Dell PowerEdge R640 10-Bay 2.5" + RFB

Configure Your System:

Currently Configured: $375.40
1 Processor Required
Series
Category
CPU (2x included in build)
Component Guide Processor. This step works in two passes. Pick a processor series first, which narrows the list, then choose the CPU that fits your workload and licensing.

Step 1: Choose a series

Both generations drop into the same R640 motherboard, so this is a price and features decision, not a compatibility one.

Intel Xeon Gen 1 (Skylake-SP, 4100/5100/6100/8100 series) is the value play. Mature, plentiful, and the best price per core in the lineup. Memory tops out at 2666 MT/s.

Intel Xeon Gen 2 (Cascade Lake-SP, 4200/5200/6200/8200 series) adds higher clocks per tier, 2933 MT/s memory support on Gold 6200 SKUs, and Optane PMem support on L-series parts. Worth the premium when per-core performance or memory features are load-bearing; otherwise Gen 1 does the same job for less.

Whichever series you pick, populate both sockets. A single-CPU build loses half its memory channels, half its DIMM slots, and a PCIe riser.

Step 2: Choose your CPU

Match the tier to the workload, and let your OS licensing steer the core count.

Running Windows Server 2022 or 2025? Two 8-core Gold CPUs land exactly on the 16-core base license. Go to 10 cores each and you are buying four extra core packs before the server performs any differently. Our pick is the Gold 6134 (8 cores, 3.2 GHz, 130W): high clocks per core, which is what Windows and SQL Server workloads actually want. The Gold 6250 is faster still, but costs many times more for a modest clock gain, and at 185W it requires the high-performance heatsink.

Virtualizing on vSphere? Ignore the 8-core strategy. VMware licenses 16 cores per socket regardless, so buy the cores your VMs need. The Gold 6230 (20 cores, 2.1 GHz, 125W) is the balanced default; the Gold 6230R (26 cores, 150W) is the VDI density play.

Silver covers light-duty and budget builds. Platinum buys maximum core count at a steep premium, and most workloads that need 24+ cores per socket are better served by the Gold 6248 or 6252 at a fraction of the price.

Fair warning on any CPU above 150W: the high-performance heatsink and fan kit are required, not optional. Confirm they are in your build before checkout.
2 Heat Sink Optional
Component Guide Heat Sink. The R640 ships with a standard heatsink or a high-performance heatsink, and the choice is not cosmetic. Get it wrong and the server throttles or refuses to POST.

Standard heatsink covers CPUs up to 150W TDP in a normal datacenter environment. That is most of what we quote, including the Gold 6134 and Gold 6230.

High-performance heatsink is required above 150W (Gold 6250, 6248R, 6258R and similar), and also on lower-TDP CPUs in two cases: rear-drive or fully populated NVMe configurations, and ambient inlet temperatures above 30°C.

If you are unsure about your rack's inlet temperature, spec the high-performance heatsink. The cost difference is small and it is a lot cheaper than a thermal event.
3 Memory (RAM) 24 DIMM slots, modules added in sets of 2 Required
Total Installed Memory
RAM Clock Speed
RAM Configuration
Component Guide Memory. Fill all 24 slots. This is our standing recommendation on the R640 and we push it hard.

There are two benefits. The first is performance: full population at 2 DPC keeps all twelve memory channels working at maximum depth, and on memory-bandwidth-sensitive workloads like VDI and high-density virtualization it consistently outperforms a partial population running at higher clock. The second is cost per gigabyte: 24 smaller modules hit the same capacity for meaningfully less than 12 larger ones, and 768 GB as 24× 32 GB is the best cost per GB in the lineup.

We know the instinct is to leave a few slots open for a later upgrade, and on paper that is sensible. In practice it rarely plays out well. One year from now you are hunting for DIMMs that match the rank and speed of what is already in the box, and a mismatch either clocks the whole server down or refuses to train. The server is in production by then, so adding memory means a maintenance window you have to schedule and justify.

Buy the capacity now, in one matched set. If you outgrow it, the right move is usually another node rather than another DIMM.

384 GB (24× 16 GB) for file, print, domain controller duty through a modest virtualization host or SQL Server instance
768 GB (24× 32 GB) is our default for a general-purpose ESXi or Hyper-V host
1.5 TB (24× 64 GB) for dense consolidation, VDI, or in-memory workloads

Past 1.5 TB you are into 128 GB LRDIMMs, which reach 3 TB but cost substantially more per GB. Most buyers are better served by a second host.
4 RAID Controllers Optional
RAID
Software RAID

Software RAID controllers can only support SATA drives

Component Guide RAID. The controller decision comes down to one question: how much does local write performance matter? The cache size is what you are paying for, and the workload tells you how much cache you need.

