Intel unveils D-series GPUs with 480GB LPDDR5X for data centers

Intel unveiled its D-series graphics processing units for data center use, introducing a memory configuration that differs from typical server hardware. According to Techmeme, the new chips feature up to 480GB of LPDDR5X memory. This capacity is designed specifically to support demanding artificial intelligence inference tasks within enterprise environments.

Why use laptop memory in servers?

The D-series utilizes LPDDR5X, a memory type commonly found in laptop computers rather than traditional data center servers. TechRadar reports that this choice allows Intel to achieve an unusually large memory capacity of 480GB. This setup aims to handle the high data throughput required by modern AI inference workloads.

What workloads does this target?

The primary focus for these new GPUs is artificial intelligence inference. By providing a massive memory pool, the hardware is suited for models that require large datasets to be loaded simultaneously. This approach addresses the specific demands of inference tasks that often exceed the memory limits of standard server graphics cards.

Is this a new market record?

TechRadar states that the 480GB capacity sets a new GPU memory record. This milestone highlights a shift in how hardware manufacturers are addressing the storage needs of complex AI models. The D-series represents a significant increase in available memory for data center operators.

Questions to ask your suppliers

Ask vendors whether their new hardware relies on consumer-grade components repurposed for enterprise use. Determine if support contracts cover these hybrid designs and verify compatibility with existing infrastructure. Check whether firmware updates address stability concerns unique to this memory configuration. Ensure procurement teams understand the long-term maintenance implications of such specialized hardware choices.

What to do and how to stay safe: Intel

• Verify if current AI workloads benefit from increased memory capacity over raw processing power.
• Assess whether existing server racks can accommodate the power and cooling needs of new GPU hardware.
• Monitor vendor documentation for specific guidance on managing LPDDR5X memory in server environments.
• Review internal procurement policies to ensure they account for non-standard enterprise hardware components.

Step-by-step guide: Browser Security Settings for Small Business Teams

General security guidance from the TechSec Feed newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

How much memory does the Intel D-series GPU have?

The D-series features up to 480GB of LPDDR5X memory according to Techmeme.

What type of memory is used in these GPUs?

They use LPDDR5X, which is typically found in laptop computers.

What is the main use case for this hardware?

The GPUs are designed for demanding artificial intelligence inference workloads.