Intel launches the D270 inference-acceleration GPU
What happened
On October 13 the Hacker News front page reported on Intel's D270 AI acceleration GPU. The product is a data center GPU tuned for inference and aimed at agentic AI, offering up to 480 GB of LPDDR5X memory, 32 Xᵉ3P cores and 256 XMX matrix engines, and targeting long-context, high-concurrency and prefill workloads. For deployment, the D270 uses a standard PCIe 5.0 interface and a 350 W air-cooled design. Intel says the product supports vLLM and the open PyTorch ecosystem and can be deployed in existing servers without dedicated networking or liquid-cooling changes.
Written by AI from the coverage · updated 3 hours ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageIntel D270 GPU AI Accelerator–Up to 480 GB LPDDR5X
Intel D270 是面向智能体 AI 的推理优化数据中心 GPU，提供最高 480 GB LPDDR5X 内存。它配备 32 个 Xᵉ3P 核心和 256 个 XMX 矩阵引擎，采用标准 PCIe 5.0 和 350 W 风冷部署，面向长上下文、高并发及预填充处理负载。Intel 表示其支持 vLLM 和开放 PyTorch 生态，可在现有服务器中部署，无需专用网络或液冷改造。
Heat over time
Heat now 9·Comparable peak 9（Oct 13 09:00）·Comparable change over 24 hours –
The trend only compares accounts observed without gaps, so its range may be smaller than the current heat. Hover or tap the chart for each hour; the left and right arrow keys step through it.