PERC H730P (2 GB cache, battery-backed) is our default on this chassis and the most common spec we quote. The 2 GB cache is appropriately sized for an 8-drive array on mixed or read-heavy workloads, and it covers the large middle of what this server actually runs.

PERC H740P (8 GB NV cache, battery-backed) is the production storage pick for write-intensive or transactional workloads: SQL Server, heavy logging, anything where write latency is load-bearing. The step up from the H730P is small money for four times the cache. If you are unsure which side of the line your workload falls on, this is the safe answer.

PERC H730 (1 GB cache, battery-backed) is the 13th-gen carryover that still works in this chassis. Viable when budget is the constraint and write performance is not, but the H730P is a small step up for double the cache. We rarely quote the H730 when the buyer sees both prices side by side.

PERC H330 (no cache) is entry-tier hardware RAID for light workloads. No cache means write performance depends entirely on the drives. Fine for a domain controller or file server; wrong for anything transactional.

S140 software RAID is dev/test only, and SATA drives only. Not a production recommendation.

One configuration this list does not cover: if you are running vSAN, Storage Spaces Direct, or Ceph, you do not want hardware RAID at all. Those stacks need direct drive access through an HBA330. Call us at 1-800-778-1545 and we will quote it.
5 Storage Drives Select up to 12 drives (0/12 Slots Used) Optional
Recommended Drives
Show New SAS SSDs 2.5"
Show New SATA SSDs 2.5"
Show New SAS HDDs 2.5"
Show Refurbished SAS SSDs 2.5"
Show Refurbished SAS HDDs 2.5"

Drive trays are included and matched to your chassis automatically

Component Guide Drive. Drives are the number one failure point on any server, and this chassis gives you the densest storage arrangement an R640 offers: ten front bays plus rear flex bays, all 2.5".

The front ten take SAS and SATA in both flavors, with NVMe where the backplane supports it. Our guidance by workload: SSDs for anything latency-sensitive (virtualization datastores, databases, application volumes), 10K SAS spinning drives where capacity per dollar matters more than IOPS, and a tiered mix, flash forward, bulk behind, when the budget covers both.

The rear flex bays are this chassis's quiet advantage: put a small RAID 1 SSD pair back there for the operating system and keep all ten front bays for data. It is the cleanest boot arrangement on the platform, and the BOSS card in the Add-Ons step accomplishes the same separation with a mirrored M.2 pair if you would rather keep the rear bays for a fast scratch tier. Either way, a boot failure and a data failure should never be the same event.

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 R640 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.

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 R640 ships with iDRAC9 either way; the license determines what you can do with it remotely.

The dividing line is the remote console. iDRAC9 Enterprise gives you full remote KVM and virtual media, meaning you can reach the server's screen, mount an ISO, and rebuild an OS from anywhere, even when the operating system is down. Express does not include the remote console, so any problem below the OS level means walking to the rack.

Our recommendation: Enterprise for any server in a datacenter, colo, or remote site. The first time a host hangs at 2 AM and you fix it from your desk instead of driving in, the license has paid for itself. It also adds automated firmware updates, two-factor authentication, and secure erase for compliance environments.

Express is the right call only when the server sits down the hall and someone can physically reach it without ceremony: a single office server, a lab box, a bench unit.

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. All four options are dual hot-swap redundant units, so redundancy and serviceability are already handled. The only decision is wattage, and your build determines it.

2x 495W covers light builds: Silver CPUs, partial memory, a few drives. Peak draw on that configuration runs around 260W, which leaves comfortable headroom.

2x 750W is our default for most builds on this chassis. A balanced configuration (Gold 6230, full memory, 8 SSDs) peaks around 460W, and even a dense VDI build stays within this unit's envelope. When in doubt, this is the answer.

2x 1100W is required territory for top-bin builds: 205W CPUs with full memory and drives peak around 720W, and any GPU configuration belongs here at minimum.

2x 1600W is the headroom play for multi-GPU inference builds or anyone who wants maximum margin.

Two rules of thumb from our deployments. Size for roughly 50 percent load at your expected peak: a PSU running at half capacity runs at peak efficiency, runs cooler, and lasts longer than one working near its limit. And when your build sits between two tiers, size up; the price step is small. Watch the estimated TDP counter at the bottom of the page as you add components to see where your build lands.

