NVIDIA Jetson Module Comparison From Orin Family to Thor: What You Need to Know
The NVIDIA Jetson ecosystem has expanded rapidly, with Orin modules powering everything from entry-level AI applications to industrial robotics. Now, the arrival of Jetson Thor brings a massive leap in compute and efficiency. In this article, we’ll break down the Jetson Orin family and compare it with the new Thor series. To make things clear, we’ll start with a detailed comparison table and then walk through each module series.
Jetson Module Comparison at a Glance
| Feature | Orin Nano 4GB (with super mode support) | Orin Nano 8GB (with super mode support) | Orin NX 8GB (with super mode support) | Orin NX 16GB (with super mode support) | Jetson AGX Orin 32GB | Jetson AGX Orin 64GB | Jetson AGX Orin Industrial | Jetson T5000 Module |
|---|---|---|---|---|---|---|---|---|
| AI Performance | Up to 34 TOPS | Up to 67 TOPS | Up to 117 TOPS | Up to 157 TOPS | 200 TOPS | 248 TOPS | 275 TOPS | 2,070 TFLOPS (FP4, sparse) |
| GPU | 512-core Ampere, with 16 Tensor Cores | 1024-core Ampere, with 32 Tensor Cores | 1024 Core Ampere, with 32 Tensor Cores | 1024 Core Ampere, with 32 Tensor Cores | 1792 Core Ampere, with 56 Tensor Cores | 2048 Core Ampere, with 64 Tensor Cores | 2048 Core Ampere, with 64 Tensor Cores | 2560-core NVIDIA Blackwell architecture GPU with 96 fifth-gen Tensor Cores Multi-Instance GPU (MIG) with 10 TPCs |
| CPU | 6-core Arm® Cortex®-A78AE | 6-core Arm® Cortex®-A78AE | 6-core Arm® Cortex®-A78AE | 8-core Arm® Cortex®-A78AE | 8-core Arm® Cortex®-A78AE | 12 core Arm® Cortex®- A78AE | 12 core Arm® Cortex®- A78AE | 14-core Arm® Neoverse®- V3AE 64-bit CPU 1 MB L2 cache per core 16 MB shared system L3 cache |
| Memory | 4GB 64-bit LPDDR5 34 GB/s | 8GB 128-bit LPDDR5 68 GB/s | 8GB 128-bit LPDDR5 102.4 GB/s | 16GB 128-bit LPDDR5 102.4 GB/s | 32GB 256-bit LPDDR5 205 GB/s | 64GB 256-bit LPDDR5 205 GB/s | 64GB 256-bit LPDDR5 205 GB/s | 128 GB 256-bit LPDDR5X 273 GB/s |
| DL Accelerator | - | - | (1x) NVDLA V2.0 | (2x) NVDLA V2.0 | (2x) NVDLA V2.0 | (2x) NVDLA V2.0 | (2x) NVDLA V2.0 | - |
| Vision Accelerator | - | - | 1x PVA v2 | 1x PVA v2 | 1x PVA v2 | 1x PVA v2 | 1x PVA v2 | 1x PVA v3 |
| Storage | Supports External NVMe | Supports External NVMe | Supports External NVMe | Supports External NVMe | 64GB eMMC | 64GB eMMC | 64GB eMMC | Supports NVMe through Pcle Supports SSD through USB3.2 |
| Video Encode | 1080p30 supported by 1-2 CPU cores | 1080p30 supported by 1-2 CPU cores | 1x 4K60 | 3x 4K30| 6x 1080p60 | 12× 1080p30 (H.265), H.264, H.265, AV1 | 1x 4K60 | 3x 4K30| 6x 1080p60 | 12x 1080p30 (H.265), H.264, H.265, AV1 | 1x 4K60 | 3x 4K30/ 6x 1080p60 | 12x 1080p30 (H.265), H.264, H.265, AV1 | 1x 4K60 (H.265) 3x 4K30 (H.265) 7x 1080p60 (H.265) 15x 1080p30 (H.265) | 2x 4K60 4× 4K30 | 8x 1080p60 | 16x 1080p30 (H.265) H.264, AV1 | 6x 4Kp60 (H.265) 12x 4Kp30 (H.265) 24× 1080p60 (H.265) 50x 1080p30 (H.265) 48× 1080p30 (H.264) 6x 4Kp60 (H.264) |
| Video Decode | 1x 4K60 (H.265) 2x 4K30 (H.265) 5x 1080p60 (H.265) 11 x 1080p30 (H.265) | 1x 4K60 (H.265) 3x 4K30 (H.265) 6x 1080p60 (H.265) 12x 1080p30 (H.265) | 1x 4K60 (H.265) 3x 4K30 (H.265) 3x 4K30 (H.265) 3x 4K30 (H.265) | 1× 8K30 (H.265) 2x 4K60 (H.265) 4x 4K30 (H.265) 9x 1080p60 (H.265) 18x 1080p30 (H.265) | 1 × 8K30 (H.265) 2x 4K60 (H.265) 4x 4K30 (H.265) 9x 1080p60 (H.265) 18x 1080p30 (H.265) | 1× 8K30 (H.265) 3x 4K60 (H.265) 7x 4K30 (H.265) 11× 1080p60 (H.265) 23x 1080p30 (H.265) | 1× 8K30 (H.265) 3x 4K60 (H.265) 7x 4K30 (H.265) 11x 1080p60 (H.265) 22x 1080p30 (H.265) | 4x 8Kp30 (H.265) 10x 4Kp60 (H.265) 22x 4Kp30 (H.265) 46х 1080p60 (H.265) 92x 1080p30 (H.265) 82x 1080p30 (H.264) 4x 4Kp60 (H.264) |
