Bulk pricing on 10x E5-2680v4 servers, any takers?

Hey folks, we've got 10 Dell PowerEdge R730 units coming back from a customer lease buyout. Dual E5-2680v4, 64GB DDR4, dual 1TB SAS drives, full warranty.

Looking to move these quickly. Individual pricing is around €420/mo per unit, but we can do €380/mo if someone wants the whole rack. These are solid workhorses for database clusters or batch processing.

Located in our Frankfurt facility (Equinix). Can provision and deploy within 48 hours. Transit latency is great if you're EU-based.

Any interest? Happy to negotiate on bandwidth commitments or longer-term contracts.

Those specs are pretty dated for 2026 tbh. E5-v4 is like 10 years old now. What's the actual power draw and cooling requirements? Some facilities charge extra for older gear that pulls 300W+. Also curious about the warranty terms—is it Dell ProSupport or just hardware replacement?

Fair point on the age, yeah. Power draw is around 280W each under load, so shouldn't trigger extra cooling fees at most facilities. We've already got them provisioned in Frankfurt with no issues. Might not be cutting-edge but the price reflects that—good value if you need reliable CPU for data work rather than bleeding-edge performance.

That's actually decent hardware for certain workloads. E5-v4 still crushes it for legacy code and batch jobs. Quick question though—are these single-socket or dual? If dual E5-2680v4, that's 28 cores/56 threads per box, which is solid for parallel workloads. What's the RAM config exactly, and are you willing to do smaller commitments (2-3 units) at closer to that €380 price? Asking for a friend running Elasticsearch clusters in EU.

Dual E5-2680v4 is actually perfect for MongoDB or Cassandra clusters if you're doing replication across multiple nodes. 28 cores per box means solid throughput for time-series workloads. At €380/mo for the rack you're basically getting enterprise gear for pennies. Only concern: DDR4 on those old Xeons maxes out around 2133MHz, so if you're doing in-memory analytics you might hit some latency walls. But for typical DB ops? Solid buy.

280W is solid for that class. We ran similar setups for Elasticsearch clusters and they handled it well, though you'll want redundant PSUs if this is production.

280W each is reasonable, but have you benchmarked memory bandwidth on those vs newer Xeons? Might matter for distributed workloads.