One planning note: if a GPU might be added later, buy the 1100W or 1600W now. PSU swaps are hot-swappable and painless, but ordering the right unit the first time is cheaper than a second shipment.
8 Network Cards Required
Component Guide Network Card. These are Dell Network Daughter Cards (NDCs), the R640's dedicated mezzanine slot. The NDC does not consume a PCIe slot, so whatever you pick here leaves your full expansion budget intact. Two decisions: speed and port type.

Speed first. Quad 1 GbE (Intel i350 or Broadcom 5720) is fine for domain controllers, file servers, and branch boxes where the pipe genuinely will not be the bottleneck. For a virtualization host it is a false economy: vMotion, backup traffic, and storage traffic will saturate 1 GbE and you will feel it. The 2x 10 Gb + 2x 1 Gb combo cards are the sweet spot for most builds: 10 Gb for production traffic, 1 Gb ports left over for management and backup networks. Step to quad 10 Gb (Intel X710, X550, or Broadcom 57840S) when you are separating production, vMotion, backup, and management onto their own links, which is the right architecture for any serious cluster.

Then port type, and this is the one that catches people. SFP+ ports take fiber or DAC cables and talk to datacenter-grade switches. The -T models (X540, X550) are 10GBASE-T copper and use ordinary RJ45 cabling. 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. Check your switch before you check this box.

Fair warning that applies to every option here: a 10 Gb card does not deliver 10 Gb unless the switch, cabling, and the device on the other end keep up. The card is the cheapest link in that chain, which is exactly why we say buy the 10 Gb card and let the rest of the network grow into it.

Need 25 GbE for all-flash SAN traffic, or additional PCIe network cards beyond the NDC? Both are in the Add Ons section below, or call 1-800-778-1545 and we will spec the whole path.
9 Operating System Required
Operating System
Component Guide Operating System. We install, license, and burn-in test your OS before shipping, so the server arrives ready to join the network.

Windows Server 2022 is the version to buy today; 2019 belongs on a new build only when an application vendor requires it. Standard licensing is per core with a 16-core base, which is why our processor guidance steers Windows buyers to two 8-core CPUs: the 16-Core license covers that build exactly. Past 16 total cores, you need the 24-Core option. 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. If the server is going into service, buy the license now.

VMware ESXi 8.0 is the one to deploy; 6.7 and 7.0 are past end of support and only belong on hosts joining an existing cluster. Licensing is yours to bring.

Proxmox VE is what we increasingly quote for buyers done with VMware licensing costs: open-source KVM virtualization with clustering and live migration, no per-core math. The R640 is a well-proven Proxmox platform.

Ubuntu Server LTS covers Linux application servers, containers, and web workloads with five years of free updates.

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.
10 Add-Ons Optional
Rails
BOSS Cards
Front Bezels
TPM Modules

WServer Warranty


Dell PowerEdge R640 10-Bay 2.5" Drives [+ RFB Rear Flex Bay] [14th Gen] Detailed Review

The R640 10-Bay + RFB is the refurbished 1U Dell PowerEdge configuration we reach for when a customer needs more than ten drive bays in a single 1U chassis. The Rear Flex Bay (RFB) adds two 2.5" hot-swap drive slots at the rear of the chassis, bringing total drive count to twelve in the same 1U footprint. Pair that with the optional NVMe-capable front backplane and this becomes the most storage-flexible R640 variant: front bays for primary storage, rear bays for boot media, cache tier, or backup target.

The architectural tradeoff matters. The RFB physically occupies space at the rear of the chassis and constrains the secondary riser configuration, which reduces full-height PCIe slot availability. If you do not need the additional rear drives and you want full PCIe slot flexibility for dual NICs plus HBA plus GPU builds, the simpler 10-Bay Standard chassis is the right call. The two extra rear drives come at the cost of one to two PCIe slot positions.

To configure a build, call 1-800-778-1545 or use the quote form below. Every refurbished unit ships under our 180-day warranty with 12+ hour burn-in testing, and volume pricing starts at 5 units.


When 10-Bay + RFB Is the Right Choice

The + RFB chassis earns its place when one of these design patterns applies: you need a dedicated boot drive pair at the rear that does not consume a front bay, you want a separate SSD cache tier physically separated from the primary storage pool, or you have a workload that genuinely needs more than ten drives in a single 1U and a 2U chassis is not an option for rack-density reasons. vSAN ReadyNode builds that want a 12-drive disk group geometry in 1U are another common pattern; the two rear bays carry the cache tier and the ten front bays carry the capacity tier.

