iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes. The Sixfab AI HAT+ is a Raspberry Pi 5-only PCIe accelerator that adds a DEEPX neural-processing unit (NPU) for local vision inference. You can buy it with 13 TOPS (DX-M1ML) or 25 TOPS (DX-M1M) of INT8 performance. The board needs both the Pi 5’s 40-pin power connection and its PCIe FFC link, plus a 27 W Pi power supply and active cooling for a complete build.
What the Sixfab AI HAT+ does
The Sixfab AI HAT+ is a HAT+ specification-compliant accelerator for Raspberry Pi 5. It places a DEEPX NPU on a 65 × 56.5 mm board and connects to the Pi over PCIe Gen 3 x1 with the supplied 16-pin FFC cable. Inference runs locally on the NPU rather than in a cloud service or on a separate desktop GPU. Sixfab positions it for real-time computer vision, including camera analytics and object detection.
There are two hardware versions. The DX-M1ML is rated at 13 TOPS (INT8) and has 1 GB of LPDDR4x accelerator memory. The DX-M1M is rated at 25 TOPS (INT8) and has 2 GB of LPDDR4x. Both use the same driver and DXNN software-development kit, so an application can move between the chips without code changes; the practical difference is available NPU memory and throughput.
Recommended Free Tools
Sixfab AI HAT+ specifications
| Specification | DX-M1ML version | DX-M1M version |
|---|---|---|
| Rated NPU performance | 13 TOPS at INT8 | 25 TOPS at INT8 |
| Dedicated accelerator memory | 1 GB LPDDR4x | 2 GB LPDDR4x |
| Published sustained NPU power | Not stated in the cited Sixfab documentation | 3 W sustained |
| MSRP | $63 | $90 |
| Board size | 65 × 56.5 mm | |
| Operating temperature | 0–70 °C commercial range | |
| Host interface | PCIe Gen 3 x1 through a 16-pin FFC cable | |
The TOPS figures are INT8 ratings, not a promise that every model or camera pipeline will reach that rate. Actual throughput depends on the network, input resolution, preprocessing, post-processing, camera drivers and the number of simultaneous streams.
#1 Best Overall
- Read Before You Buy — No Video Output: These adapters support charging and USB 2.0 data transfer, but cannot transmit video signals. Except for standard USB webcams (which use USB data only), they are not compatible with HDMI/DisplayPort cables, video-capable USB-C hubs, or docking stations with video output.
- Convert USB-A Ports to USB-C: Designed to connect USB-C earphones, cables, flash drives, card readers, and other USB-C accessories to standard USB-A ports. Plug-and-play with no drivers or software required.
- Aluminum Alloy Housing: Built with a sturdy aluminum alloy shell that aids in heat dissipation and protects against daily wear and scratches. Designed to maintain a stable and secure connection.
- Compact & Travel-Friendly: The ultra-compact design allows the adapter to stay plugged into your device without blocking adjacent ports or adding bulk, reducing wear and tear on your original USB ports.
- 12-Month Warranty: Backed by a 12-month manufacturer warranty for peace of mind. Designed to meet strict quality control standards for reliable everyday performance.
Raspberry Pi 5 compatibility and parts you need
Raspberry Pi 5 is the only supported host platform. Raspberry Pi 4, Compute Module 4, Compute Module 5 on the CM5 IO Board and non-Raspberry-Pi single-board computers are explicitly unsupported. The HAT+ label therefore does not mean that it will work with any 40-pin Raspberry Pi.
Required hardware
- Raspberry Pi 5.
- Sixfab AI HAT+, in either DX-M1ML or DX-M1M form.
- The included 16-pin FFC cable for the Pi 5 PCIe connection.
- A Raspberry Pi 27 W USB-C power supply (5 V/5 A is the recommended specification).
- A Raspberry Pi 5 Active Cooler. The Pi, power supply and cooler are sold separately from the Sixfab board.
- A microSD card or other Pi 5 boot medium, camera(s), and enclosure or mounting hardware if your deployment needs them.
Both electrical paths matter: the HAT+ uses the Pi’s 40-pin GPIO connection for power and the FFC cable for the PCIe data link. Connecting only one of them will not produce a working accelerator. Turn the Pi off and disconnect its power before fitting or removing the HAT and FFC cable; an incorrectly seated flex cable can look like a software failure.
Which Sixfab version should you buy?
