LattePanda Sigma 2 Packs 180 TOPS for Edge AI
The new Intel-powered board brings high-performance AI inference to a compact footprint for industrial and edge applications.
Updated on Oct. 23, 2026 in Computers
Live Poll
Do you plan to purchase specialized hardware for local AI tasks this year?
LattePanda has released the Sigma 2, a new single-board computer powered by an Intel Core Ultra X7 358H processor. The device is a commercially available system designed to provide high-performance edge computing in a compact 146 by 102 millimeter footprint.
Why it matters
The Sigma 2 brings high-end mobile silicon to the single-board computer market, targeting developers who require significant local AI processing capabilities without the thermal or space constraints of a full-sized workstation. Its integration of specialized NPU and GPU hardware enables complex local workloads that were previously difficult to run on sub-notebook-sized form factors.
The system features 64GB of LPDDR5X-8533 memory and a total of 180 TOPS in peak AI performance, measured against its 4.5-watt idle power consumption. It packs these specs into a 146 by 102 millimeter board, including dual 10GbE ports and a 1TB NVMe SSD.
The players
LattePanda
A manufacturer of high-performance single-board computers that integrate desktop-grade processors with industrial connectivity.
Intel
The semiconductor firm providing the Core Ultra X7 358H processor, which features a dedicated Neural Processing Unit for localized AI acceleration.
The details
The Sigma 2 utilizes the Intel Core Ultra X7 358H processor, which distributes AI workloads across three distinct silicon blocks to achieve its 180 TOPS total output. To handle peripheral management, the board includes an onboard RP2350 microcontroller—a chip specialized for low-latency sensor acquisition and actuator control—that communicates with external hardware through GPIO, or General Purpose Input/Output, interfaces. This architecture allows the device to manage heavy computational tasks while maintaining dedicated, deterministic control over external connected hardware.
Timeline
Mid-November 2026 marks the end of the promotional pre-order pricing window.
The Tech Race
The Sigma 2 marks a departure from hobbyist-grade SBCs, positioning itself against industrial PC modules that prioritize high AI throughput. It moves beyond the limitations of ARM-based development boards by leveraging Intel's x86-based Core Ultra silicon to reach performance tiers previously unreachable in this form factor.
The device runs both Ubuntu 24.04 LTS and Windows 11, making it compatible with existing x86 development environments and professional software stacks. Developers requiring localized, high-speed networking and AI inference can deploy the unit immediately, though power budgets should account for its 40 degrees Celsius maximum operating temperature.
The takeaway
The Sigma 2 demonstrates that high-performance x86 AI compute is becoming viable in specialized, compact edge deployments. Watch for benchmarking reports regarding thermal management as developers begin integrating the unit into closed-chassis, fanless environments.
What happens next
The promotional pre-order price of $2,599 expires in mid-November 2026, at which point the device will shift to its regular price of $2,799.
Further reading
Explore more high-performance hardware trends in our Computers section.
Source note: This article includes information reported by LinuxGizmos.
Live Poll
Do you plan to purchase specialized hardware for local AI tasks this year?