If your workload runs fine on ten front bays and you want PCIe flexibility, the 10-Bay Standard chassis is the simpler architecture. If your storage architecture is NVMe-first across all front bays, the 10-Bay NVMe chassis is the right call. The + RFB is specifically for the case where rear drive bays are a hard requirement.


Storage - 10 Front Bays + 2 Rear (RFB)

Ten 2.5" hot-swap front bays on a SAS/SATA backplane (NVMe-capable backplane optional; confirm at quote time) plus two additional 2.5" hot-swap drive bays at the rear of the chassis via the Rear Flex Bay assembly. Total: twelve hot-swap 2.5" bays in 1U.

Common configurations:

  • Front bays for data plus rear bays for boot: Ten front bays available for the primary data pool (SAS SSD, mixed SAS/SATA, or all-flash), with the two rear bays running a mirrored OS boot pair. Cleanest separation of boot from data and keeps every front bay available for production storage.
  • Front bays for primary plus rear bays for cache: Production data on the front backplane with two SSDs in the rear bays acting as a write-through or read cache tier. Useful for tiered storage architectures and the cache-tier slot in vSAN OSA disk groups.
  • Front bays for cold plus rear bays for hot: Higher-capacity spinning disk or near-line storage on the front bays, with high-endurance SAS SSDs in the rear bays for the active working set.

NVMe note: Front-bay NVMe on this chassis depends on the specific backplane ordered. Not every 10-Bay R640 ships with the NVMe-capable backplane. Confirm at quote time before assuming front NVMe is part of your configuration. For NVMe-first deployments where every front bay is NVMe, the 10-Bay NVMe chassis is the correct chassis, not this one.

BOSS alternative for boot: If you prefer the standard BOSS module and ReadyRails II accessories for boot, you free both rear bays for data use. BOSS does not consume a front bay, a rear bay, or a RAID controller channel. We recommend BOSS as the default boot solution on most builds; rear-bay boot is the right call when the customer specifically wants front-accessible boot drives for serviceability or when both rear bays are needed for the cache-tier role and BOSS would conflict with that design.


Storage Controllers at 12-Bay Scale

The 12-bay configuration (10 front plus 2 rear) brings the RAID controller decision into sharper focus than the standard 10-bay. Two key considerations: the rear bays attach via a separate cable harness to the same controller as the front backplane, and write-cache sizing matters more on a 12-drive array than on a 10-drive one.

  • PERC H740P (8 GB NV cache, battery-backed): The default recommendation on this chassis. The 8 GB non-volatile cache is the right size for a 12-drive array, and the battery backup is essential for any production data workload.
  • PERC H730P (2 GB cache, battery-backed): Acceptable for read-heavy or modest write workloads on this chassis, but the 2 GB cache is on the small side for 12 drives if write throughput matters. Quote H740P unless cost is the constraint.
  • PERC H730 (1 GB cache, battery-backed): The 13th-gen-era controller that Dell maintained Mini-PERC slot compatibility for on 14th gen. It works on this chassis and appears frequently on refurbished R640 units as a carryover from prior deployments. Viable but generally a downgrade vs the H730P or H740P on Cascade Lake workloads, and the 1 GB cache is undersized for write activity across 12 drives. Quote it when budget is the driving constraint; otherwise step up.
  • PERC H330 (no cache): Light-workload hardware RAID. Not recommended on a 12-drive array carrying production data.
  • HBA330 (pass-through): For software-defined storage where vSAN, Storage Spaces Direct, or Ceph manages redundancy across the 12-drive array. Front and rear bays are presented as raw devices to the software layer.
  • S140 (software RAID): Dev/test only. Not a production recommendation, particularly at 12-bay scale.

The single-controller-shared-across-front-and-rear architecture means a controller failure takes out access to all twelve drives simultaneously. This is the same single-point-of-failure profile as the 10-Bay Standard, but the higher drive count makes the controller choice more consequential.


Processors

CPU options: Dual 1st Generation Intel Xeon Scalable (Skylake-SP) or 2nd Generation Intel Xeon Scalable (Cascade Lake-SP), socket LGA 3647 on the Intel C620-series chipset. Skylake and Cascade Lake are drop-in compatible. Up to 28 cores per CPU. The platform vocabulary matches the rest of the R640 family.

