Dell PowerEdge R6515 8-Bay 2.5"
Configure Your System:
Currently Configured:
$3,964.70
1 Processor Required
Component Guide
Processor.
The R6515 is a single-socket AMD EPYC server, and the design philosophy is worth understanding before you pick a chip: one EPYC socket carries the core counts, memory channels, and PCIe lanes that traditionally required two processors. One socket to power, one socket to cool, and for several licensing models, one socket to license.
The choice is the shape of the compute.
Higher clocks, fewer cores: the frequency-optimized end of the EPYC range suits per-core licensed software and single-thread-sensitive work. Running Windows Server? Its 16-core base license pairs naturally with the 8 and 16 core parts, and every core past the base is a license line, so on Windows the restrained chip is often the financially correct one.
More cores, one socket: the high-core parts are the EPYC headline, dense virtualization and parallel throughput from a single package that two-socket platforms of this era struggle to match. If the workload is VM density, containers, or scale-out compute, this is where the R6515 shines.
Virtualizing on vSphere? VMware's per-socket licensing math is one of EPYC's honest advantages: one populated socket is one socket to license, and the core counts here make that one socket go a very long way.
Whichever chip you pick, the full memory bandwidth and PCIe complement is there; nothing routes through a second socket, because there is not one.
Not sure where your workload lands? We configure these daily; call 1-800-778-1545.
The choice is the shape of the compute.
Higher clocks, fewer cores: the frequency-optimized end of the EPYC range suits per-core licensed software and single-thread-sensitive work. Running Windows Server? Its 16-core base license pairs naturally with the 8 and 16 core parts, and every core past the base is a license line, so on Windows the restrained chip is often the financially correct one.
More cores, one socket: the high-core parts are the EPYC headline, dense virtualization and parallel throughput from a single package that two-socket platforms of this era struggle to match. If the workload is VM density, containers, or scale-out compute, this is where the R6515 shines.
Virtualizing on vSphere? VMware's per-socket licensing math is one of EPYC's honest advantages: one populated socket is one socket to license, and the core counts here make that one socket go a very long way.
Whichever chip you pick, the full memory bandwidth and PCIe complement is there; nothing routes through a second socket, because there is not one.
Not sure where your workload lands? We configure these daily; call 1-800-778-1545.
2 Heat Sink Optional
Component Guide
Heat Sink.
Heatsinks on the R6515 are handled for you: your build ships with the cooling configuration matched to the processor you pick, sized to Dell's requirements for this chassis. There is nothing to decide on this page.
One socket in a 1U is a cleaner thermal problem than two, which works in this platform's favor, but high-core EPYC silicon still runs warm under load, and the chassis manages it exactly as long as the airflow path stays intact. Keep every drive bay blank and internal shroud installed; the R6515 cools front to back, and an open path lets air skip the components that need it.
Give it real datacenter intake air where you can; it will run in a well-ventilated closet, but a 1U under load is not quiet company, and component life is longest in a cold aisle. If your plans include accelerator cards or unusually warm environments, call 1-800-778-1545 before checkout and we will confirm the thermal picture.
One socket in a 1U is a cleaner thermal problem than two, which works in this platform's favor, but high-core EPYC silicon still runs warm under load, and the chassis manages it exactly as long as the airflow path stays intact. Keep every drive bay blank and internal shroud installed; the R6515 cools front to back, and an open path lets air skip the components that need it.
Give it real datacenter intake air where you can; it will run in a well-ventilated closet, but a 1U under load is not quiet company, and component life is longest in a cold aisle. If your plans include accelerator cards or unusually warm environments, call 1-800-778-1545 before checkout and we will confirm the thermal picture.
3 Memory (RAM) 16 DIMM slots, modules added in sets of 2 Required
Component Guide
Memory.
The R6515 pairs its single EPYC socket with eight memory channels and 16 DIMM slots, and that ratio shapes the advice.
