Intel
Power AI Everywhere with Intel & Lablup
Lablup partners with Intel to bring Intel® Gaudi® AI Accelerators to Backend.AI, an AI infrastructure operating platform built for scale.
Learn more about Backend.AI
Intel and AI acceleration
Intel has been at the center of High Performance Computing (HPC) for decades. Gaudi AI Accelerators handle deep learning training and inference, while Xeon Scalable Processors power data centers worldwide.
The oneAPI ecosystem gives developers a single programming model across different hardware, and Intel Tiber AI Cloud provides on-demand access to Gaudi instances.
×
Backend.AI and Intel Gaudi
Backend.AI’s Sokovan™ orchestrator distributes workloads across Intel Gaudi’s compute engines to reduce bottlenecks. It schedules model training and inference down to memory allocation and parallelism, so Gaudi operates closer to its rated throughput.
The full stack for AI
Gaudi AI Accelerators
Purpose-built accelerators for deep learning training and inference at scale
Xeon Scalable Processors
Enterprise-grade CPUs powering data centers and AI workloads worldwide
oneAPI Ecosystem
Open, unified programming model for heterogeneous computing
Intel Tiber AI Cloud
Cloud platform providing on-demand access to Intel AI hardware
Sokovan™ Orchestrator
Workload distribution across Gaudi compute engines
Container-level Virtualization beta
Patented technology for fine-grained accelerator resource management
Multi-tenant AI Platform
Secure, isolated environments for teams sharing AI infrastructure
End-to-end MLOps
Complete lifecycle management from experimentation to production
Intel’s AI hardware and Lablup’s platform software, combined in a single stack
What our leaders say
“Lablup unlocks the full potential of Intel Gaudi AI Accelerators with Backend.AI — AI infrastructure operating platform. Backend.AI orchestrator, Sokovan intelligently distributes workloads across Intel Gaudi’s compute engines, significantly reducing bottlenecks. Every model training and inference task is precisely orchestrated—from memory usage to parallelism—maximizing customers AI performance. Backend.AI doesn’t just run on Gaudi—it unleashes Gaudi.
Jeongkyu Shin, CEO of Lablup Inc.
“The partnership between Intel and Lablup represents a perfect fusion of hardware excellence and software intelligence. Together, we’re redefining what’s possible for GenAI and LLMs, delivering not just performance but unprecedented efficiency at scale. Whether in the cloud or your data center, this collaboration brings the AI industry exactly what it needs right now: choice without compromise.
Arijit Bandyopadhyay, CTO of Enterprise Analytics & AI, Head of Strategy - Enterprise & Cloud (CSV Group) at Intel Corporation
Our journey with Intel
Click each item to read the press release or blog post.
Learn more about Backend.AI & Intel
Your AI infrastructure, built on Intel Gaudi.
Talk to our team about deploying Backend.AI on Gaudi 2 and Gaudi 3.