Our SKU recommendations on this chassis: Intel Xeon Gold 6230 (20 cores, 125W) for the balanced converged-infrastructure build, which is the most common + RFB workload pattern. For vSAN ReadyNode builds with 12 drives feeding a high-VM-density cluster, Gold 6248 (20 cores, 150W) or Gold 6254 (18 cores, 200W) deliver the per-core headroom that vSAN's data services consume.

Heatsink requirement: Any CPU above 150W TDP requires Dell's high-performance heatsink kit and high-performance fan kit. On this chassis specifically, the RFB drive assembly slightly affects rear-chassis airflow, so we are more conservative about top-bin CPU configurations here than on the Standard 10-Bay. Builds with 165W or higher CPUs plus a GPU plus full 12-drive population should run through Dell's thermal restriction tables at quote time; we work this validation in as part of the build review.

Single-socket warning: A single-CPU + RFB build is supported but cuts the platform in half. With one CPU populated only 12 of the 24 DIMM slots are accessible, half the PCIe lanes are inactive, and the NDC and several PCIe slots route through the second CPU and become unavailable. The 12-drive storage layout already implies a workload that justifies dual-socket; single-socket on this chassis is unusual outside of lab and dev contexts.


Memory

Architecture: 24 DDR4 DIMM slots, 12 per CPU across 6 channels at 2 DIMMs per channel. The 6-channel Purley layout is the defining memory feature.

Supported DIMM types:

  • RDIMM: Standard enterprise choice. Up to 64 GB per DIMM, 1.5 TB total at full population.
  • LRDIMM: Up to 128 GB per DIMM, 3 TB total. The path past 1.5 TB without Optane.
  • Intel Optane Persistent Memory (PMem): Cascade Lake L-series CPUs only. Up to 7.68 TB combined with LRDIMM. The vSAN-with-Optane-cache configurations specifically use PMem in App Direct mode and are a known + RFB workload; we walk through the cache-sizing math at quote time when this is in scope.
  • NVDIMM-N: Niche; rarely the right answer in 2026.

Memory speed: DDR4-2933 on Cascade Lake Gold 6200 / 5222 at 1 DPC, DDR4-2666 at full 2 DPC population, DDR4-2666 on all Skylake SKUs. Full 24-DIMM population on the + RFB is common because the workloads that justify a 12-drive 1U (vSAN, dense virtualization, mixed converged) are memory-bandwidth-sensitive. The bandwidth gain from full-channel population is worth the speed step down to 2666.

Mixing rules: Match ranks, capacity, and timing within a channel. We do not quote mixed configurations for production builds; matched-set DIMMs avoid subtle stability issues and simplify later expansion.


Networking and PCIe Expansion

Network Daughter Card (NDC): Dell's NDC mezzanine handles primary networking and does not consume any PCIe slot. This matters more on the + RFB than on most R640 variants because the RFB constrains the rear riser. NDC options:

  • 4x 1 GbE: Entry-tier, not recommended for the converged workloads that typically justify a 12-drive chassis.
  • 2x 10 GbE SFP+ + 2x 1 GbE: Baseline for most + RFB builds carrying enterprise virtualization workloads.
  • 4x 10 GbE SFP+: The right call for vSAN clusters and converged builds where storage traffic and application traffic need separation across links.
  • 2x 25 GbE SFP28: Strongly recommended for 12-drive all-flash builds and vSAN all-flash nodes. 10 GbE can become the bottleneck on a 12-drive SAS SSD array under load.

PCIe expansion - this is where the RFB tradeoff lives: The Rear Flex Bay assembly physically occupies space at the rear of the chassis and constrains the secondary riser. Riser 2 options that deliver a full-height PCIe slot on the Standard 10-Bay chassis are reduced or eliminated on the + RFB depending on the specific RFB SKU. Practical impact: expect 2 PCIe slots, not 3, on most + RFB builds. Confirm exact riser availability at quote time against your specific chassis revision.

Builds requiring three full-height PCIe cards (typical pattern: dual 25 GbE NIC plus external SAS HBA plus GPU) are not a clean fit on this chassis. The 10-Bay Standard is the right answer for that PCIe budget.


GPU Support

The 1U thermal envelope plus the RFB's rear-airflow impact means GPU support on this chassis is more constrained than on the Standard 10-Bay. Single-width low-profile cards (NVIDIA T4 class) are workable for inference workloads where a single GPU plus a 12-drive storage tier is the architecture. Multi-GPU configurations on the + RFB are not something we quote often; the riser constraint and the airflow impact stack up against this combination, and at that point the Dell PowerEdge R740 16-Bay 2.5" 2U is the better answer.

