New UCS Servers Expand the Portfolio: More Third Generation Fabric Computing from Cisco
Back in March we announced the third generation of UCS, with significant expansions to the I/O and systems management capabilities of the platform as well as a new lineup of servers. This month we’re continuing to expand the UCS server lineup with the addition of four new models. The latest batch of M3 systems are comprised of three Intel Xeon “EN” class machines (E5-2400 series processors) as well as a four socket “EP” (E5-2600 series) blade server. Specifically: the UCS B22 and B420 M3 blades and the C22 and C24 M3 rack servers. These new servers round out the UCS portfolio with an even stronger set of products optimized for scale-out and light general-purpose computing as well as a new price/performance 4S category in the mid-range.
If you prefer watching than reading , here is a nice conversation between Intel Boyd Davis , VP & GM, Data Center Infrastructure group, Cisco Jim McHugh, VP UCS Marketing, and Scott Ciccone, Sr. Product Marketing Manager, highlighting the key benefits of these new models.
To figure out how these fit in, let’s step back and consider the broader evolution of server technology in play here:
1) Cisco has made server I/O more powerful and much simpler.
One of the key differentiators of UCS is the way in which high-capacity server network access has been aggregated through Cisco Virtual Interface Cards and infused with built-in high performance virtual networking capabilities. In “pre-UCS” server system architectures, one of the main design considerations was the type and quantity of physical network adapters required. Networking, combined with computing sockets/cores/frequency/cache, system memory, and local disk are historically the primary resources considered in the balancing act of cost, physical space and power consumption, all of which are manifested in the various permutations of server designs required to cover the myriad of workloads most efficiently. Think of these as your four server subsystem food groups. Architecture purists will remind us that everything outside the processors and their cache falls into the category of “I/O” but let’s not get pedantic because that will mess up my food group analogy. In UCS, I/O is effectively taken off the table as a design worry because every server gets its full USRDA of networking through the VIC: helping portions of bandwidth, rich with Fabric Extender technology vitamins that yield hundreds of Ethernet and FC adapters through one physical device. Gone are the days of hemming and hawing over how many mezz card slots your blade has or how many cards you’re going to need to feed that hungry stack of VM’s on your rack server. This simplification changes things for the better because it takes a lot of complication out of the equation.
2) Processors have become incredibly powerful, bringing new choices for design optimization.
While Cisco has been hard at work making server networking better, our friends at Intel haven’t been asleep at the wheel. With the advent and advance of multi-core processing, the workhorse two socket server has become a real performance monster. In fact, for some applications the amount processing power required, relative to the other food groups I mentioned above, is outstripped by the capabilities of the mainstream processor family, which in today’s incarnation is Intel’s Xeon E5 2600 series. In response to this phenomenon, Intel subdivided the Xeon lineup to include a new “EN” class of processors, the E5-2400 series, which ease back on the gas pedal of Moore’s law for designs that don’t require as much processing power in relation to local storage and memory. This creates a new class of cost & performance optimized systems for lighter workloads or for storage heavy systems (think big data) at the entry end of the portfolio. Three of our new UCS M3 series systems fall in this category: the B22, C22 and C24. At the same time, Intel has brought four socket server options, formerly the province of the mission critical, “EX” end of the spectrum, down into the mainstream. An example of this is our new UCS B420 blade. So if you want four socket core count and performance but don’t necessarily need the comprehensive RAS features of an EX class system, you now have a price/performance optimized solution for that need. So the trend here is that as processors continue to become more powerful, new options are emerging for customers to optimize on the spectrum of cost, performance and RAS features.
3) New options are blurring the boundaries of main memory and local disk.
Let’s save this topic for another post but the disk and memory food groups are starting to merge with SSD, flash and the rise of storage options far more interesting than the rotating media of yore. This means fewer system varieties can be offered with storage options that span a wider range of needs.
4) Thanks to the UCS Manager, the shape of the server boxes doesn’t make life complicated any more.
One of the things announced in March are advancements that Cisco has made in eliminating the management silos that existed between rack and blade server architectures “pre-UCS.” Historically, server manufacturers have had to look at their rack and blade server lines as nearly independent, and build out a full spectrum of options in both form factors. This was because customers often needed to manage them differently and if they deployed blades for density and power concerns, might opt to go all-in on blades and need to find every option under the sun in the blade lineup. With UCS, both rack and blade servers exist in a common pool of resources. Administrators choose the balance of the four food groups they need and manage them all the same way, regardless of whether it happens to come as a rack server or a blade server. Sometimes, you just need a rack server to fit all the components you need, and trying to make a blade out of it isn’t efficient. Example: our new C24 M3, which offers 24 drive bays. With UCS, what is crucially new and improved is that the shape of the box doesn’t dictate management policy anymore and this eliminates the need for superfluous server design types created by rigidly parallel lines of blade and rack servers.
I wanted to step through these factors to bring out the point that, on the design front, you can cover a lot more workload and environmental requirements with a lot fewer server designs these days. Where legacy architecture could spawn an overly complex portfolio of 30+ servers (or custom designs for those that could afford it) UCS can now address the space capably with less than half that number. This is good news for customers who can now get more done with fewer machine types (and a lot fewer machines.) With the addition of these four new servers, Cisco has filled out a “right sized” portfolio of machines that answer the broadest set of workload requirements with more simplicity than ever before. The full specs and “what it’s best for” information for these new models is up and posted on our UCS product pages so please take a look. I’ll come back with another post on some of thinking that led us to the design optimizations you see in these systems and the lineup today.