Choose DX-M1ML for a cost-sensitive single stream
The $63 DX-M1ML provides 13 TOPS and 1 GB of LPDDR4x. It is the sensible choice for one camera, a compact detector or classifier, and deployments where the model fits comfortably in that memory budget. Its lower price can matter when several Pi-based endpoints are being installed. Sixfab describes the driver and SDK as shared with the DX-M1M, so selecting this chip does not require a separate application codebase.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchRank #2
- 5-in-1 USB-C Hub: Experience comprehensive connectivity featuring a Power Delivery input, two USB-A 2.0 ports, a USB-A 3.0 port, and an HDMI port. (Note: The USB-C power delivery input port is only for connecting an external wall charger to power your laptop and cannot power peripheral devices.)
- 90W Pass-Through Charging: Achieve optimal charging with 90W pass-through power to your laptop, supported by a total input of 100W, with the hub reserving 10W for operational efficiency. (Note: Wall charger not included.)
- Quick Data Transfers: Accelerate your productivity with rapid data transfers using a high-speed 5Gbps USB 3.0 port and two 480Mbps USB 2.0 ports.
- 4K HDMI Display: Enhance your visual experience with a hub capable of delivering 4K resolution at 30Hz in both mirror and extend modes. Please note that this hub is compatible with MacBook (macOS 12 and newer), Windows 10 and 11, ChromeOS, and laptops equipped with DP Alt Mode and Power Delivery. Note: This device is not compatible with Linux.
- What You Get: Anker USB-C Hub (5-in-1, 4K HDMI), welcome guide, 18-month warranty, and our friendly customer service.
Choose DX-M1M for larger models or more cameras
The $90 DX-M1M doubles the stated TOPS rating to 25 and doubles accelerator memory to 2 GB. Select it when the network is larger, when you need more headroom for input queues and post-processing, or when one Pi must handle multiple feeds. Sixfab reports up to four 1080p streams at 25–30 frames per second in its stated DX-M1M methodology. That is a vendor test condition, not a universal guarantee: your camera codec, model and pipeline may produce a lower rate.
| Decision | Better fit | Reason |
|---|---|---|
| One 1080p stream and lowest board cost | DX-M1ML | 13 TOPS and 1 GB are aimed at a smaller, single-stream deployment. |
| Several 1080p streams | DX-M1M | Higher throughput and 2 GB memory provide more concurrency headroom. |
| Model may exceed 1 GB or needs larger buffers | DX-M1M | Its 2 GB LPDDR4x pool reduces memory pressure. |
| Switching hardware after software development | Either | Sixfab says the driver and SDK are shared and application code does not need changes. |
How to assemble and configure it
- Prepare the Pi. Use a Raspberry Pi 5 with Raspberry Pi OS Trixie, 64-bit, and update the operating system before adding the accelerator. Shut down fully and unplug USB-C power.
- Install cooling. Fit the Raspberry Pi 5 Active Cooler so the host CPU can sustain workloads. The Sixfab board is specified for a 0–70 °C commercial operating range; that does not remove the Pi 5’s own cooling requirement.
- Mount the HAT+. Align the board with the Pi 5’s 40-pin header and secure it without forcing the connector. The 40-pin connection is required for the HAT’s power path.
- Connect PCIe. Insert the supplied 16-pin FFC cable in the Pi 5 PCIe connector and the HAT+ connector, observing the contacts and latch orientation. The accelerator communicates over PCIe Gen 3 x1.
- Apply the recommended power. Use the 27 W Raspberry Pi supply rated at 5 V/5 A. A marginal supply can cause resets or peripheral dropouts when cameras and the NPU are active.
- Install the runtime. In Raspberry Pi OS Trixie 64-bit, install Sixfab’s documented
dxrt-runtimepackage and the C++ or Python DXNN SDK. The runtime exposes the DEEPX device to applications. - Convert a model. The DXNN compiler accepts ONNX models exported from PyTorch, TensorFlow, Keras or Ultralytics. Export the network to ONNX, compile it to the DXNN format, and use the resulting file from your C++ or Python application.
- Run a known-good test. Sixfab’s basic validation setup uses a Raspberry Pi 5, the AI HAT+ and a YOLOv8n DXNN file. Confirm that this small detector runs before adding higher-resolution cameras, tracking or several streams.
The exact package commands and compiler options can change with the SDK release, so use the instructions shipped with the version of dxrt-runtime and DXNN SDK you install rather than copying options from an older tutorial.
