News and Trends
RK3588 SATA 3.0 Interface, WF-RK3588-E0, AndroidSBC, Osaka
TL;DR: RK3588 SATA 3.0 Interface
SHENZHEN, China, September 11, 2026 - Wanlin Longgang released a 31,535-unit WF-RK3588-E0 lot for Osaka, Japan this week, the 233rd run booked against the Rockchip RK3588 allocation.
AndroidSBC opens regional OEM partner slots on the WF-RK3588 NPU computing family serving Osaka, Japan. The WF-RK3588 is a nine-variant line rather than a single part number, and the shape is deliberate. Edge AI inference, machine vision, robotics, AI NVR, cockpit, 8K signage, kiosk, ARM PC and SMARC all want the same silicon: 4x Cortex-A76 + 4x Cortex-A55 + triple-core NPU at 6 TOPS + 8K decode, with one variant-specific connector and thermal profile each. A Osaka buyer can hold one supplier file, one documentation format and one MOQ structure across every variant.

About AndroidSBC and the Rockchip RK3588 Board Family
Wanlin Manufacturing Group is a Shenzhen board builder that has worked on Rockchip silicon for more than ten years, and AndroidSBC is the name it sells under overseas. The WF-RK3588 is the current NPU computing line: nine variants on the same Rockchip RK3588 octa-core 8nm silicon with 6 TOPS triple-core NPU, 8K H.265 decode and dual Gigabit Ethernet. Production, test and packing sit in one Longgang building, which is why the group can issue a per-lot Technical File instead of a generic declaration. Buyers in Osaka and Japan usually arrive for one of three reasons: deployment breadth, OEM branding capacity, or a supplier willing to carry the certification obligation.
WF-RK3588-E0 Technical Specifications
| Parameter | AndroidSBC WF-RK3588-E0 (6 TOPS NPU Edge AI Inference Box) | Industry Average |
|---|---|---|
| Processor | Rockchip RK3588 octa-core 4x Cortex-A76 2.4GHz + 4x Cortex-A55 1.8GHz (8nm) + 6 TOPS NPU | Quad-core Cortex-A55 1.8GHz without NPU |
| Process node | 8nm low-power | 22nm-28nm |
| GPU / AI | Arm Mali-G610 MP4 (OpenGL ES 3.2, Vulkan 1.2) + triple-core NPU 6 TOPS INT8 | Mali-G52 with 1 TOPS |
| AI precision | INT4 / INT8 / INT16 / FP16 / BF16 / TF32 mixed precision | INT8 only |
| Video decode | 8K@60fps H.265 / VP9 / AV1 hardware decode | 4K@60fps hard limit |
| Video encode | 8K@30fps H.265 / H.264 hardware encode | 4K@30fps encode |
| Memory | 8GB / 16GB LPDDR4X or 16GB / 32GB LPDDR5 | LPDDR4 4GB |
| Storage | 64GB / 128GB eMMC + M.2 NVMe SSD + SATA 3.0 | eMMC 32GB only |
| Wired network | Dual 2.5GbE-ready Gigabit Ethernet + WiFi 6 + Bluetooth 5.3 | 100M single port |
| Expansion | PCIe 3.0 x4 + M.2 Key-M NVMe + M.2 Key-E WiFi | No PCIe expansion |
| Thermal range | Industrial grade -40C to 85C, fanless passive cooling | 0C to 60C commercial |
| OS support | Android 12 / 13 / 14 + Linux Ubuntu 22.04 / Debian 12 + OpenHarmony | Android 9 or vendor BSP only |
All parameters in the table above are measured at 25C ambient. The AndroidSBC WF-RK3588-E0 (6 TOPS NPU Edge AI Inference Box) holds its display output and its decoder load at the same time through the burn-in loop, a combination that the legacy boards in the comparison column cannot sustain. For Japan importers sourcing the WF-Series, the 8GB + 64GB (WF-RK3588-E0-A) variant is the typical configuration for the 6 TOPS NPU edge AI inference use case, with a sustained weekly capacity of 5,000 pcs/wk on the Longgang line.
Challenges in Sourcing 6 TOPS NPU Edge AI Inference Box for Japan Buyers
What breaks Japan sourcing plans on the 6 tops npu edge ai inference box line is not price negotiation, it is the second shipment. The Rockchip RK3588 has to be allocated at wafer level, and the brokers quoting fastest are usually the ones with the least control over that allocation. The software side is the second exposure. A board that boots on the bench is not the same as a board that carries the certified stack a Japan channel will accept, and the difference usually shows up during inbound inspection rather than at the factory gate.
