Liquid LEAP Now Supports AMD Ryzen for Fast On-Device AI
Liquid AI, the MIT-born chief in environment friendly basis fashions, at the moment introduced native assist for AMD’s newest Ryzen
and Ryzen AI processors inside the Liquid Edge AI Platform (LEAP), unlocking new efficiency good points for builders constructing high-quality, on-device AI options.
With this replace, builders acquire instant entry to Liquid’s low-latency, memory-optimized fashions, now natively accelerated on AMD’s newest “Zen 5” with AMD Radeon built-in graphics. The result’s a unified software program and {hardware} path for deploying privacy-preserving, real-time AI purposes instantly on PCs, with out reliance on cloud inference.
Efficiency With out Compromise
Early benchmarks of Liquid’s LFM fashions operating by the LEAP SDK on AMD {hardware} display:
- Excessive throughput effectivity (tok/sec) at low energy draw utilizing AMD’s built-in compute accelerators
- Zero dependency on cloud APIs—preserving knowledge privateness and lowering latency
- Seamless deployment for each client and enterprise software program
With native processor assist, builders can ship light-weight, high-performing AI experiences that scale from client laptops to enterprise endpoints with out modification or added complexity.
Goal-Constructed for the Edge
Liquid’s LFM structure is engineered for environment friendly edge inference. The brand new AMD-optimized method with LEAP additional reduces integration time for builders, offering a direct route for mannequin deployment.
“This integration is about collapsing complexity for builders,” stated Mathias Lechner, CTO and Co-Founding father of Liquid AI. “By tuning LEAP for AMD Ryzen processors, we’re enabling groups to ship AI merchandise quicker—with the pace, effectivity, and management that edge purposes demand.”
“Builders desire a friction-free path to ship AI that’s quick, non-public and power-efficient,” stated Mark Papermaster, chief know-how officer and govt vp at AMD. “Pairing AMD Ryzen and Ryzen AI processors with Liquid AI’s memory-lean LFM fashions within the LEAP SDK collapses complexity, giving creators sub-100-millisecond responsiveness on-device whereas retaining knowledge on the consumer’s PC. It’s a strong instance of the open-ecosystem, holistic design method we champion to carry energy-efficient intelligence.”
Availability
The LEAP SDK with AMD acceleration is on the market at the moment, supporting the LFM-2 mannequin households. Builders can discover instruments, benchmarks, and documentation at leap.liquid.ai.
The publish Liquid LEAP Now Supports AMD Ryzen for Fast On-Device AI first appeared on AI-Tech Park.