EVERYTHING YOU NEED TO KNOW ABOUT JETSON ORIN NANO — Part 2

The performance

The Jetson Orin Nano offers 80x the AI performance of the older Jetson Nano, but this is based on INT8 precision on the new device and FP16 on the old. When using FP32 precision on both devices, the gain drops to 5.4x, which is still impressive. In addition to switching to a new GPU architecture, the Jetson Orin Nano has eight times as many CUDA cores as the Jetson Nano, plus 32 Tensor cores as well. Two additional cores are running at marginally faster clock speeds, and there is double the memory with the move to LPDDR5 which offers twice as much bandwidth. In addition, the processor has moved to a newer Arm Cortex architecture, which offers two and a half times the bandwidth of the Jetson Nano. However, there are some sacrifices, including the lack of NVDLAs, PVAs, and hardware video encoders. The focus is on on-device edge AI, and the Jetson Orin Nano performs significantly better than the Jetson Nano on various networks. The Jetson Orin Nano is efficient and configurable in 15W full-performance and 7W reduced-power modes.

Jetson Orin Nano VS Jetson Nano

It is important not to compare the two Nanos directly, as they serve vastly different markets, and are only similar in name. NVIDIA’s $149 Jetson Nano remains an entry-level artificial intelligence platform designed for education and makers, while the Jetson Orin Nano is much higher performance than the older, higher-end Jetson TX2 and Xavier NX-based systems designed for more serious projects and industrial deployments.

There are 4 Cortex A57 cores clocked at 1.47 GHz on the original Jetson Nano, along with 128 Maxwell GPU cores and 4GB of LPDDR4 RAM. It has a lower power rating than its newer counterpart. It is clear that edge technology has advanced a lot in the past few years when you compare the specs side-to-side.

Specification comparison

Jetson Orin Nano

Jetson Nano

CPU

6-core Arm Cortex-A78AE v8.2 64-bit CPU

Quad-core ARM Cortex-A57 MPCore processor

1.5MB L2 + 4MB L3

GPU

Nvidia Ampere architecture with 1024 Nvidia CUDA cores and

Nvidia Maxwell architecture with 128 Nvidia CUDA cores

32 Tensor cores

Memory

8GB 128-bit LPDDR5

4 GB 64-bit LPDDR4, 1600MHz 25.6 GB/s

68 GB/s

Storage

Micro SD

16 GB eMMC 5.1

NVMe M.2 via Carrier Board

Micro SD

Power

7W to 15W (5V Input Voltage)

20W (Max 5V at 4 Amps)

Dimensions

69 x 45 x 21 mm

69.6 x 45 x 20 mm

Camera

2x MIPI CSI-2 22-pin Camera Connectors

12 lanes (3x4 or 4x2) MIPI CSI-2 D-PHY 1.1

M.2 Key M

x4 PCIe Gen 3

X2 PCIe Gen 3

M.2 Key E

PCIe (x1), USB 2.0, UART, I2S, and I2C

1 x M.2 Key E

USB

4 x USB 3.2 Gen2

4x USB 3.0

1 x Type C for debug and device mode

1 x USB 2.0 Micro-B

Networking

Gigabit Ethernet

Gigabit Ethernet

RTL8822CE 802.11ac PCIe Wireless Network Adapter

Display

DisplayPort 1.2

HDMI 2.0 and eDP 1.4

GPIO

40 Pin GPIO

40 Pin GPIO

12 Pin Button Header

4 Pin Fan Header

Power

DC 9–19V Barrel Jack

DC Barrel Jack 20W (Max 5V at 4 Amps)

Dimensions

100 x 79 x 21 mm (Height includes Orin Nano module and cooling solution)

100 x 80 x 29mm (Height includes Jetson Nano module and cooling solution)

In passing, there is no visible difference between the Orin Nano and the Jetson Nano, except that the Orin Nano has a fan built into the heatsink and there is no HDMI port, instead, there is a USB-C port which replaces the Jetson Nano’s micro USB port. Even when running at the full 15W, the fan remained quiet, unlike other SBC fans which have been tested before. The fan remained silent even when testers ran one of Nvidia’s recommended inference benchmarks.

Final Words

The new Jetson Orin Nano has impressed reviewers with its powerful hardware and software capabilities. It is one of the fastest and graphically most powerful single-board computers tested, and NVIDIA’s support and tools make AI development fast and easy. Although it lacks some hardware features, this does not take away from the impressive performance achievable from such a small device. While it is more expensive than most SBCs at $499, it offers more features and capabilities. For beginners in AI, the $149 Jetson Nano may be a better option, but for those experienced in the field, the Jetson Orin Nano is a must-have for tackling more demanding projects.

Whether you are a maker looking for a powerful SBC for AI or the classic Raspberry Pi product, Elecrow provides you with the necessary electronic components for your DIY electronic projects with these powerful embedded SBCs, and Elecrow can even help you turn your DIY electronic ideas into reality and mass produce them for commercial purposes and make your own profits! Join Elecrow’s Partner Seller Program for more details.