2 Sources
[1]
Sipeed's new K3 RISC-V SBCs can run 30B-parameter LLMs at 10 tokens per second
Sipeed has launched its new K3 series Single Board Computers, powered by the RISC-V ISA. Using SpacemiT's new "Fusion Architecture" with dedicated matrix multiplication blocks, Sipeed claims these systems can run 30B LLMs locally at over 10 tokens per second. SpacemiT, a fabless Chinese semiconductor designer, is the silicon architect, while Sipeed serves as the hardware integrator. The K3 SoC is packaged with 32GB of LPDDR5-6400 memory and 10GbE networking, offering an alternative to proprietary AI hardware for enthusiasts and researchers. Prices start at $299 and go up to $629 for the 32GB flagship. At the silicon level, the K3 SoC features 8 general-purpose X100 cores, each with 4 MB of L2 cache. The company claims the X100 performs similarly to ARM's Cortex-A76 core. The K3 also offers 8 A100 AI matrix units with TCM (Tightly Coupled Memory) and supports up to 1024-bit RVV 1.0 vector processing. These deliver up to 60 TOPS of performance (format unspecified) and support BF16, FP16, FP8, INT8, and INT4 data types. Both X100 and A100 cores connect to the memory controller via a coherent interconnect bus, sharing the same memory. This is not like a traditional NPU, which usually resides separately. While not as fast as GPU cores, the K3 architecture supports zero-copy, allowing CPU and AI cores to share the same memory space. The dual 32-bit controllers support LPDDR4x-4200/LPDDR5-6400 memory, delivering up to 51GB/s of bandwidth. The SoC is rated at a TDP of 15-25W. Sipeed offers the K3 in two form factors: the CoM (Computer-on-Module) and the Pico-ITX. The K3 CoM260 is a 69.6mm x 45mm module, pin-compatible with NVIDIA's Jetson Orin Nano carrier boards. The Pico-ITX version, similar to a Raspberry Pi, measures 100mm x 86mm and features 2 USB Type-C ports (Power Delivery and Alt-DP), 1 10 GbE port, and 1 1 GbE port. The K3 officially supports Ubuntu 26.04 and ROS. In terms of AI throughput, the company claims inference speeds exceeding 10 tokens per second with a 30B-parameter model. The 32GB version should be able to fit a quantized version of Qwen 3.6 A3B 35B (~22GB), though space will be tight. Smaller MoE models, such as Gemma 4 26B A4B (~15GB), may be more suitable. The Sipeed K3 series, starting at $299 for the 8GB model, provides an accessible gateway into the RISC-V landscape. While it may not yet challenge NVIDIA's dominance in high-end GPUs, the K3 represents an important milestone as one of the first RVA23-compliant platforms with a full Ubuntu 26.04 LTS environment. The platform gives enthusiasts and researchers a practical way to explore local inference on an open instruction-set architecture.
[2]
Sipeed Crams 32GB LPDDR5 and a 60 TOPS NPU Into a Compact RISC-V Board That Hits 15 Tokens/s on Qwen-3.5 35B AI LLMs
Sipeed has unveiled its new RISC-V powered SBC platform, the K3 series, which can run up to Qwen3.5 35B AI LLMs and start at $299. Chinese manufacturer, Sipeed, has partnered with SPACEMIT to launch its brand new K3 series SBCs or Single-Board Computers (such as the Raspberry Pi). These PCs are normally aimed at AI, Edge, and Networking applications. Sipeed has leveraged SPACEMIT's Key Stone K3 AI CPUs, which are based on the RISC-V architecture, and offer some decent AI LLM capabilities for their size. Some highlights of the K3 series include: Starting with the specifications, the SPACEMIT K3 AI CPU comes with 8 X100 high-performance cores and 8 A100 AI cores. These are part of the fusion architecture, which delivers up to 130,000 DMIPS for general-purpose computing. The chip operates at a clock speed of 2.4 GHz and is equivalent to the ARM-A76 CPU in terms of performance. For the AI side, the K3 AI NPU offers up to 60 TOPS of AI compute capabilities, supporting various data types such as BF16, FP16, FP8, INT8, and INT4. The SBCs are rated to easily run up to 30B AI LLMs & have scored a 84% intelligence rating of a 235B model. As a demonstration, Sipeed states that the K3 can run Qwen-3.5 35B easily, delivering up to 15 Tokens/s. In terms of memory support, the CPU comes with LPDDR4X-4266 and LPDDR5-6400 support in up to 32 GB capacities. The total memory bandwidth is rated at up to 51 GB/s for the LPDDR5 solution. Sipeed is listing three models of the K3 in 8 GB, 16 GB, and 32 GB flavors. Coming to the design, the K3 CoM260 Kit is a 69.6mm x 45mm module that features a 260-pin SO-DIMM slot and is compatible with NVIDIA's Jetson Orin series. The platform supports Ubuntu 26.04 & ROS, with a rich set of IO capabilities. As for the K3 Pico-ITX, the board comes in a 100mm x 86mm form factor, and has rich IO support, such as: The full specifications of the Sipeed K3 series are listed below: There's a $10 US difference between the Kit and the ITX board, but you are looking at a starting price of $299-$309 for the 8 GB SBCs and up to $629-$639 for the 32 GB configurations. If you want a small AI machine that can handle up to 30B LLMs, then these definitely look enticing.