FPGA single-card builds are supported and behave similarly to single-GPU configurations on this chassis. For any GPU configuration on the + RFB, we validate against Dell's thermal restriction tables at quote time.


Management - iDRAC9 Generation

iDRAC9 Enterprise: Required for production deployment. Remote KVM, virtual media, predictive analytics, Group Manager, Quick Sync 2, Silicon Root of Trust. The 12-drive workload pattern usually means the chassis is unattended at the rack, which makes iDRAC9 Enterprise functionally non-negotiable.

Security baseline: Silicon Root of Trust anchors firmware verification in immutable silicon. System Lockdown mode prevents unauthorized firmware changes after deployment. TPM 2.0 module supported and recommended for any deployment subject to NIST 800-171, CMMC, FedRAMP, HIPAA, or PCI DSS compliance frameworks.

Lifecycle Controller and OpenManage Enterprise: Same Dell management plane as the rest of the R640 family. Lifecycle Controller for per-chassis firmware orchestration; OpenManage Enterprise for fleet-scale firmware compliance, configuration drift detection, and warranty status tracking. Worth integrating on any fleet of more than 20 units.


Power and Cooling

The 12-drive RFB configuration adds two drives to the power and thermal budget vs the standard 10-bay. We size PSUs against the actual build, not against the chassis maximum. As a guideline:

  • Light (Silver CPUs, partial RAM, mixed HDD plus SSD): 2x 495W Platinum, peak draw approximately 310W (slightly higher than the Standard chassis due to two extra drives)
  • Balanced (Gold 6230, full RAM, all-SAS SSD across 12 bays): 2x 750W Platinum, peak draw approximately 510W
  • Heavy (Gold 6248, full RAM, all-SSD 12-bay plus GPU): 2x 1100W Platinum, peak draw approximately 740W

Thermal: Eight hot-plug redundant fans standard. The rear drive assembly does affect rear-chassis airflow slightly. For builds with top-bin 165W+ CPUs plus a GPU, the high-performance fan kit is strongly recommended on this chassis specifically, more so than on the Standard 10-Bay. ASHRAE A3 (40C) ambient support is achievable with the high-performance fan kit, but the operating margin is tighter than on the Standard chassis; we are conservative about A3 sign-off on high-TDP + RFB builds.


Physical Specs & Platform Notes

  • Form factor: 1U rack server. 42.8mm H x 434mm W. The RFB assembly extends chassis depth slightly vs the Standard 10-Bay; expect 760-790mm rear-to-front including the RFB rather than the Standard's 735-760mm. Confirm rail kit clearance in shallow racks before order.
  • PCIe expansion: Up to 2 PCIe Gen3 slots on most + RFB SKUs (down from 3 on the Standard 10-Bay). Riser 2 options are reduced by the RFB assembly. Builds requiring 3+ full-height cards do not fit cleanly on this chassis.
  • Parts availability: Strong. The RFB assembly itself is less common in the secondary market than the standard 10-Bay backplane, but Dell parts coverage remains active and refurbished + RFB units are readily available. PERC controllers, NDC cards, fan modules, and PSUs are the same as the rest of the R640 family.
  • Accessories we recommend: Dell LCD bezel (P/N 521RX security bezel, 7M3F1 LCD bezel without security, 9NN24 with security - confirm part at quote time), Dell ReadyRails II sliding rail kit, and the Dell cable management arm (CMA). The CMA matters more on this chassis than most because the rear-bay assembly requires the chassis to be pulled forward for any rear-drive service event.
  • Platform notes: Rear drives are accessed by pulling the chassis forward; they are hot-swap but not as fast to service as the front bays. The single PERC managing both front and rear arrays means controller failure isolates all twelve drives. NDC swap and CPU replacement still require powered-down access. BIOS NVMe bifurcation settings must be configured correctly if NVMe expansion cards are added.

Our Assessment

Where it excels: 12-drive vSAN ReadyNodes where the disk group geometry calls for a cache-tier separation between front and rear bays. Converged-infrastructure hosts that benefit from rear-mounted mirrored boot drives keeping the OS off the data pool. Tiered storage builds with a hot working set on the rear bays and capacity on the front. Workloads that legitimately need 12 hot-swap drives in a 1U where rack density rules out the 2U R740xd.

