We take cryptocurrency — so the coins you hold can buy real hardware.Pay in crypto, spend it on real hardware. Free insured shipping worldwide over $399. Plain, unbranded boxes, sent from Germany.Free insured shipping over $399. Your rate is locked for 30 minutes once checkout opens.Rate locked 30 minutes. Every item sealed, serial-checked and covered by a 24-month warranty.Sealed, 24-month warranty.

Saved items Sign in
HPE ProLiant DL384 Gen12 barebone server chassis front view

HPE

HPE ProLiant Compute DL384 Gen12 barebone server

SKU CH-K1AV-ZGJBT MPN HPE ProLiant Compute DL384 Gen12

Barebone server for generative AI workloads with NVIDIA GH200 NVL2 superchips

  • Supports up to two NVIDIA GH200 NVL2 superchips per server
  • Outstanding coherent memory and memory bandwidth for larger models
  • Superchip performance delivers superb performance per GPU in HPE ProLiant portfolio
  • Simplifies scaling with data demands for mixed workloads and RAG

HPE ProLiant DL384 Gen12 for AI

The HPE ProLiant Compute DL384 Gen12 is a barebone server built for accelerated computing. It supports up to two NVIDIA GH200 NVL2 superchips to run large generative AI models. Teams can scale out infrastructure for retrieval-augmented generation and mixed workloads in a hybrid environment. This platform targets organisations that need coherent memory and high bandwidth for AI factories.

Dual superchip configuration decides fit

The server holds up to two NVIDIA GH200 NVL2 superchips, which determines whether the density matches your model size. Buyers who need more than two superchips per node or who require a different GPU architecture should evaluate other platforms. The barebones design means you add the specific components your workload demands.

Highlights

  • Supports up to two NVIDIA GH200 NVL2 superchips per server
  • Outstanding coherent memory and memory bandwidth for larger models
  • Superchip performance delivers superb performance per GPU in HPE ProLiant portfolio
  • Simplifies scaling with data demands for mixed workloads and RAG

Specifications

No published specifications for this item yet.

Questions about this item

What does the 384 figure mean for model size once the server is deployed?

The 384 figure represents the maximum number of superchips the system can scale to, letting teams run larger models across the expanded coherent memory pool.

What interface standard connects the superchips and what does it cap in practice?

NVIDIA GH200 NVL2 links the superchips, capping the configuration at two per server while delivering the coherent memory bandwidth needed for generative AI workloads.

New to this? Buying server hardware with crypto — what to check, what it costs to pay, and what is never asked for.

Also in this aisle

People compared these

Same shelf, same checkout — eight coins and a 30-minute rate lock.