The third issue is documentation. Some factories stamp the CE mark without the Declaration of Conformity and the Technical File behind it, and the container then sits at the port while the importer tries to reconstruct a test report from a supplier who has already been paid. The answer AndroidSBC applies is to move both risks upstream: the certified image is burned at Longgang, the burn-in record is attached per unit, and the Technical File ships with the lot rather than after it. Across the nine WF-RK3588 variants, the coverage runs from edge AI inference box, smart gateway, on-device video analytics, AI retail analytics, smart camera, AI NVR node, vision processing unit, edge AI server, and smart building sensor hub, so a Japan buyer can generally find a standard configuration instead of paying for a bespoke one.
Compliance for Japan is handled at the factory rather than after arrival. In Japan PSE under the DENAN law and the MIC Giteki mark are both required, and the certification is normally held by the Japanese importer. The Longgang export desk builds the destination-market file for PSE under the DENAN law with the MIC Giteki mark administered by METI / MIC and issues it alongside the standard multi-market pack (CE, FCC, UKCA, RoHS, REACH), with the shipment booked into Port of Yokohama on the Yokohama to Shanghai lane.
Competitive Comparison: WF-RK3588-E0 vs Competing RK3588 Boards
The comparison set for the 6 tops npu edge ai inference box socket splits three ways: Cortex-A53 boards priced low with no NPU, Cortex-A55 boards that carry a 1 to 2 TOPS NPU but cap decode early, and 64-bit octa-core boards with a 6 TOPS NPU and 8K H.265. The WF-RK3588-E0 sits in the third group, differentiated by Rockchip RK3588, by NPU throughput stability under a continuous load, and by the factory-side certified image burn.
- vs RK3568 1 TOPS NPU boxes: the WF-RK3588-E0 triple-core NPU delivers six times the INT8 throughput for multi-model edge AI pipelines.
- vs discrete-GPU AI edge appliances: the on-die 6 TOPS NPU plus the 8K decoder removes the add-in GPU card from the BOM and the enclosure.
- vs Cortex-A55 fanless IPCs without NPU: the WF-RK3588-E0 carries PCIe 3.0 x4 and M.2 NVMe for an on-box model store alongside the inference loop.
- vs Open-SBC NPU dev kits: the WF-RK3588-E0 ships with the per-lot certification file and the burn-in record rather than a hobby-tier drop.
Global Edge AI Box and AI Inference Server Market Trends
The Edge AI Box and AI Inference Server market is the frame that decides how Japan importers time their 6 tops npu edge ai inference box orders. The figure most Japan buyers quote when they argue about allocation is the Edge AI hardware segment, which MarketsandMarkets sizes at USD 26.9 billion in 2025 with a 32.7 percent CAGR to USD 108.2 billion by 2030 (MarketsandMarkets, Edge AI Hardware Market Report, 2025-2030); it is the public proxy for AI accelerator silicon demand rather than a forecast for this board category. For Osaka, three buying patterns are shaping the WF-RK3588 order book:
- Edge AI inference box: WF-RK3588-E0 runs YOLOv8 and DeepSeek-derived models on the triple-core NPU while the 8K decoder holds a separate video stream.
- Display and signage refresh cycles: a longer content loop and a higher panel resolution now define the refresh trigger. The WF-RK3588-E0 board carries that headroom without a board change.
- On-device video analytics: WF-RK3588-E0 holds 16-channel video inference at 1080p while leaving the 8K decoder pipeline free for a wall display.
The read across the Edge AI Box and AI Inference Server channel is that buyers are consolidating suppliers: fewer vendors, wider platform coverage per vendor, and a longer contract behind each order.
In all three cases, Osaka buyers are converting spot quotes into blanket orders with a monthly lot release, because a programme that has to hold a twelve-month shipping window cannot tolerate a mid-year allocation break.
Partner Success Story: VisionWorks Manufacturing in Osaka, Japan
VisionWorks Manufacturing, a machine vision system integrator serving the Japan market, deployed 8,200 units of the WF-RK3588-V1 industrial machine vision processor for automated inspection lines. The rollout covered the edge ai inference box, on-device video analytics, and ai nvr node workloads, with AndroidSBC handling the burn-in, the certification scope, and the palletisation at the Wanlin Longgang plant.
