Daily Archives: March 30, 2010

Technical Details on the IBM HX5 Blade Server (UPDATED)

(Updated 4/22/2010 at 2:44 p.m.)
IBM officially announced the HX5 on Tuesday, so I’m going to take the liberty to dig a little deeper in providing details on the blade server. I previously provided a high-level overview of the blade server on this post, so now I want to get a little more technical, courtesy of IBM.  It is my understanding that the “general availability” of this server will be in the mid-June time frame, however that is subject to change without notice.

Block Diagram
Below is the details of the actual block diagram of the HX5.  There’s no secrets here, as they’re using the Intel Xeon 6500 and 7500 chipsets that I blogged about previously.

As previously mentioned, the value that the IBM HX5 blade server brings is scalability.  A user has the ability to buy a single blade server with 2 CPUs and 16 DIMMs, then expand it to 40 DIMMs with a 24 DIMM MAX 5 memory blade.  OR, in the near future, a user could combine 2 x HX5 servers to make a 4 CPU server with 32 DIMMs, or add a MAX5 memory DIMM to each server and have a 4 CPU server with 80 DIMMs. 

The diagrams below provide a more technical view of the the HX5 + MAX5 configs. Note, the “sideplanes” referenced below are actualy the “scale connector“.  As a reminder, this connector will physically connect 2 HX5 servers on the tops of the servers, allowing the internal communications to extend to each others nodes.  The easiest way to think of this is like a Lego .  It will allow a HX5 or a MAX5 to be connected together.  There will be a 2 connector, a 3 connector and a 4 connector offering. 

 (Updated) Since the original posting, IBM released the “eX5 Porfolio Technical Overview: IBM System x3850 X5 and IBM BladeCenter HX5” so I encourage you to go download it and give it a good read.  David’s Redbook team always does a great job answering all the questions you might have about an IBM server inside those documents. 

If there’s something about the IBM BladeCenter HX5 you want to know about, let me know in the comments below and I’ll see what I can do.

Thanks for reading!

Dell M910 Blade Server – Based on the Nehalem EX

Dell appears to be first to the market today with complete details on their Nehalem EX blade server, the PowerEdge M910. Based on the Nehalem EX technology (aka Intel Xeon 7500 Chipset), the server offers quite a lot of horsepower in a small, full-height blade server footprint.

Some details about the server:

  • uses Intel Xeon 7500 or 6500 CPUs
  • has support for up to 512GB using 32 x 16 DIMMs
  • comes standard two embedded Broadcom NetExtreme II Dual Port 5709S Gigabit Ethernet NICs with failover and load balancing.
  • has two 2.5″ Hot-Swappable SAS/Solid State Drives
  • 3 4 available I/O mezzanine card slots
  • comes with a Matrox G200eW w/ 8MB memory standard
  • can function on 2 CPUs with access to all 32 DIMM slots

Dell (finally) Offers Some Innovation
I commented a few weeks ago that Dell and innovate were rarely used in the same sentence, however with today’s announcement, I’ll have to retract that statement. Before I elaborate on what I’m referring to, let me do some quick education. The design of the Nehalem architecture allows for each processor (CPU) to have access to a dedicated bank of memory along with its own memory controller. The only downside to this is that if a CPU is not installed, the attached memory banks are not useable. THIS is where Dell is offering some innovation. Today Dell announced the “FlexMem Bridge” technology. This technology is simple in concept as it allows for the memory of a CPU socket that is not populated to still be used. In essence, Dell’s using technology that bridges the memory banks across un-populated CPU slots to the rest of the server’s populated CPUs. With this technology, a user could start of with only 2 CPUs and still have access to 32 memory DIMMs. Then, over time, if more CPUs are needed, they simply remove the FlexMem Bridge adapters from the CPU sockets then replace with CPUs – now they would have a 4 CPU x 32 DIMM blade server.

Congrats to Dell. Very cool idea. The Dell PowerEdge M910 is available to order today from the Dell.com website.

 Let me know what you guys think.

Details on Intel’s Nehalem EX (Xeon 7500 and Xeon 6500)

Intel is scheduled to “officially” announce today the details of their Nehalem EX CPU platform, although the details have been out for quite a while, however I wanted to highlight some key points.

Intel Xeon 7500 Chipset
This chipset will be the flagship replacement for the existing Xeon 7400 architecture.  Enhancements include:
•Nehalem uarchitecture
•8-cores per CPU 
•24MB Shared L3 Cache
• 4 Memory Buffers per CPU
•16 DIMM slots per CPU for a total of 64 DIMM slots supporting up to 1 terabyte of memory (across 4 CPUs)
•72 PCIe Gen2 lanes
•Scaling from 2-256 sockets  
•Intel Virtualization Technologies

Intel Xeon 6500 Chipset
Perhaps the coolest addition to the Nehalem EX announcement by Intel is the ability for certain vendors to cut the architecture in half, and use the same quality of horsepower across 2 CPUs.  The Xeon 6500 chipset will offer 2 CPUs, each with the same qualities of it’s bigger brother, the Xeon 7500 chipset.  See below for details on both of the offerings.

Additional Features
Since the Xeon 6500/7500 chipsets are modeled off the familiar Nehalem uarchitecture, there are certain well-known features that are available.  Both Turbo Boost and HyperThreading have been added to the and will provide users for the ability to have better performance in their high-end servers (shown left to right below.)

HyperThreading

Memory
Probably the biggest winner of the features that Intel’s bringing with the Nehalem EX announcement is the ability to have more memory and bigger memory pipes.  Each CPU will have 4 x high speed “Scalable Memory Interconnects” (SMI’s) that will be the highways for the memory to communicate with the CPUs.  As with the existing Nehalem architecture, each CPU has a dedicated memory controller that provides access to the memory.  In the case of the Nehalem EX design, each CPU has 4 pathways that each have a Scalable Memory Buffer, or SMB, that provide access to 4 memory DIMMs.  So, in total, each CPU will have access to 16 DIMMs across 4 pathways.  Based on the simple math, a server with 4 CPUs will be able to have up to 64 memory DIMMs.  Some other key facts:
• it will support up to 16GB DDR3 DIMMs
•it will support up to 1TB with 16GB DIMMS
•it will support DDR3 DIMMs up to 1066MHz, in Registered, Single-Rank, Dual-Rank and Quad-Rank flavors.

Another important note is the actual system memory speed will depend on specific processor capabilities (see reference table below for max SMI link speeds per CPU):
•6.4GT/s SMI link speed capable of running memory speeds up to 1066Mhz
•5.86GT/s SMI link speed capable of running memory speeds up to 978Mhz
•4.8GT/s SMI link speed capable of running memory speeds up to 800Mhz

Here’s a great chart to reference on the features across the individual CPU offerings, from Intel:

Finally, take a look at some comparisons between the Nehalem EX (Xeon 7500) and the previous generation, Xeon 7400:

That’s it for now.  Check back later for more specific details on Dell, HP, IBM and Cisco’s new Nehalem EX blade servers.