en.Wedoany.com Reported - Intel showcased AI deployment scenarios including content creation, AI computing acceleration, and agent deployment at the Volcengine FORCE Original Power Conference, unveiling the Xeon 6/6+ processors, Arc Pro B70 GPU, and Core Ultra processors.
Guo Wei, Vice President of Intel's Marketing Group and General Manager of Intel China, stated that from the lab to large-scale deployment, the evolution of AI brings opportunities for full-system collaboration. To support complex workloads such as inference and Agentic AI in real data center environments, the challenge lies in achieving coordinated operation of products like CPUs, GPUs, and IPUs. The deep collaboration between Intel and Volcengine aims to integrate technological advantages and provide an integrated solution for Agentic AI.
To address performance bottlenecks in the large-scale deployment of intelligent agents, Intel and Volcengine jointly launched the Intel Xeon 6 Intelligent Agent Deployment Optimization Solution. This solution layers the system from bottom to top, covering core data such as system-level operational parameters and sandbox resource utilization. By capturing runtime latency in the agent execution chain and external tool invocation phases, it builds an integrated performance monitoring system.
In terms of agent deployment, Intel released the Xeon 6+ processor, integrating up to 288 efficiency cores. Each processor core can run four agents, enabling a single processor to support stable operation of over 1,000 agents. Intel's AMX technology targets high-concurrency online scenarios, achieving breakthroughs in core operators such as video transcoding and Whisper speech recognition, allowing customers to improve performance with zero hardware cost overhead.
In terms of reliability and operational costs, Intel TDX and the Volcengine AICC Confidential Computing Platform jointly create a trusted end-to-end data protection solution for intelligent agents. This solution, based on Intel Xeon processors and TDX confidential computing technology, enables full lifecycle isolation of AI data and computing resources in multi-tenant environments, from user input, memory retrieval, model inference, agent decision-making, to tool invocation. Intel demonstrated the AMX-powered ArkClaw user TCO optimization solution, which efficiently preprocesses and postprocesses data via the CPU, significantly reducing the number of large model calls. The two parties built multiple localized skills covering ASR, TTS, PDF parsing, VSR, and Embedding, effectively reducing token consumption for online agent services and optimizing TCO.
In the edge and terminal domains, Intel collaborated with Volcengine for strategic deployment. The Intel Arc Pro B70 features large video memory and cost-effectiveness, suitable for creative work and AI inference scenarios. In the XR content creation field, PICO 4DGS and Intel explored using the B70 to accelerate spatial video 3D reconstruction, reducing designer wait times. With the B70's 32GB video memory and 367 Tops computing power, AgentSphere achieves higher concurrency, lower latency, and less jitter in multi-agent collaboration. Through an out-of-the-box standard all-in-one solution, it enables enterprises to more conveniently and quickly build an AI employee management platform.
For consumer scenarios, Intel leverages Core Ultra processors to provide AI computing power for AI PCs, collaborating with teams such as Kouzi, TRAE, Xiaoyunque, and Cherry Voice Assistant, extending to scenarios including content creation, intelligent PC voice, and gaming.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