Eight matched DIMMs, one per channel, engages the full memory bandwidth of the platform, and that is our default starting point; EPYC's performance story leans on its memory system, and an underpopulated channel layout leaves it on the table. Sixteen DIMMs doubles capacity across the same channels, and it is the natural full build.
Populate in multiples of eight and keep the layout even. The configurator pairs DIMM speeds with what your processor supports.
Buy it as one matched set, now. Hunting for matching DIMMs a year into production means a maintenance window and rank-matching roulette, and a mismatch clocks the whole bank down. Sixteen slots is also the ceiling, so growth on this chassis eventually means swapping DIMMs for bigger ones; size for the workload's future, not its present.
Common landing spots: 128 GB (8x 16 GB) for general duty at full bandwidth, 256 GB (16x 16 GB or 8x 32 GB) is our default virtualization host, and 512 GB (16x 32 GB) when dense consolidation is the reason you bought this many cores.
Eight matched DIMMs, one per channel, engages the full memory bandwidth of the platform, and that is our default starting point; EPYC's performance story leans on its memory system, and an underpopulated channel layout leaves it on the table. Sixteen DIMMs doubles capacity across the same channels, and it is the natural full build.
Populate in multiples of eight and keep the layout even. The configurator pairs DIMM speeds with what your processor supports.
Buy it as one matched set, now. Hunting for matching DIMMs a year into production means a maintenance window and rank-matching roulette, and a mismatch clocks the whole bank down. Sixteen slots is also the ceiling, so growth on this chassis eventually means swapping DIMMs for bigger ones; size for the workload's future, not its present.
Common landing spots: 128 GB (8x 16 GB) for general duty at full bandwidth, 256 GB (16x 16 GB or 8x 32 GB) is our default virtualization host, and 512 GB (16x 32 GB) when dense consolidation is the reason you bought this many cores.
4 RAID Controllers Optional
Component Guide
RAID.
The workload makes this choice, and the list runs from simple to serious.
PERC H740P (8 GB battery-backed cache) is the configurator default and real muscle: a large cache and throughput that keeps a busy backplane honest. For virtualization datastores, databases, and any local array doing real work, it is the right default and we leave most builds on it.
PERC H730P (2 GB cache) and H730 (1 GB cache) are the sensible savings on lighter builds where the array works modestly; on a full backplane with write load, the H740P's cache earns its price.
PERC H330 has no cache. It handles a RAID 1 boot mirror honestly, which is the right amount of controller for a host whose real storage lives on a SAN or software-defined layer, and nothing heavier.
Building vSAN, Ceph, ZFS, or another software-defined stack? Those want direct drive access through an HBA rather than a RAID controller, and a high-core single-socket node is a textbook fit for that architecture; call 1-800-778-1545 and we will spec the right HBA for that build.
Array guidance regardless of controller: RAID 10 where write latency matters, RAID 6 over RAID 5 on large spinning drives, and a hot spare on any array whose rebuild you would rather not race.
PERC H740P (8 GB battery-backed cache) is the configurator default and real muscle: a large cache and throughput that keeps a busy backplane honest. For virtualization datastores, databases, and any local array doing real work, it is the right default and we leave most builds on it.
PERC H730P (2 GB cache) and H730 (1 GB cache) are the sensible savings on lighter builds where the array works modestly; on a full backplane with write load, the H740P's cache earns its price.
PERC H330 has no cache. It handles a RAID 1 boot mirror honestly, which is the right amount of controller for a host whose real storage lives on a SAN or software-defined layer, and nothing heavier.
Building vSAN, Ceph, ZFS, or another software-defined stack? Those want direct drive access through an HBA rather than a RAID controller, and a high-core single-socket node is a textbook fit for that architecture; call 1-800-778-1545 and we will spec the right HBA for that build.
Array guidance regardless of controller: RAID 10 where write latency matters, RAID 6 over RAID 5 on large spinning drives, and a hot spare on any array whose rebuild you would rather not race.
