Core 4 Solutions has listed a used 300-rack AI cluster on eBay for $59.7 million or best offer, advertising up to 9,600 Nvidia A100 GPUs, 4,200 AMD EPYC processors, and overnight delivery for $50.
The Minnesota-based IT hardware seller’s eBay listing describes the equipment as a decommissioned cluster containing 300 racks, with two available. The listing includes free local pickup and free FedEx Ground or FedEx Home Delivery. It also states that the hardware is used and that the seller does not accept returns.
Core 4 Solutions says the cluster was decommissioned from a hyperscaler cloud environment and consists of fully functional ZT A100 compute racks. The seller’s description also includes a 30-day warranty and export compliance requirements covering the use and transfer of the hardware.
The numbers add up quickly
Each rack contains four compute nodes. Every node includes eight Nvidia A100 40GB SXM4 GPUs, two AMD EPYC Rome 64-core processors, 1TB of DDR4 memory, eight ConnectX-6 200Gb/s network cards, an FPGA, NVMe storage slots, and six power supplies.
That gives each rack 32 A100 GPUs, or 9,600 across the full cluster. Each rack also includes three management nodes, with two AMD EPYC processors per node. Combined with the eight processors in the compute nodes, that adds up to 14 CPUs per rack, or 4,200 processors overall.
As Tom’s Hardware reported, the full configuration would provide 268,800 CPU cores and roughly 1.4 petabytes of memory.
The shipping options are unusually flexible for a purchase of this size. Alongside free FedEx Ground delivery, the listing reportedly offers two-day shipping for $25 and overnight shipping for $50. Buyers can also arrange local pickup from the seller’s Minnesota location.
The A100 GPUs and EPYC Rome processors are older data center hardware, but they remain suitable for AI inference, machine learning, and other high-performance computing workloads. The listing’s $59.7 million price tag makes it one of the most expensive pieces of used computer hardware currently available on eBay.
I’m Abhishek Sharma, an author at TigerJek.com. I enjoy exploring games, testing different strategies, and turning what I learn into clear, useful guides. My goal is to help players understand the game better and improve without the usual confusion.