Where to look instead: If you do not need rear bays, the 10-Bay Standard chassis gives you the full PCIe slot budget intact. If your storage is NVMe-first across all bays, the 10-Bay NVMe is the cleaner architecture. If your build requires three or more full-height PCIe slots, the RFB's riser constraint is the wrong tradeoff and the Standard chassis is the right call. If you need more than 12 total drives or any LFF capacity, the Dell PowerEdge R740xd 12-Bay 3.5" 2U platform is the next step up.

Bottom line: The + RFB is a precision pick. It delivers a specific design point - 12 drives in 1U with the rear-bay separation that some workloads genuinely need - in exchange for a measurable reduction in PCIe slot availability. When the workload calls for the rear bays, this chassis is excellent. When it does not, the Standard 10-Bay is the simpler and more flexible build. We ask the workload question first and pick the chassis from the answer.


Where the R640 Fits in 2026

The R640 family is 2 to 3 generations behind current Dell production (R650 / R660). The 10-Bay Standard page covers the generational ladder and the support status in full. + RFB specifically: the rear-bay design point carried forward into the R650 and R660 generations with similar architecture, so the migration path is straightforward when the workload eventually justifies the platform refresh. For 2026 procurement, the + RFB earns its place when 14th gen fleet standardization, budget, or vendor certification keeps the workload on R640 hardware.


Honest Limitations

  • RFB constrains PCIe slot availability. The Rear Flex Bay physically occupies riser space, reducing full-height PCIe slot count vs the Standard 10-Bay. Builds requiring 3+ full-height PCIe cards should look at the Standard chassis instead.
  • NVMe backplane is not universal on the 10-bay front. Front NVMe support requires the correct backplane SKU. Confirm at quote time. For NVMe-first storage, the 10-Bay NVMe chassis is the correct configuration.
  • 2 rear bays, not 4. If you need more than 2 rear drives or want all-rear NVMe, the R640 chassis cannot deliver. The R740xd 2U platform is the next step up for high rear-bay counts.
  • Rear bays share the controller with the front backplane. All twelve drives present to a single PERC or HBA. Separate RAID groups can be created across the front and rear bays, but the controller is shared and is a single point of failure for the whole array.
  • Rear-bay service requires chassis pull-forward. The rear bays are hot-swap but service is slower than front-bay swap because the chassis must be pulled forward on its rails to access them. The CMA helps; bring it on every + RFB build.
  • Tighter thermal margin on top-bin builds. The RFB assembly affects rear airflow. High-TDP CPUs plus GPU on this chassis run closer to the thermal envelope than equivalent builds on the Standard 10-Bay.
  • Full R640 family limitations apply. PCIe Gen3, 14th gen platform, 1U thermal envelope. See the R640 10-Bay Standard page for the full Honest Limitations list including the generational positioning vs R650 / R660.

Workload Fit

This server is right for Consider alternatives for
12-drive storage in 1U (10 front + 2 rear) Builds needing 3+ full-height PCIe slots
Rear-mounted mirrored boot drives Native front-bay NVMe across all bays
Front data plus rear cache tier architectures 4+ rear drive requirements (consider R740xd)
Tiered storage with hot/cold separation GPU compute or AI training workloads
vSAN nodes wanting a 12-drive disk group Greenfield deployments needing PCIe Gen4 / DDR5

Where to Look Instead

  • Don't need rear bays? The R640 10-Bay 2.5" Standard Chassis is the primary R640 build with the full PCIe slot budget intact.
  • NVMe-first storage architecture? The R640 10-Bay 2.5" NVMe replaces SAS/SATA with PCIe-attached NVMe across all front bays.
  • Compute-first, fewer drives, wider thermal envelope? The R640 8-Bay 2.5" is the right call when drive count is not the constraint but top-bin CPU thermals are.
  • Need 4+ rear drives or higher total drive count in 2U? The Dell PowerEdge R740xd 12-Bay 3.5" is the 2U high-density companion to the R640 1U lineup; up to 24 SFF front bays plus rear bays available.
  • vSAN HCI pre-validated? The R640 VxRail E560F is the vSAN-certified version of the 10-bay chassis for VxRail cluster expansion.
  • HPE-side equivalent? The HPE ProLiant DL360 Gen10 10-Bay 2.5" is the direct counterpart on the same Intel Purley platform. HPE's rear-bay equivalent is the universal media bay on the DL380 Gen10.
  • Step up to 15th gen? The Dell PowerEdge R650 8-Bay 2.5" is the Ice Lake-SP successor with PCIe Gen4 and rear-bay design points that carried forward from the R640.
  • Step down to 13th gen for budget? The Dell PowerEdge R630 10-Bay 2.5" is the 13th gen predecessor for budget-constrained refurbished builds.