Running YOLO and multi-camera vision locally
YOLO models are a supported use case when they are exported through the ONNX-to-DXNN workflow. A practical first deployment is YOLOv8n on one camera: capture frames on the Pi, send the tensor inference to the DEEPX NPU, then perform confidence filtering, non-maximum suppression and application logic on the host. Keeping the camera and inference on the Pi avoids sending video to a cloud service.
Rank #3
- Sleek 7-in-1 USB-C Hub: Features an HDMI port, two USB-A 3.0 ports, and a USB-C data port, each providing 5Gbps transfer speeds. It also includes a USB-C PD input port for charging up to 100W and dual SD and TF card slots, all in a compact design.
- Flawless 4K@60Hz Video with HDMI: Delivers exceptional clarity and smoothness with its 4K@60Hz HDMI port, making it ideal for high-definition presentations and entertainment. (Note: Only the HDMI port supports video projection; the USB-C port is for data transfer only.)
- Double Up on Efficiency: The two USB-A 3.0 ports and a USB-C port support a fast 5Gbps data rate, significantly boosting your transfer speeds and improving productivity.
- Fast and Reliable 85W Charging: Offers high-capacity, speedy charging for laptops up to 85W, so you spend less time tethered to an outlet and more time being productive.
- What You Get: Anker USB-C Hub (7-in-1), welcome guide, 18-month warranty, and our friendly customer service.
For multiple cameras, plan the pipeline rather than multiplying a single-stream frame rate. Each stream consumes camera bandwidth, decode time, host memory and NPU scheduling capacity. The DX-M1M’s published methodology reaches up to four 1080p streams at 25–30 FPS, but the result is tied to Sixfab’s model and test setup. Use lower input sizes, staggered inference or a smaller model if your application cannot sustain the target rate. The 13-TOPS DX-M1ML is better treated as a single-stream part unless your own measurements show otherwise.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Measure end-to-end latency, not only NPU inference time.
- Record dropped frames and queue depth for every camera.
- Check whether decoding and resizing, rather than inference, are the bottleneck.
- Leave memory headroom for tracking, recording and your user interface.
Power, cooling and operating limits
Sixfab documents 3 W sustained power for the DX-M1M NPU. That figure describes the accelerator and should not be read as the total consumption of a Pi 5 with cameras, storage, USB devices and networking. The recommended 27 W Pi supply gives the complete system more margin than a generic low-power adapter.
The board’s stated commercial operating range is 0–70 °C. Use airflow or an enclosure designed for ventilation, keep the Pi 5 Active Cooler installed, and monitor temperatures during the longest expected workload rather than during an idle test. Thermal throttling, brownouts and an FFC cable that is not fully latched can all present as intermittent inference failures.
Rank #4
- Dual Converters, Infinite Potential：Includes 2× USB C male to USB A female adapters and 2× USB A male to USB C female adapters. Perfect for a wide range of uses—tablets with Bluetooth keyboards, expand USB ports on macbook, and more. Two different converters for all your daily needs
- Next-Level 10Gbps & 3A Charging: No more slow 480Mbps, this usb to usb c adapter has a transfer speed of up to 10Gbps, allowing you to do more transferring in less time. This usb adapter fits both USB A and USB C charger, supporting up to 3A fast charging
- Upgraded Exquisite Craftsmanship: With an aluminum alloy housing and metal connector, the usbc to usb adapter is extremely durable and sturdy. Rigorously tested to withstand more than 10,000 times of plugging and unplugging, ensuring long-lasting performance
- Broad Compatible: The usb c to usb adapter widely supports all USB C/ USB A devices like laptops, tablets, cellphones, car chargers, and phone chargers. Such as compatible with MacBook Pro/Air 2023/2022, Thunderbolt 4/3 Devices,Apple MagSafe Watch 9/8/7/SE/Ultra, iPad Pro 2022/2021, Samsung Galaxy S23/S20/S10, and iPhone 17/16/15 Pro. Plug and play
- Please Note: To reach 10Gbps speed, keep the cable under 3.3 ft. For USB A Male to USB C adapters, try flipping the USB C connector. USB C Male to USB A adapters support bidirectional 10Gbps transfer within 3.3 ft
Sixfab versus Raspberry Pi AI HAT+
Raspberry Pi also sells an AI HAT+ family based on Hailo accelerators. The right comparison is not TOPS alone: the NPU architecture, SDK, model-conversion path, memory, power behavior and camera software integration determine how much work your application requires.
