STMicroelectronics, SpacemiT, DeepComputing, and Espressif have all shipped new processors and development platforms in recent months. Together, they span everything from a sub-dollar microcontroller to a 60-TOPS RISC-V laptop mainboard — here's what's worth knowing about each.

STMicroelectronics' new STM32C5 family targets cost-sensitive embedded designs that have surpassed the Cortex-M0+.

The chips run a 144MHz Arm Cortex-M33 core with an FPU and DSP instructions, and depending on the model, offer up to 1MB of flash, 256KB of RAM, Ethernet, USB, OctoSPI, and dual FDCAN. ST rates the C5 at roughly 593 CoreMark, about three times a typical Cortex-M0+ part. All while keeping pricing affordable (starting around $0.64 per unit at 10,000-unit quantities).

With the combination of processing power and price, the STM32C5 will fit right in for smart-home controllers, industrial sensors, motor-control systems, electronic locks, computer peripherals, and similar products that need more capability without having to use a more costly high-end MCU.

The SpacemiT K3 is a much bigger leap in capability: an RVA23-compliant RISC-V processor built for Linux, edge AI, and robotics.

It stars an eight-core RISC-V CPU (up to 2.4GHz) with a dedicated RISC-V AI engine, delivering up to 130 KDMIPS of general compute and up to 60 TOPS of AI performance. Because both the CPU and AI cores share a unified architecture and memory space, general-purpose and AI workloads run within the same Linux environment rather than requiring a separate accelerator and driver stack.

SpacemiT's K3 Pico-ITX dev board features a compact form with up to 32GB of LPDDR5 memory, onboard UFS storage, NVMe expansion, Wi-Fi 6, and 10Gb Ethernet, with more than enough I/O options.

There’s a lot of projects this board could used in including; local AI servers, autonomous robots, industrial vision systems, network appliances, and RISC-V development workstations. The hardware itself is impressive, but how well it performs outside a controlled demo will still come down to driver maturity, graphics support, and the state of AI software on RISC-V.

The K3 is also powering DeepComputing's new DC-ROMA RISC-V Mainboard III, a drop-in replacement mainboard for the Framework Laptop 13.

DeepComputing calls it the first laptop mainboard to support the RVA23 RISC-V profile. It's built around the eight-core K3, with up to 60 TOPS of AI performance, 16GB or 32GB of LPDDR5 memory, and an M.2 slot supporting both SATA and NVMe storage.

Because it drops into the modular, repairable Framework Laptop 13 chassis, developers get RISC-V in a complete portable system rather than having to build a workstation around other dev boards. DeepComputing places the DC-ROMA III primarily for developers working on Linux ports, RISC-V applications, compilers, drivers, and local AI. it's not being marketed as a mainstream x86 laptop replacement, but it's a really cool step toward practical RISC-V personal computers.

Espressif announced on July 27 that the ESP32-S31 has entered mass production and is now available for purchase, with development boards already shipping and chip/module supply expected to continually ramp through the rest of the year.

The ESP32-S31 uses two RISC-V cores running at up to 320MHz; one of them adds SIMD instructions and a 128-bit data path aimed at faster graphics, audio, signal-processing, and AI workloads. It's also Espressif's first S-series chip to move off the Xtensa architecture completely.

For the wireless connectivity options it features Wi-Fi 6, Bluetooth 5.4 (LE and Classic), Zigbee, Thread, and Matter. On the multimedia side, the chip adds camera and LCD interfaces, capacitive touch, JPEG hardware acceleration, 2D graphics acceleration, and notably a Gigabit Ethernet MAC, a step up from the 100Mbps RMII interface on previous ESP32 chips.

These releases are a great example of just how much the embedded hardware market has and continues to grow every week. The STM32C5 gives budget microcontroller projects new capabilities, the ESP32-S31 adds great multimedia features alongside wireless connectivity, and SpacemiT’s K3 is helping push RISC-V into powerful AI development boards and even a modular laptop.

Most platform evaluations start with a demo request and end three weeks later in a conference room. This one takes 15 minutes and puts you directly inside Gladly's interface — navigating it on your own terms.

See how AI surfaces real-time customer context before a conversation starts. Watch how a single conversation thread pulls in purchase history, channel history, and account details without a handoff.

No installation. No commitment. Start the interactive demo and see the platform for yourself.

See it yourself