5 Storage Drives Select up to 8 drives (0/8 Slots Used) Optional
Component Guide
Drive.
Drives are the number one failure point on any server, so this decision deserves more attention than the spec sheet suggests.
The 2.5" R6515 takes SAS and SATA in both flavors, with NVMe where the backplane supports it. Our guidance by workload: NVMe when storage latency is the product, and on a platform with this many PCIe lanes from one socket, NVMe is playing to the architecture's strength; SATA and SAS SSDs for mainstream flash duty where they remain the value play; 10K SAS spinning drives only for bulk that genuinely does not need speed. A dense single-socket virtualization node wants its hot data local and fast.
Buy new. We steer you away from refurbished on drives specifically, and we say that as a company that sells refurbished servers all day. The chassis, board, and power supplies in a refurbished R6515 have proven themselves through our burn-in; a used drive with unknown hours is a different bet, and a rebuild is exactly when you do not want to lose a second one.
Boot the right way: the BOSS card in the Add-Ons step carries a mirrored pair of M.2 drives dedicated to the operating system, keeping every front bay for data and separating a boot failure from a data failure permanently. It is standard practice on this platform and we recommend it on every build.
Fair warning on capacity math: RAID overhead, hot spares, and filesystem headroom mean usable space lands well under the sum of the labels. Size for the usable number, and when in doubt, one drive size up is cheap insurance.
The 2.5" R6515 takes SAS and SATA in both flavors, with NVMe where the backplane supports it. Our guidance by workload: NVMe when storage latency is the product, and on a platform with this many PCIe lanes from one socket, NVMe is playing to the architecture's strength; SATA and SAS SSDs for mainstream flash duty where they remain the value play; 10K SAS spinning drives only for bulk that genuinely does not need speed. A dense single-socket virtualization node wants its hot data local and fast.
Buy new. We steer you away from refurbished on drives specifically, and we say that as a company that sells refurbished servers all day. The chassis, board, and power supplies in a refurbished R6515 have proven themselves through our burn-in; a used drive with unknown hours is a different bet, and a rebuild is exactly when you do not want to lose a second one.
Boot the right way: the BOSS card in the Add-Ons step carries a mirrored pair of M.2 drives dedicated to the operating system, keeping every front bay for data and separating a boot failure from a data failure permanently. It is standard practice on this platform and we recommend it on every build.
Fair warning on capacity math: RAID overhead, hot spares, and filesystem headroom mean usable space lands well under the sum of the labels. Size for the usable number, and when in doubt, one drive size up is cheap insurance.
6 Remote Access Required
Component Guide
Remote Access.
The R6515 ships with iDRAC9, the best out-of-band management Dell has built, and the license determines what you can do with it.
The dividing line is the remote console. iDRAC9 Enterprise gives you remote KVM and virtual media through a clean HTML5 interface: reach the server's screen, mount an ISO, and rebuild an OS from anywhere, even when the operating system is down. The lower tiers leave you walking to the rack for any problem below the OS.
Enterprise is our pick, and on a production virtualization host the math is short: the license costs a fraction of one emergency site visit, and the first bad firmware day pays for it.
Enterprise also brings automated firmware updates, telemetry streaming, and secure erase for the day the server retires holding company data.
Skip it only when the server sits down the hall and someone can physically reach it without ceremony, and even then, think about whether that stays true for the machine's whole life.
If you are deploying 5 or more units, take Enterprise on all of them. Fleet management without remote console access does not work in practice.
The dividing line is the remote console. iDRAC9 Enterprise gives you remote KVM and virtual media through a clean HTML5 interface: reach the server's screen, mount an ISO, and rebuild an OS from anywhere, even when the operating system is down. The lower tiers leave you walking to the rack for any problem below the OS.
Enterprise is our pick, and on a production virtualization host the math is short: the license costs a fraction of one emergency site visit, and the first bad firmware day pays for it.