| Factor | Sixfab AI HAT+ | Raspberry Pi AI HAT+ |
|---|---|---|
| Accelerator | DEEPX DX-M1ML or DX-M1M; applications use the DXNN stack. | Hailo-8L or Hailo-8; applications use Hailo’s software stack. |
| Published performance options | 13 TOPS or 25 TOPS, both at INT8. | 13 TOPS for Hailo-8L or 26 TOPS for Hailo-8. |
| Dedicated memory | 1 GB LPDDR4x on DX-M1ML; 2 GB on DX-M1M. | Memory capacity is not stated in the supplied comparison facts. |
| Host | Raspberry Pi 5 only; PCIe Gen 3 x1 and a 16-pin FFC cable. | Designed for Raspberry Pi 5 integration; verify the current Hailo accessory and software requirements for your image. |
| Model workflow | Export from PyTorch, TensorFlow, Keras or Ultralytics to ONNX, then compile to DXNN. | Uses Hailo’s conversion and runtime tools rather than DXNN. |
| Best fit | Projects already targeting DEEPX/DXNN, or deployments needing the DX-M1M’s stated multi-stream headroom. | Projects standardized on Hailo tooling or Raspberry Pi’s Hailo-supported camera ecosystem. |
Because the two products use different compilers and runtimes, a model that runs on one is not automatically deployable on the other without conversion and validation. Compare the model families you actually intend to use, not just the headline 13-, 25- or 26-TOPS number.
Free tools Windows power users keep installed
One-click scans. No signup required.
Price and buying checklist
Sixfab lists a $63 MSRP for the DX-M1ML and $90 for the DX-M1M. Those are product MSRPs, not a complete system price. Add the Raspberry Pi 5, 27 W USB-C supply, Active Cooler, storage, cameras and any enclosure. Regional stock, tax and shipping can change the amount you pay.
Best Value
- 5-in-1 Connectivity: Equipped with a 4K HDMI port, a 5 Gbps USB-C data port, two 5 Gbps USB-A ports, and a USB C 100W PD-IN port. Note: The USB C 100W PD-IN port supports only charging and does not support data transfer devices such as headphones or speakers.
- Powerful Pass-Through Charging: Supports up to 85W pass-through charging so you can power up your laptop while you use the hub. Note: Pass-through charging requires a charger (not included). Note: To achieve full power for iPad, we recommend using a 45W wall charger.
- Transfer Files in Seconds: Move files to and from your laptop at speeds of up to 5 Gbps via the USB-C and USB-A data ports. Note: The USB C 5Gbps Data port does not support video output.
- HD Display: Connect to the HDMI port to stream or mirror content to an external monitor in resolutions of up to 4K@30Hz. Note: The USB-C ports do not support video output.
- What You Get: Anker 332 USB-C Hub (5-in-1), welcome guide, our worry-free 18-month warranty, and friendly customer service.
- Confirm the listing names the correct board: “Sixfab AI HAT+ DX-M1ML for Raspberry Pi 5” or “Sixfab AI HAT+ DX-M1M for Raspberry Pi 5.”
- Check that the 16-pin FFC cable is included.
- Budget for the 27 W Raspberry Pi supply and Active Cooler, which are separate accessories.
- Choose DX-M1ML for a lower-cost single stream; choose DX-M1M for model memory, throughput or multiple-stream headroom.
- Verify current regional availability and warranty terms before ordering; the stated MSRPs do not establish stock in every country.
Troubleshooting the most common failures
The Pi boots but no accelerator is visible
Power down, reseat both ends of the FFC cable and verify that the HAT is fully engaged on the 40-pin header. Confirm that the host is a Raspberry Pi 5 and that the documented runtime is installed on 64-bit Raspberry Pi OS Trixie. Unsupported hosts will not become compatible through software changes.
Inference starts, then the system resets
Replace an underspecified USB-C supply with the recommended 27 W/5 V/5 A unit, disconnect unnecessary USB loads and check the Pi 5’s cooling. A system-wide power or thermal problem is different from the DX-M1M’s 3 W accelerator figure.
The model will not compile
Check that the source model was exported to ONNX and that its operators and input shapes are supported by the DXNN compiler version you installed. Start with the YOLOv8n validation file, then add layers or higher resolution incrementally. A model that exceeds the DX-M1ML’s 1 GB accelerator memory may need the DX-M1M or an optimized network.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Multiple cameras miss the target frame rate
Measure decode, resize and post-processing separately from NPU time. Reduce input resolution or model size, schedule streams deliberately and test the DX-M1M if your workload needs the vendor’s four-stream methodology. Treat 25–30 FPS as a stated test result, not a guaranteed rate for every camera and model.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
