> ## Content Index
> Fetch the complete content index at: https://itsfoss.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# SAKURA-II Brings Energy-Efficient Edge AI to Raspberry Pi 5
- URL: https://itsfoss.com/news/sakura-ii-raspberry-pi-5/
- Published: 2025-06-15T05:03:04.000Z
- Updated: 2025-06-15T05:03:04.000Z
- Description: The SAKURA-II is an interesting bit of kit for the Raspberry Pi 5.
- Author: Sourav Rudra
- Tags: Gadgets 🎛️, News, #Import 2025-11-08 21:15

Specializing in energy-efficient [edge AI](https://en.wikipedia.org/wiki/Edge%5Fcomputing?ref=itsfoss.com) hardware and software solutions, [EdgeCortix](https://www.edgecortix.com/en/?ref=itsfoss.com) is a Tokyo-based manufacturer of AI accelerators and platforms. 

The company’s core offering is the [SAKURA-II](https://www.edgecortix.com/en/products/sakura?ref=itsfoss.com) series, which leverages its proprietary Dynamic Neural Accelerator (DNA) architecture to deliver high-performance, low-latency AI inferencing for real-time applications.

In its latest move, EdgeCortix [has introduced](https://www.edgecortix.com/en/press-releases/edgecortixs-sakura-ii-ai-accelerator-brings-low-power-generative-ai-to-raspberry-pi-5-and-other-arm-based-platforms?ref=itsfoss.com) an M.2 form factor SAKURA-II AI accelerator module that is compatible with the [Raspberry Pi 5](https://itsfoss.com/raspberry-pi-5/) and any single-board computer (SBC) featuring an M.2 2280 slot with PCIe Gen 3.0 x4 support.

## 📝 SAKURA-II M.2: Key Specifications

![an illustration that shows a block diagram of the sakura-ii m.2 on the right, and the module itself on the bottom left](https://itsfoss.com/content/images/2025/06/sakura-ii-m2-block-diagram.png)

**The SAKURA-II's* [**block diagram*](https://en.wikipedia.org/wiki/Block%5Fdiagram?ref=itsfoss.com)**.*

Equipped with **60 TOPS of INT8 performance** and **30 TFLOPS of BF16 compute capability**, the SAKURA-II M.2 can be configured with up to 16 GB of LPDDR4X RAM and up to 68 GB/sec memory bandwidth, enabling it to handle large models and complex workloads with ease.

Thanks to its M.2 2280 form factor, **it can easily connect with the Raspberry Pi 5, or to any other single-board computer or embedded system** that requires low-power, high-performance AI acceleration for tasks such as real-time object detection, large language model inference, or multi-modal edge AI applications.

It’s specifically designed for space-constrained environments and to efficiently run generative AI models and workloads such as [Llama 2](https://www.llama.com/llama2/?ref=itsfoss.com), [Stable Diffusion](https://stability.ai/stable-image?ref=itsfoss.com), [DETR](https://github.com/facebookresearch/detr?ref=itsfoss.com), and [Vision Transformers](https://en.wikipedia.org/wiki/Vision%5Ftransformer?ref=itsfoss.com), all while maintaining **a power consumption figure of around 10 watts**.

There's also a more powerful PCIe version of the SAKURA-II, available in single and dual chip configurations, designed for higher-throughput edge servers and industrial applications.

## 🛒 Getting the SAKURA-II M.2

Priced at **$349**, the SAKURA-II M.2 can be sourced from the [official website](https://www.edgecortix.com/en/evaluation-hardware?ref=itsfoss.com). There is no mention whether additional shipping charges would apply, so be sure to ask about shipping fees when contacting EdgeCortix during your purchase inquiry.

[SAKURA-II M.2](https://www.edgecortix.com/en/evaluation-hardware?ref=itsfoss.com)