Enterprise also brings automated firmware updates, telemetry streaming, and secure erase for the day the server retires holding company data.
Skip it only when the server sits down the hall and someone can physically reach it without ceremony, and even then, think about whether that stays true for the machine's whole life.
If you are deploying 5 or more units, take Enterprise on all of them. Fleet management without remote console access does not work in practice.
7 Power Supply 2x included in build Required
Component Guide
Power Supply.
There is no decision to make on this page, and that is the single-socket dividend: the R6515 ships with a redundant pair of 550W hot-swap power supplies, and one EPYC socket with a bay of drives lives comfortably inside it.
What the pair buys you: full redundancy, either supply can carry the whole server if its partner fails, and hot-swap serviceability, so a failed unit is replaced without a shutdown. Run each cord to a separate circuit or PDU and a single upstream failure does not touch the workload.
Watch the estimated TDP counter at the bottom of the page as you add components; on this platform it mostly serves to confirm how much headroom the pair has left. If your plans include accelerator cards, call us before checkout and we will confirm the power picture for the exact build.
What the pair buys you: full redundancy, either supply can carry the whole server if its partner fails, and hot-swap serviceability, so a failed unit is replaced without a shutdown. Run each cord to a separate circuit or PDU and a single upstream failure does not touch the workload.
Watch the estimated TDP counter at the bottom of the page as you add components; on this platform it mostly serves to confirm how much headroom the pair has left. If your plans include accelerator cards, call us before checkout and we will confirm the power picture for the exact build.
8 Network Cards Required
Component Guide
Network Card.
The R6515 uses the OCP 3.0 mezzanine slot for its primary network card: it does not consume a PCIe slot, and on a single-socket platform whose PCIe lanes are one of its selling points, that matters. Two decisions: speed and port type.
Quad 1 GbE covers light duty honestly, and it is the budget answer when the clients are desktops on gigabit switching.
10 Gb is the floor the moment this server hosts a busy hypervisor, shared storage, or backup traffic on a deadline, and a high-core R6515 consolidating a rack of VMs concentrates all of their traffic onto this one card. The dual and quad 10 Gb options, and 25 Gb where the switching exists, match what the platform can push.
Then port type, and this is the one that catches people. SFP+/SFP28 ports take fiber or DAC cables and talk to datacenter-grade switches. BASE-T ports are 10 Gb copper on ordinary RJ45. Neither is better; they must simply match your switch. Order SFP against a copper-only switch and the server racks up with nothing to plug into.
Need more ports, or separate links for storage and VM traffic? PCIe network cards stack on top of whatever you choose here. Call 1-800-778-1545 and we will spec the whole path, switch ports included.
Quad 1 GbE covers light duty honestly, and it is the budget answer when the clients are desktops on gigabit switching.
10 Gb is the floor the moment this server hosts a busy hypervisor, shared storage, or backup traffic on a deadline, and a high-core R6515 consolidating a rack of VMs concentrates all of their traffic onto this one card. The dual and quad 10 Gb options, and 25 Gb where the switching exists, match what the platform can push.
Then port type, and this is the one that catches people. SFP+/SFP28 ports take fiber or DAC cables and talk to datacenter-grade switches. BASE-T ports are 10 Gb copper on ordinary RJ45. Neither is better; they must simply match your switch. Order SFP against a copper-only switch and the server racks up with nothing to plug into.
Need more ports, or separate links for storage and VM traffic? PCIe network cards stack on top of whatever you choose here. Call 1-800-778-1545 and we will spec the whole path, switch ports included.
9 Operating System Required
Component Guide
Operating System.
We install, license, and burn-in test your OS before shipping, so the server arrives ready to join the network.
The good news on this platform: EPYC is a first-class citizen everywhere that matters, so every modern OS supports the R6515 fully and the choice is about your environment, not compatibility.