| Camera | Up to 4 cameras (8 via virtual channels***) 8 lanes MIPI CSI-2 D-PHY 2.1 (up to 20Gbps) | Up to 4 cameras (8 via virtual channels***) 8 lanes MIPI CSI-2 D-PHY 2.1 (up to 20Gbps) | Up to 4 cameras (8 via virtual channels***) 8 lanes MIPI CSI-2 D-PHY 2.1 (up to 20Gbps) | Up to 4 cameras (8 via virtual channels***) 8 lanes MIPI CSI-2 D-PHY 2.1 (up to 20Gbps) | 16 lane MIPI CSI-2 connector | 16 lane MIPI CSI-2 connector | 16 lane MIPI CSI-2 connector | Up to 20 cameras via HSB Up to 6 cameras through 16x lanes MIPI CSI-2 Up to 32 cameras using Virtual Channels C-PHY 2.1 (10.25 Gbps) D-PHY 2.1 (40 Gbps) |
| PCI Express | 1 x4 + 3 x1 (PCle Gen3, Root Port, & Endpoint) | 1 x4 + 3 x1 (PCle Gen3, Root Port, & Endpoint) | 1 x4 + 3 x1 (PCle Gen4, Root Port, & Endpoint) | 1x4 + 3 x1 (PCle Gen4, Root Port, & Endpoint) | Up to 2x8 + 1x4 + 2 x1 (PCle Gen4, Root Port, & Endpoint) | Up to 2 x8 + 1 x4 + 2 x1 (PCle Gen4, Root Port, & Endpoint) | Up to 2 x8 + 1x4 + 2 ×1 (PCle Gen4, Root Port, & Endpoint) | Up to Gen5 (x8 lanes) Root port only—C1 (x1) and C3 (x2) Root Point or Endpoint—C2 (x1), C4 (x8), and C5 (x4) |
| Mechanical | 69.6mm x 45mm 260-pin SO-DIMM connector | 69.6mm x 45mm 260-pin SO-DIMM connector | 69.6mm x 45mm 260-pin SO-DIMM connector | 69.6mm x 45mm 260-pin SO-DIMM connector | 100mm x 87mm 699-pin Molex Mirror Mezz Connector Integrated Thermal Transfer Plate | 100mm x 87mm 699-pin Molex Mirror Mezz Connector Integrated Thermal Transfer Plate | 100mm x 87mm 699-pin Molex Mirror Mezz Connector Integrated Thermal Transfer Plate | 100 mm x 87 mm 699-pin B2B connector Integrated Thermal Transfer Plate (TTP) with heatpipe |
| Power | 7W - 25W | 7W - 25W | 10W - 40W | 10W - 40W | 15W - 40W | 15W - 75W | 15W - 60W | 40W - 130W |
What is the difference between Jetson AGX Orin, Orin NX, Orin Nano and AGX Thor?
Jetson AGX Thor Series
The NVIDIA Jetson Thor series is designed for the most demanding robotics and physical AI platforms. Delivering up to 2070 FP4 TFLOPS of compute and 128 GB memory, Thor is configurable between 40 W and 130 W. Compared to AGX Orin, it offers 7.5x higher AI compute and 3.5x better energy efficiency, making it the ultimate choice for next-generation autonomous systems.
Recommended FORECR products:
Jetson AGX Orin Series
With up to 275 TOPS, Jetson AGX Orin is NVIDIA’s most powerful AI computer for energy-efficient machines today. It delivers 8x the performance of its predecessor while supporting multiple concurrent AI inference pipelines and high-speed sensor interfaces. Applications range from manufacturing and logistics to healthcare and retail.
Recommended FORECR products:
Jetson Orin NX Series
The Jetson Orin NX is the sweet spot between power and compact design. In its smallest form factor, it offers up to 157 TOPS, along with double the CUDA cores. It’s ideal for autonomous machines that need both high compute and efficient energy use in tight spaces.
Recommended FORECR products:
Jetson Orin Nano Series
The entry-level Jetson Orin Nano sets a new baseline for edge AI, offering up to 67 TOPS at 7 W to 25 W power. That’s 140x more performance than the original Jetson Nano, enabling developers to bring AI workloads into compact, low-power devices.
Recommended FORECR products:
Final Thoughts
The Jetson Orin family gives developers a wide range of compute options, scaling from low-power Nano modules to industrial-grade AGX Orin. With Jetson Thor, NVIDIA is redefining the upper limits of AI performance for robotics and edge computing. The choice now depends on your application’s power budget, size constraints, and compute needs.
Be Among the First to Experience the THRMAX Family