Ready to Configure?

Tell us your workload, target storage layout across the 10 front and 2 rear bays, NVMe vs SAS/SATA front backplane preference, target memory footprint, NDC choice, and quantity. Our account team returns a fully specced build with formal pricing within 24 hours, including a validated PCIe slot allocation that accounts for the RFB's riser constraint, a confirmed front-backplane SKU, and thermal validation on top-bin CPU configurations. Every refurbished unit ships with the Wholesale Servers 180-day warranty and 12+ hour burn-in testing, and volume pricing starts at 5 units. Call 1-800-778-1545 or use the quote form below.

Estimated TDP Draw: 0W
0 lines

Dell PowerEdge R640 10-Bay 2.5" + RFB Refurbished

10-Bay 2.5" + RFB

Subtotal $375.40
Estimated TDP Draw 0W
Build total $375.40

Choosing Memory for Your Dell PowerEdge R640

Fill all 24 slots. This is our standing recommendation on the R640 and we push it hard. There are two benefits. The first is performance: full population at 2 DPC keeps all twelve memory channels working at maximum depth, and on memory-bandwidth-sensitive workloads like VDI and high-density virtualization it consistently outperforms a partial population running at higher clock. The second is cost per gigabyte: 24 smaller modules hit the same capacity for meaningfully less than 12 larger ones, and 768 GB as 24× 32 GB is the best cost per GB in the lineup. We know the instinct is to leave a few slots open for a later upgrade, and on paper that is sensible. In practice it rarely plays out well. One year from now you are hunting for DIMMs that match the rank and speed of what is already in the box, and a mismatch either clocks the whole server down or refuses to train. The server is in production by then, so adding memory means a maintenance window you have to schedule and justify. Buy the capacity now, in one matched set. If you outgrow it, the right move is usually another node rather than another DIMM. 384 GB (24× 16 GB) for file, print, domain controller duty through a modest virtualization host or SQL Server instance 768 GB (24× 32 GB) is our default for a general-purpose ESXi or Hyper-V host 1.5 TB (24× 64 GB) for dense consolidation, VDI, or in-memory workloads Past 1.5 TB you are into 128 GB LRDIMMs, which reach 3 TB but cost substantially more per GB. Most buyers are better served by a second host.

Choosing Your iDRAC License

The R640 ships with iDRAC9 either way; the license determines what you can do with it remotely. The dividing line is the remote console. iDRAC9 Enterprise gives you full remote KVM and virtual media, meaning you can reach the server's screen, mount an ISO, and rebuild an OS from anywhere, even when the operating system is down. Express does not include the remote console, so any problem below the OS level means walking to the rack. Our recommendation: Enterprise for any server in a datacenter, colo, or remote site. The first time a host hangs at 2 AM and you fix it from your desk instead of driving in, the license has paid for itself. It also adds automated firmware updates, two-factor authentication, and secure erase for compliance environments. Express is the right call only when the server sits down the hall and someone can physically reach it without ceremony: a single office server, a lab box, a bench unit. 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.

Choosing Your Power Supply

All four options are dual hot-swap redundant units, so redundancy and serviceability are already handled. The only decision is wattage, and your build determines it. 2x 495W covers light builds: Silver CPUs, partial memory, a few drives. Peak draw on that configuration runs around 260W, which leaves comfortable headroom. 2x 750W is our default for most builds on this chassis. A balanced configuration (Gold 6230, full memory, 8 SSDs) peaks around 460W, and even a dense VDI build stays within this unit's envelope. When in doubt, this is the answer. 2x 1100W is required territory for top-bin builds: 205W CPUs with full memory and drives peak around 720W, and any GPU configuration belongs here at minimum. 2x 1600W is the headroom play for multi-GPU inference builds or anyone who wants maximum margin. Two rules of thumb from our deployments. Size for roughly 50 percent load at your expected peak: a PSU running at half capacity runs at peak efficiency, runs cooler, and lasts longer than one working near its limit. And when your build sits between two tiers, size up; the price step is small. Watch the estimated TDP counter at the bottom of the page as you add components to see where your build lands. One planning note: if a GPU might be added later, buy the 1100W or 1600W now. PSU swaps are hot-swappable and painless, but ordering the right unit the first time is cheaper than a second shipment.

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.