Windows Server 2022 is the mainstream pick, and Windows Server 2025 is fully supported if you want the newest platform. Licensing is per core with a 16-core base, and on this chassis the core count you chose in the processor step is the licensing decision: a 16-core EPYC lands exactly on the base license, while the high-core parts multiply the license math even faster than they multiply the VMs. We will run both numbers with you at quote time. On CALs, count your users and round up.
Fair warning on Evaluation Editions: free and fully functional for 180 days, then they expire. Right for labs and proof-of-concept, wrong for production.
VMware ESXi is fully supported, and the single-socket licensing math is one of the reasons EPYC nodes are popular vSphere hosts; we will install the version your cluster runs.
Proxmox VE is our quiet favorite for new deployments: open-source KVM with clustering and live migration, no per-core license math at all, and a high-core R6515 running it is about the cheapest legitimate consolidation host per VM that money buys.
Ubuntu Server LTS covers Linux and container duty with five years of free updates, and No OS is there for teams imaging from their own deployment infrastructure.
Not sure which fits? Tell us the workload at quote time, or call 1-800-778-1545.
The good news on this platform: EPYC is a first-class citizen everywhere that matters, so every modern OS supports the R6515 fully and the choice is about your environment, not compatibility.
Windows Server 2022 is the mainstream pick, and Windows Server 2025 is fully supported if you want the newest platform. Licensing is per core with a 16-core base, and on this chassis the core count you chose in the processor step is the licensing decision: a 16-core EPYC lands exactly on the base license, while the high-core parts multiply the license math even faster than they multiply the VMs. We will run both numbers with you at quote time. On CALs, count your users and round up.
Fair warning on Evaluation Editions: free and fully functional for 180 days, then they expire. Right for labs and proof-of-concept, wrong for production.
VMware ESXi is fully supported, and the single-socket licensing math is one of the reasons EPYC nodes are popular vSphere hosts; we will install the version your cluster runs.
Proxmox VE is our quiet favorite for new deployments: open-source KVM with clustering and live migration, no per-core license math at all, and a high-core R6515 running it is about the cheapest legitimate consolidation host per VM that money buys.
Ubuntu Server LTS covers Linux and container duty with five years of free updates, and No OS is there for teams imaging from their own deployment infrastructure.
Not sure which fits? Tell us the workload at quote time, or call 1-800-778-1545.
WServer Warranty
Dell PowerEdge R6515 8-Bay 2.5" Detailed Review
Build Your Own Dell PowerEdge R6515 8-Bay Server
The Dell PowerEdge R6515 8-bay 2.5" server is a high-efficiency 1U rack server designed to deliver exceptional performance using a single-socket architecture. Powered by AMD EPYC processors, the R6515 offers impressive core density and value, making it ideal for virtualization, cloud platforms, and data-intensive workloads.
This Build-Your-Own R6515 platform allows you to configure your system for your exact deployment needs. Customize your server with CPU, memory, RAID controller, and storage options including SAS, SATA, or SSD drives, enabling optimized performance for both speed and capacity.
With its 8-bay SFF (2.5") chassis, the R6515 provides flexible storage options while maintaining a compact footprint—perfect for virtualization clusters, scalable infrastructure, and cost-efficient data center deployments.
At Wholesale Servers, every system is professionally tested and built to support bulk purchasing, enterprise workloads, and reliable IT operations.
Included with every server at no charge. Covers parts and labor on hardware failures for 180 days.
Get full hardware replacement coverage
including a rebuilt replacement server in the case of non-recoverable
system failure.
Get full hardware replacement coverage
including a rebuilt replacement server in the case of non-recoverable
system failure.
Get full hardware replacement coverage
including a rebuilt replacement server in the case of non-recoverable
system failure.
Save Your Design
Click the Add to Quote button at the bottom of your screen to save your design as a draft order for future reference and to check for discounts, lead time, and availability. Most servers ship within 1-3 days.
