Review: Dell EMC PowerEdge MX740c Server Sled

May 20, 2021 0 By Lorena Mejia

The Dell EMC PowerEdge MX740c (SHOP HERE) is a compute sled that can support two Intel Xeon Scalable processors, plus more than 3TB when outfitted using persistent memory modules. Of course, we can’t talk about the MX740c server sled without looking at the whole package: the Dell EMC PowerEdge MX7000 chassis.

The PowerEdge MX7000 chassis is a 7U enclosure that supports the power, cooling, centralized management, and a number of high-performance I/O modules. As a modular system, it’s quick to scale either in today’s software-defined data center or your enterprise business. It features redundant power, networking, management, and storage fabric. There are two compute sleds to choose from at this time. Either our single width Dell PowerEdge MX740c featuring dual processors or the double-width MX840c with 4 sockets and support for over 6TB of memory. Sprinkle in some Data Centric Persistent memory modules (DCPMM) and you get even more memory. There’s even a single-width 16x bay SAS drive storage module, the PowerEdge MX5016s, so there are options for configuration. It can be outfitted for dense virtual workloads and in-memory database, plus mission-critical applications— all within the same chassis.



Dell EMC PowerEdge MX7000 chassis with MX740c server sled

With the 7U MX7000 chassis, you can place up to eight MX740c server sleds, up to 4x double-width MX840 nodes, or a combination of the two depending on the workloads you want to address.

You can also incorporate a maximum of 7x single-width MX5016s storage modules. There’s even mention of supporting up to 16x half-height sleds, but these have not been released yet. This all sounds very familiar to the M1000e enclosure at 11U, and you could get a few more sleds in the M1000e.

Dell PowerEdge MX7000 with 7 MX5016s stacked

The MX7000 does take up less space and theoretically with advances in compute memory and storage, you should be able to get the same if not better performance even in this smaller chassis. Not to mention, the M1000e is approaching the end of its life! Dell has indicated the MX7000 blade server is guaranteed to support the next three microprocessor generations, so this system does have some legs for the long run. The absence of a backplane is also a calculation on Dell’s part to support the latest technologies going forward. The MX7000 in general is described as a kinetic, modular, composable. Add a few more adjectives from the thesaurus and we’re good to go!

Dell M1000e versus MX7000

With no backplane in the MX7000 chassis, server sleds and nodes connect to the fabric and networking modules through expansion cards. This is a good thing, right? It is, because it’s one less thing to break. You can also switch out the storage and networking fabric modules without worrying about feature compatibility with the non-existent backplane. Bottom line, our MX740c sled connects directly to the fabric interface with redundancy through a direct orthogonal connection. This basically means they connect at right angles to each other with the vertically mounted server sleds in front and the horizontally mounted Fabric and storage modules in back. It sounds much cooler than it is. Kudos to marketing!

No backplane on MX7000 chassis

We keep getting sidetracked by the enclosure because a lot of the functionality resides IN the chassis. Our MX740c sled features two 2nd generation Intel Xeon Scalable processors with up to 28 cores and 56 threads. And no, you can’t run those 9200 series Xeon Scalable CPUs in this blade or any others on this chassis. We also don’t think there is any plan to include them.

No compatible Xeon 9200 series CPUs on Dell PowerEdge MX740c sled

Two Xeons each support 12x memory module slots and 6x memory channels each at speeds of up to 2933MHz, but only with second generation Intel Xeon Scalable processors. First generation Xeon Scalable CPUs will only support a maximum memory speed of 2666MHz.

There are 24x active slots in a two-processor configuration. You have a choice of RDIMMs with a maximum capacity of 1.5TB or LRDIMMs at up to 3TB. You can also add a maximum of 12x data centric persistent memory modules, like Intel’s Optane modules.

RDIMM and LRDIMM maximum memory capacity on Dell PowerEdge MX740c

With persistent memory modules this sled will support up to 7.68TB! That would be 6.1TB Optane and the rest RDIMMs or LRDIMM memory modules. You can even install up to 192GB of NVDIMMs to support in-memory data base applications. Unlike NVDIMMS, Optane will do the trick too and don’t require a battery in the event of a power failure.

