AI-Native Tools for Builders. Create. Debug. Explore...
shenzhen,China
Joined September 2018
- Tweets6K
- Following250
- Followers29.1K
- Likes4.1K
Sipeed retweeted
10 Gbps USB 3.2 logic analyzer with 32 channels.
cnx-software.com/2026/10/09/…
The @SipeedLab SLogic32U3 USB-C logic analyzer supports up to 32 channels via four mini HDMI connectors, 350 MHz signal bandwidth, up to 1400 MS/s per channel with four channels, or 200 MS/s per channel with 32 channels, and is offered with an optional ADC module.
The logic analyzer works on Windows 10/11 64-bit, Linux 64-bit x86, macOS (Apple Silicon), and Android (browser). Most people will likely use the SLogicView software developed by Sipeed, but it is also compatible with ngscopeclient, Sigrok/Pulseview, SLogicWeb, and the ALL LOGIC community app.
Sipeed retweeted
4K USB-C KVM with AI integration, support for Microsoft Recall-like feature.
cnx-software.com/2026/07/01/…
@SipeedIO NanoKVM-Go/Go+ works with any device featuring a USB-C port with DisplayPort Alt mode, including several Android smartphones, iPhone 15+, mini PCs, laptops, and so on. Only one cable is required to easily enable hardware-level remote access.
Both variants are AI-Native KVM with all KVM functions exposed as an MCP server, and the Go+ model, featuring more memory and storage, enables "Ambient Screen Intelligence" to allow people to search for screenshot history.
Sipeed retweeted
(near) Instant dumping of the Bitlocker VMK using @SipeedIO #SLogic16U3 and #ngscopeclient 🥰. Full disclosure: i know nothing about C++, filter was fully vibe coded (with a Claude free plan)
Sipeed retweeted
Introducing z386: an 80386-class FPGA CPU built around the original 386 microcode.
It’s more compact than ao486 and useful for studying how the 386 really worked.
Also: z386-MiSTer, an experimental PC MiSTer core built on z386.
github.com/nand2mario/z386
github.com/nand2mario/z386_M…
Sipeed retweeted
First part of SpacemiT K3 RISC-V Pico-ITX Chassis Kit.
cnx-software.com/2026/05/23/…
In this review, we check the hardware of the 16-core RISC-V mini PC with an unboxing and a teardown, before booting to the Ubuntu-based Bianbu 4.0 OS and checking basic features and system information.
The second part of the review will test features and performance in more detail.
Sipeed retweeted
Good progress on #386fpgacore last month 🚀
Now running on Tang Console 138K + DE10-Nano (MiSTer)
+I/D caches, lots of fixes
Doom runs
75MHz with decent CPI
3DBench: 24.1 FPS
Possibly the fastest 386 ever 🙂
Sipeed retweeted
Online guide to select the best hardware for local LLM/AI deployments.
cnx-software.com/2026/03/30/…
The website relies on Qwen 3.5 models and shows the price, performance, power consumption, and more for a range of hardware from Raspberry Pi 5 16GB to NVIDIA DGX Spark.
In one specific test (Qwen 3.5 9B), a system with a $260 Intel Arc graphics card performs just as well as $4,000 computers. The project also accepts user submissions, but the process is a little convoluted.
Sipeed retweeted
MisterClaw: World's first @SipeedIO PicoClaw AI assistant on MiSTer FPGA (DE10-Nano) #MisterFPGA #OpenClaw #PicoClaw #RetroGaming #MisterClaw
github.com/wolfshow/MisterCl…
Sipeed retweeted
It works 🚀
First demo of the new #386fpgacore running on real FPGA hardware (Sipeed Tang Console 138K).
VGA output, 3DBench, Norton Commander, and Turbo C all running.
Still slow and buggy — but a lot of programs already work.
Sipeed retweeted
[Statement] I've recently discovered impersonators on X using my identity.
This is my only X account. Any other accounts claiming to be me are fake
— please do not trust them.
I'll be sharing my thoughts and insights from the #PicoClaw development process here. Follow along!
Sipeed retweeted
Norton Commander running on the new 386 FPGA core 🧭 2.4 MHz sim speed...
Interrupts, disk I/O, and VGA text mode are all up and running ✅
Now working on JemmEx extended memory manager - some protected mode corner cases remain. After that is actual FPGA. #386fpgacore
Sipeed retweeted
Milestone 🚩 for the #386fpgacore:
Protected mode is now mostly implemented — real mode was completed end of January.
Still ahead:
• Hardware interrupts
• FPGA validation
• Cleanup + optimization
• Large programs
• Doom 😈
Test suite: github.com/barotto/test386.a… (~1B cycles😀)
Sipeed retweeted
i built a local LLM inference engine that runs a 1B parameter model on a $10 board with 256mb ram. model sits on the sd card, streams one layer at a time through 45mb of ram
You can use it as local LLM model backend for PicoClaw
no python no cloud no api keys
80kb binary | pure c | zero dependencies
github: github.com/RightNow-AI/picol…
Sipeed retweeted
Slowly making my way 🏃 through protected mode on the 386 FPGA core.
Built a single-step test suite with 86Box as the model.
Protection test unit is done (a fun one!) — segment loading + paging mostly working (~80% pass).
Plenty more edge cases and new functionality to tackle 🚀
Sipeed retweeted
🎉 100% pass on SingleStepTest/80386, 80186_tests.zip and test386.asm (real mode) .
This 386 FPGA core is coming together faster than I expected. Next: verify on real FPGA. Then the hard part begins: protected mode. Many PM mirco-ops are poorly understood... expect pain 😅
Sipeed retweeted
Compact modular AI camera with performance similar to Jetson Nano Orin 4GB (33 TOPS).
cnx-software.com/2026/01/27/…
@SipeedIO MaixCAM2 is built around an Axera AX630 AI SoC with 3.2 TOPS (INT8) to 12.8 TOPS (INT4) of AI performance. It comes with 1GB or 4GB RAM, a 32GB eMMC flash, and a microSD card slot for storage, a 2.4-inch touchscreen display, and an 8MP / 4Kp30 camera sensor with LED flash, as well as two microphones and a speaker, making it suitable for local LLM and VLM applications.
It also features two PMOD connectors for expansion, such as a thermal imager and a Time of Flight (ToF) ranging sensor. It ships with over 40 AI samples for evaluation, and can be programmed in Python using the MaxiPy framework.