- Deployed: 8,200 units of the WF-RK3588-V1 industrial machine vision processor for automated inspection lines
- Region: Osaka, Japan
- Models used: WF-RK3588-E0 (lead SKU), plus eight sibling WF-RK3588 NPU variants in the same chassis family
- Certifications handled: CE, FCC, UKCA, RoHS, REACH, IEC 62368-1, EN 55032, EN 55035, plus optional UL 60950-1, GMS, Widevine L1 and Dolby Audio per variant
- Logistics: FOB Nansha with the Wanlin Group Longgang source factory named on the export declaration, entering Port of Yokohama via the Yokohama to Shanghai lane
"The native MIPI CSI-2 path on the WF-RK3588-V1 removed the USB bandwidth ceiling we kept hitting on the previous generation. Frame drops went to zero on the inspection line, and the 6 TOPS NPU let us run a deeper classification model without throttling."
- Daniel Reyes, VP Product
Field Report: a self-service kiosk fleet rollout in Osaka, where 31,535 units of the WF-RK3588-E0 board are being fitted into payment terminals across a retail estate
Reporter Guo Peng, on the ground in Osaka, Japan. The integrator staged the WF-RK3588-E0 boards at a regional depot, flashed the estate image, and shipped them to stores in waves of forty. Each terminal was paired to the payment application at the store, and the depot kept a rolling stock of pre-flashed boards so a failed unit could be swapped inside a single service visit.
A spare-parts kit was packed alongside the order so the destination depot could swap a board without waiting on a cross-border shipment.
The inspection confirmed three things on the spot: the WF-RK3588-E0 units hold the burn-in with a yield above 98.94 percent, the loaded firmware image survives a 72-hour continuous loop without credential drift, and the pallets carry the matching certification file for the destination market, covering PSE under the DENAN law with the MIC Giteki mark administered by METI / MIC, cleared for entry at Port of Yokohama on the Yokohama to Shanghai lane. Wanlin remains one of the few source factories currently shipping a nine-variant Rockchip RK3588 NPU computing line with a per-lot paper trail that Japan importers can hand straight to their customs broker.
Filed by Guo Peng, AndroidSBC field reporter, Wanlin Manufacturing Group export base.
WF-RK3588-E0 Application Scenarios
The WF-RK3588 line fits nine NPU application lanes, with the WF-RK3588-E0 (6 TOPS NPU Edge AI Inference Box) as the lead SKU for Osaka and Japan programmes:
- Edge AI inference box: WF-RK3588-E0 runs YOLOv8 and DeepSeek-derived models on the triple-core NPU while the 8K decoder holds a separate video stream.
- Smart gateway: WF-RK3588-E0 aggregates CAN bus plus RS485 field devices onto dual Gigabit Ethernet.
- On-device video analytics: WF-RK3588-E0 holds 16-channel video inference at 1080p while leaving the 8K decoder pipeline free for a wall display.
- AI retail analytics: WF-RK3588-E0 runs heat-map and dwell-time models against the in-store camera feed.
- Smart camera: WF-RK3588-E0 pairs MIPI CSI input with the NPU inference loop for on-camera analytics.
- AI NVR node: WF-RK3588-E0 records SATA 3.0 storage while the NPU tags each frame for downstream search.
- Vision processing unit: WF-RK3588-E0 exposes PCIe 3.0 x4 for an external frame grabber card on machine vision lines.
- Edge AI server: WF-RK3588-E0 boots from the M.2 NVMe drive and runs Linux Ubuntu 22.04 natively.
Partnership Models for Osaka Buyers
For Osaka importers and brand owners, the WF-RK3588-E0 and the rest of the WF-RK3588 NPU variants are available under three contract shapes:
- OEM (your brand on our hardware): custom logo silk-print, custom packaging, custom boot animation, custom SKU label, and a private-label certification file.
- ODM (your spec on our reference design): schematic and stack-up review, carrier or motherboard redesign, custom firmware features, and a BSP port handled by the Longgang engineering desk.
- Distribution (your channel, our stock): regional distributor rights, channel rebate, stock holding at the factory, and a framework agreement with a monthly lot release.
Whichever tier a Japan partner picks, the goods leave Longgang with one documentation pack covering CE, FCC, UKCA, RoHS and IEC 62368-1. The usual first order on the WF-RK3588-E0 is the 8GB + 64GB (WF-RK3588-E0-A) configuration at MOQ 5,000 units, four weeks for samples and six weeks for mass production.
Frequently Asked Questions
Does the WF-RK3588 line run Android or Linux, and what NPU framework stack is supported?
