NVIDIA Orin system on chip (SoC) NVIDIA Orin system on chip (SoC)
CPU Octa-core
  • 1 x 3.2 GHz Cortex-X3
  • 2 x 2.8 GHz Cortex-A715
  • 2 x 2.8 GHz Cortex-A710
  • 3 x 2.0 GHz Cortex-A510
12-core Arm Cortex-A78AE v8.2 (up to 2.2 GHz) 8-core Arm® Cortex®-A78AE v8.2 (up to 2.0 GHz)
GPU Qualcomm® Adreno™ 740 NVIDIA Ampere (2048 CUDA cores, 64 Tensor cores) 1024-core NVIDIA Ampere architecture GPU with 32 Tensor Cores
AI performance
  • HTP INT8 48 TOPs
  • GPU FP16 12 TOPs
Up to 275 TOPS Up to 100 TOPS
ISP Acceleration Qualcomm Spectra™ Image Signal Processor - Cognitive ISP, Triple 18-bit ISPs
  • Gen 4.2 VIC, two GPix/s
  • 1x PVA v2.0
  • VIC
  • 1x PVA v2.0
DSP Acceleration
  • Compute DSP: V73 AI-optimized tensor processor, six threads scalar DSP
  • Sensor DSP: Qualcomm® Sensing Hub 3.0
2x NVDLA v2 2x NVDLA v2
Encode/Decode
  • UHD video processing unit
  • Video decode up to 4K240/8K60
  • Video encode up to 4K120/8K30
  • Concurrent 4K60 decode and 4K60 encode for wireless display
  • AV1 decode
  • Native decode support for H.265 Main 10, H.265 Main, H.264 High, and VP9 profile 2
  • Native encode support for H.265 Main 10, H.265 Main, H.264 high formats
Encode:
  • 2x 4K60 (H.265)
  • 4x 4K30 (H.265)
  • 8x 1080p60 (H.265)
  • 16x 1080p30 (H.265)
Decode:
  • 1x 8K30 (H.265)
  • 3x 4K60 (H.265)
  • 7x 4K30 (H.265)
  • 11x 1080p60 (H.265)
  • 22x 1080p30 (H.265)
Encode:
  • 1x 4K60 (H.265)
  • 3x 4K30 (H.265)
  • 6x 1080p60 (H.265)
  • 12x 1080p30 (H.265)
Decode:
  • 1x 8K30 (H.265)
  • 2x 4K60 (H.265)
  • 4x 4K30 (H.265)
  • 9x 1080p60 (H.265)
  • 18x 1080p30 (H.265)
Camera Support
  • Inputs: 5 cameras
  • 4 x CSI D-PHY-1.2 4-lane 2.5 Gbps/lane
  • 2 x CSI C-PHY-2.0 3-trio/D-PHY-1.2 4-lane compatible, C-PHY 13.68 Gbps/trio
  • 2 x CSI D-PHY-1.2 2-lane
  • 5 x concurrent, 3 IFE + 2 IFE-lites
  • Low Power AI subsystem for AON camera (CSI4 with PM8010) (TBD)
  • 3 x 36 MP@30fps – Triple cameras
  • 64 MP + 36 MP@30fps –Dual cameras
  • 1 x 108 MP@30fps – Single camera
  • Up to 6 cameras (16 via virtual channels)
  • 16 lanes MIPI CSI-2
  • D-PHY 2.1 (up to 40Gbps)
  • C-PHY 2.0 (up to 164Gbps)
  • Up to 4 cameras (8 via virtual channels)
  • 8 lanes MIPI
  • CSI-2 D-PHY 2.1 (up to 20Gbps)
Memory 12 GB LPDDR5x, memory up to 4200 MHz 64GB 256-bit LPDDR5 204.8GB/s 16GB 128-bit LPDDR5 102.4GB/s
Storage 128 GB (UFS 3.1/4.0) 64GB eMMC 5.1 (Supports external NVMe)
Security
  • Qualcomm® Trusted Execution Environment (TEE) v5.3
  • Qualcomm® Type-1 Hypervisor enables multiple trusted VMs (TVMs)
Platform Security Controller (PSC):
  • RISC-V subsystem, PKC crypto (RSA3K)
Sensor Processing Engine (SPE):
  • Arm Cortex-R5 processor
  • Always-on (AON) cluster
  • Platform Security Controller (PSC)
  • Security Engine (SE)
Power Consumption 6W - 14W 15W-60W 10W-25W
Integrated connectivity
  • Wi-Fi 7
  • Bluetooth® 5.3
None None

References

  1. Thundercomm TurboX C8550, product brief. Retrieved September 13, 2024, from [1]
  2. Thundercomm TurboX C8550 SOM . Retrieved September 13, 2024, from [2]
  3. Thundercomm TurboX C8550 SOM datasheet. Retrieved September 13, 2024, from [3]
  4. Pocessors QCS8550. Retrieved September 13, 2024, from [4]
  5. Qualcomm® QCS8550/QCM8550 Processors. Retrieved September 13, 2024, from [5]
  6. NVIDIA Jetson Modules. Retrieved September 13, 2024, from [6]
  7. NVIDIA Jetson Orin Modules. Retrieved September 13, 2024, from [7]
  8. NVIDIA Jetson Orin NX Series Modules Data Sheet. Retrieved September 13, 2024, from [8]
  9. NVIDIA Jetson AGX Orin Series Data Sheet. Retrieved September 13, 2024, from [9]


Index