RDIMM and Intel Optane on Dell PowerEdge MX740c

Storage is on the front of the sled with 6x 2.5-inch drive bays that can support SAS, SATA HDDs/SSDs or U.2 PCIe NVMe drives. However, if you do go with NVDIMM you will only have 4x drives bays available because of the additional battery required to protect the DIMMs during a power loss.

Dell PowerEdge MX740c compatible with SAS, SATA, NVMe

To preserve your upfront storage and increase bootup speed, choose an M.2 Boot optimized Storage Solution or BOSS. The BOSS supports two M.2 drives and can be used to host your OS in mirror mode for redundancy. You can install a dual SD card module (IDSDM) for a hypervisor that features two SD cards on one side and a flash drive on the other for additional storage that can be used by iDRAC to store system firmware updates and patches. There is a lot more you can do with iDRAC and Dell has a 363-page guide for that. You can use iDRAC to deploy, monitor, manage, configure, update, and troubleshoot. Both the BOSS and IDSDM are common features on other stand-alone PowerEdge servers.

Of course, if you go with NVMe drives up front, there are PCIe connectors, but you will need a PowerEdge RAID Controller (PERC) for SAS drives. You still have quite a few options for support of both local and external drives. The PERC and the BOSS both have a dedicated PCIe slot on the system board as does the mini mezzanine connector for connection to the storage Fabric C.

Dell PERC controller

There are two redundant MX9002m management modules on the back of the chassis to support the embedded management solution. In this case, it’s the Dell OpenManage Enterprise Modular Edition (OME). You get enterprise level management features that are designed for modern DATA Center applications. The PowerEdge FX and the VRTX enclosures all rely on the Chassis Management Controller (CMC) for compatibility with third-party vendors. Open Manage Enterprise Modular Edition taps into the integrated Dell Remote Access Controller (iDRAC 9.0) modules on the server boards. QuickSync II and OpenManage Mobile are also options for at-chassis management of the system. Quick Sync should be ordered when it is purchased, as it’s not available for an in-the-field add-on. There’s even an augmented reality OpenManage Mobile!

Fabric C on Dell PowerEdge MX740c server sled

Two I/O mezzanine cards at the rear of the sled are PCIe 3.0 compatible and have dual ports for redundancy connecting to I/O Fabrics A and B. A third PCIe slot is a mini-mezzanine slot and connects to the storage Fabric C. The Mezzanine cards connect directly to the I/O Fabric modules at the back of the MX7000 chassis, which can be configured for different network connections speeds using different modules. Same for the pair of storage specific bays for Fabric C. Fabric C is currently for use in SAS or Fiber Channel storage connectivity, but also future-ready to support next gen technologies. The MX5016s storage sled uses Fabric C to communicate directly with the I/O modules.

The lack of a backplane means to support future technologies, all that will be required is to replace the I/O module of which there are currently 4x, and replace the I/O mezzanine cards in the sleds, if necessary. The server sleds on the other hand do have a few storage backplanes depending on your storage requirements.

Rear of MX7000 Chassis with slots for Fabrics

There are only two compute sleds at the moment with the MX740c, the MX840c, and a single MX5016s storage module. There are currently 4x I/O modules to choose from. You can bet Dell will be adding more blades utilizing AMD EPYC CPUs plus optional I/O and storage sleds featuring all NVMe drives. This is a system that not only delivers great performance, but has even more potential to support next gen hardware and applications. It’s ideal for a variety of workloads including dense virtualization, software-defined workloads, including Software-Defined Storage (SDS) and Hyperconverged infrastructures (HCI), plus big data environments.

Front panel on the Dell PowerEdge MX7000 chassis

Are you interested in the Dell EMC PowerEdge MX740c Server sled or maybe the full enclosure? IT Creations is a partner with Dell EMC and we have these PowerEdge MX740c compute sleds in stock! Check out our website for more servers or components.

We carry many server nodes and chassis, including the Dell PowerEdge FC640 node for the FX2!