Share
Copy Link
Sipeed has launched its K3 series RISC-V-powered single board computers that can run 30B-parameter AI LLMs locally at up to 15 tokens per second. Built with SpacemiT's Fusion Architecture and featuring up to 32GB LPDDR5 memory plus a 60 TOPS NPU, the K3 offers an accessible open-source alternative to proprietary AI hardware starting at $299.
Sipeed has launched its K3 series single board computers, marking a significant step forward for RISC-V SBC platforms in AI applications. Built in partnership with SpacemiT, a fabless Chinese semiconductor designer, these compact systems can run 30B-parameter LLMs locally at speeds ranging from 10 to 15 tokens per second
2
. The K3 series starts at $299 for the 8GB model and scales up to $629 for the 32GB flagship configuration, positioning itself as an open-source alternative to proprietary AI hardware for enthusiasts and researchers exploring local inference capabilities .
Source: TweakTown
At the heart of the Sipeed K3 lies SpacemiT's Key Stone K3 SoC, featuring what the company calls "Fusion Architecture" with dedicated matrix multiplication blocks. The chip integrates 8 X100 cores for general-purpose computing, each equipped with 4MB of L2 cache and performing comparably to ARM's Cortex-A76 core . Alongside these sit 8 A100 AI matrix units with Tightly Coupled Memory, supporting up to 1024-bit RVV 1.0 vector processing . The entire system operates at 2.4 GHz and delivers up to 130,000 DMIPS for general-purpose computing
2
.The 60 TOPS NPU built into the K3 supports multiple data types including BF16, FP16, FP8, INT8, and INT4, providing flexibility for running quantized LLMs . Unlike traditional NPU designs that operate separately, both X100 cores and A100 AI matrix units connect to the memory controller via a coherent interconnect bus, enabling zero-copy operations where CPU and AI cores share the same memory space . The dual 32-bit controllers support LPDDR4x-4200 and LPDDR5-6400 memory, delivering up to 51GB/s of bandwidth . Sipeed demonstrates the platform running Qwen-3.5 35B at 15 tokens per second, with the board scoring an 84% intelligence rating of a 235B model
2
.Related Stories
Sipeed offers the K3 in two distinct form factors designed for Edge applications and networking use cases. The K3 CoM260 Kit measures 69.6mm x 45mm and features a 260-pin SO-DIMM slot, making it pin-compatible with NVIDIA's Jetson Orin Nano carrier boards
2
. The Pico-ITX version resembles a Raspberry Pi at 100mm x 86mm, featuring 2 USB Type-C ports with Power Delivery and Alt-DP support, plus 1 10GbE port and 1 1GbE port . The SoC operates at a TDP of 15-25W, making it suitable for compact deployments . Both platforms officially support Ubuntu 26.04 and ROS, providing a full LTS environment .The 32GB version can accommodate a quantized version of Qwen 3.6 A3B 35B requiring approximately 22GB, though space constraints mean smaller models like Gemma 4 26B A4B at around 15GB may prove more practical . Sipeed offers three memory configurations: 8GB, 16GB, and 32GB, with pricing ranging from $299-$309 for the entry-level 8GB models to $629-$639 for the top-tier 32GB configurations, with a $10 difference between the Kit and ITX board versions
2
. As one of the first RVA23-compliant platforms with full Ubuntu 26.04 LTS support, the K3 represents a milestone for researchers and enthusiasts seeking to explore local inference on an open instruction-set architecture . While it won't challenge NVIDIA's dominance in high-end GPUs, the platform offers a practical entry point into the RISC-V landscape for AI experimentation and development.Summarized by
Navi
1
Technology

2
Technology

3
Business and Economy
