Intel Transforms Power, Performance, and

Affordability, Accelerating AI Everywhere

 

 

Computex 2024 Unveils a Game-Changing Trio

 

 

At Computex 2024, Intel introduced three major innovations Lunar Lake architecture, Gaudi AI accelerators, and Intel Xeon 6 processors designed to make AI accessible across networking, cloud, client, and edge systems. Together, these breakthroughs combine high performance, energy efficiency, and cost-effectiveness to redefine what’s possible in AI computing.

Major Announcements at a Glance

• Xeon 6 processors with Efficient‑cores (E‑cores) optimize scale‑out, high‑density workloads, achieving 4.2× rack-level performance and 2.6× performance-per-watt gains.

• Gaudi 2 & 3 accelerator kits deliver generative AI performance at up to one-third the cost of rival platforms.

• Lunar Lake introduces an efficient client chip architecture with up to 40% lower SoC power and 3× more AI compute than its predecessor.

Altogether, these announcements reinforce Intel’s vision of AI-powered computing with Intel at Computex 2024.

CEO Gelsinger’s Vision & Ecosystem Strategy

Pat Gelsinger called AI “one of the most consequential eras of innovation.” To support this, he emphasized Intel’s unique hardware-to-software strategy. Additionally, he highlighted partnerships with Microsoft, Acer, and other major players, underscoring a firm commitment to open standards and rapid deployment across ecosystems.

Xeon 6 Redefines Data-Center Efficiency

Intel launched the Sierra Forest (E-core) variant on June 4, 2024, with Granite Rapids (P-core) following in Q3. In comparison to previous generations, these processors mark a significant leap forward.

• Sierra Forest offers up to 4.2× rack-level performance and 2.6× efficiency compared to 2nd Gen Xeon, enabling 3:1 server consolidation.

• As a result, data centers can free up infrastructure for AI workloads while using less power and physical space.

Gaudi Accelerators: AI at a Fraction of the Cost

Intel confirmed pricing that disrupts the current market:

• Gaudi 2 kit (8 accelerators + UBB): $65,000  roughly one-third the cost of similar offerings.

• Gaudi 3 kit: $125,000, which is about two-thirds the cost of high-end alternatives.

Consequently, organizations can access advanced generative AI without overspending. In fact, analysts note that Gaudi 3 accelerates LLM training by up to 40% and inference by as much as 2× compared to Nvidia H100 clusters.

Furthermore, tier‑1 vendors including Dell, Lenovo, ASUS, Foxconn, Gigabyte, and others are already integrating these kits into their platforms, showing rapid adoption across industries.

Lunar Lake: The Next-Gen AI PC Engine

Intel also introduced Lunar Lake, designed for ultra-efficient AI-powered laptops:

• Up to 40% lower SoC power.

• 3× more AI compute capability.

• An NPU delivering 48 TOPS of AI performance.

• A new Xe2 GPU providing 1.5× better graphics and 67 TOPS AI acceleration.

Meanwhile, Intel’s improved architecture combines P-cores, Skymont E-cores, and new AI-tuned GPUs to maximize performance. As a result, users will experience faster, more responsive devices without sacrificing battery life.

Looking ahead, Intel expects Lunar Lake to power 80+ AI PC designs from over 20 OEMs, with shipments beginning in Q3 2024.

Why This Matters

Intel’s Computex 2024 announcements reflect a strategic and technological shift toward democratizing AI:

• Xeon 6 boosts data-center efficiency, scalability, and sustainability.

• Gaudi accelerators significantly reduce AI development costs while improving output.

• Lunar Lake brings advanced AI capabilities directly to thin-and-light laptops.

In summary, these innovations deliver AI everywhere from cloud to edge, server to laptop—unlocking the full potential of AI across industries.