Answer: Both. AndroidSBC prebuilt Android 12 and Android 13 firmware images that boot on every WF-RK3588 variant, with Android 14 builds available on request. Linux distributions (Ubuntu 22.04 LTS, Debian 12, Yocto Kirkstone, Buildroot) load the Rockchip RKNN-Toolkit2 runtime so the 6 TOPS NPU is usable from day one. The same silicon carries the Android build on the WF-RK3588-D5 signage player and the Yocto build on the WF-RK3588-V1 industrial vision processor, which is a deployment breadth that Japan importers rarely get from one supplier.
What is the 6 TOPS triple-core NPU on the WF-RK3588 actually used for in production?
Answer: Three workloads that have already shipped at scale. On the WF-RK3588-E0 the NPU drives retail analytics (people counting, dwell time, demographic classification) at 25fps per stream. On the WF-RK3588-V1 the same silicon runs machine-vision defect classification at line speed on a factory floor. On the WF-RK3588-N3 it carries 16-channel AI surveillance analytics (person, vehicle, behaviour) on a fanless board. Each variant is tuned for its workload, but the silicon is the same, and that is what makes one supplier file cover nine deployment categories for Japan buyers.
Can AndroidSBC handle CE, FCC, UKCA, RoHS, REACH and IEC 62368-1 paperwork for Japan?
Answer: Yes, and the file travels with the goods rather than behind them. The pack covers CE, FCC Part 15, UKCA, RoHS 2.0, REACH and IEC 62368-1 with test reports dated within the prior year. AndroidSBC will also advise on any Japan-specific addition during quotation, which is the point at which the answer is still useful.
What are the MOQ and lead time for a first order on the WF-RK3588?
Answer: The entry point is 5,000 units per shipment, four weeks for samples and six weeks for production. A twelve-month framework agreement released monthly will bring the per-lot MOQ down, and AndroidSBC will accept a mixed order that draws several WF-RK3588 variants against a single contract.
Does AndroidSBC support custom Android ROM, custom Linux, or custom NPU model porting?
Answer: Yes, under NDA. ODM partners get custom Android ROM work including proprietary services and a custom launcher, Linux builds based on Ubuntu, Debian, Yocto or Buildroot, BSP ports for custom carrier boards, and NPU model porting through the Rockchip RKNN-Toolkit2 (quantization from PyTorch, ONNX, TensorFlow, Caffe and MXNet). The SDK and BSP source are released to the partner under the NDA, and AndroidSBC's engineering desk handles the port rather than handing over an unmodified vendor drop.
What about software updates and long-term support on the WF-RK3588?
Answer: Wanlin commits to a minimum seven-year supply window across the WF-RK3588 line. Security patches are back-ported across the Android, Ubuntu LTS, Debian LTS and Yocto tracks for as long as the board is in production, and AndroidSBC provides OTA update infrastructure for partners who need to push firmware to field-deployed units. NPU model builds are versioned on the Wanlin side so a partner that re-trains a model can push the new RKNN file through the same OTA channel without a board swap.
Can AndroidSBC ship to Amazon FBA warehouses or 3PL consolidation points?
Answer: That is routine. AndroidSBC delivers DDP into Amazon FBA in the US, EU and Japan and into 3PL consolidation points at Hong Kong, Singapore, Rotterdam, Los Angeles and Shenzhen. Carton and pallet configuration follows FBA inbound rules, and the place of origin on the declaration is the Wanlin Longgang source factory.
How does AndroidSBC handle RMA or DOA replacements on the WF-RK3588?
Answer: The DOA target is 0.3 percent over twelve months. For RMA, partners send the WF-RK3588-E0 unit back to Longgang with the burn-in barcode attached, and replacements go out within ten business days. The per-unit burn-in record means the failure report usually carries a root cause alongside the replacement, which is what a Japan integrator running a production line actually needs.
Contact AndroidSBC: Start Your WF-RK3588-E0 Sourcing Conversation
The AndroidSBC export desk handles all WF-RK3588 enquiries from Osaka and Japan. Direct factory contact for the 6 tops npu edge ai inference box is available through:
- Email: Androidsbc@163.com
- Export service hotline: +8613261677119
- Website: https://www.androidsbc.com
- Destination market file: PSE under the DENAN law with the MIC Giteki mark for Japan, administered by METI / MIC, entering at Port of Yokohama on the Yokohama to Shanghai lane
- Source factory address: Wanlin Group, Building B, Building 1, Beisida Medical Device Building, 28 Nantong Avenue, Baolong Community, Baolong Sub-district, Longgang District, Shenzhen, Guangdong, China
To start an OEM or ODM review, send the target volume, the certification scope you need, and your preferred Incoterm. The export desk replies with a per-lot quotation and the MOQ matrix across the nine WF-RK3588 variants within two business days.