5 things to make you go hmmm about the M5 server
It’s not unusual for vendors to wax lyrical about their latest and greatest offerings. But there are plenty of reasons to love unified computing. A full stack of data centre technologies combined into a system that’s pre-engineered, tested and supported to operate as a whole.
Following the launch of our new M5 servers, here’s five more reasons why you should consider putting the power of unified computing to work for your business:
- Don’t call it a comeback
Cisco entered the server market 8 years ago with the introduction of our Unified Computing System (UCS). Some said we were crazy. But our solution proved a true game changer. One that was 100% tailored to the needs of our customers:
- Fully assembled stacks and appliances rigorously tested, documented and certified as Validated Designs for all the leading business applications
- Delivering the best possible performance and flexibility
- With single-number solution support, end to end
Every now and again expensive gadgets give birth to important technologies. The graphics processing units (GPUs) in video-game machines for example turned out to be perfect for powering artificial-intelligence (AI) systems.
AI, machine learning and deep learning projects are on the radar for many businesses, while demand for high performance computing (HPC), virtualised desktops (VDI) and workstation applications has seen GPUs become ubiquitous in the data centre.
One of the biggest asks from our customers has been to focus on accelerators to ensure the best use of flash memory and GPUs. We’ve responded with broad support for NVMe across our server portfolio.
The result is a 5X or better capacity increase on our M5 C-Series rack servers, plus leading GPU density in our B-Series blades, which offer the industry’s only half-width blade with dual GPU support.
The 5th generation of UCS, our M5 servers provide unbeatable performance and hold 9 world record benchmarks. They feature new Intel CPUs that bring increased core counts and faster memory. In Cisco lab testing comparing M4 to M5 generation machines, we’ve seen performance improvements up to 89%.
While all vendors have access to Intel’s new processors, only UCS servers are able to realise their full potential. Why? Because only Cisco gives our servers the ‘unified’ dimension.
We integrate the CPUs into one system, providing the perfect resource distribution to exploit the full performance of the CPUs.
Combining faster CPUs, more memory and accelerators increases the adaptability of our 2-way systems for workloads and accelerates the time to ROI when you need to upgrade your environment.
- Plug in baby
The best way to prepare for the unexpected is to have an architecture that adapts quickly and efficiently to change. This is an area in which the UCS system excels.
Just like the S-Series Storage server we announced recently, M5 servers offer the same plug-and-play modularity and seamless integration Cisco users enjoy across our entire portfolio of converged and hyperconverged solutions.
The M5 generation has all the hooks in place for integration with our Starship project, which is launching later this year in the form of a SaaS platform. It will deliver cloud-based management of your data centre and, through analytics, the capability for your systems to constantly optimise themselves.
- It’s a kind of magic
The competitive advantage you realise with UCS lies not only with the portfolio itself, but also in the operating model.
From the software embedded in the custom ASICs of our Virtual Interface Cards, through the UCS Management software embedded in the system, to the orchestration software in Cisco Enterprise Cloud Suite (ECS), it’s software that creates the magic of UCS: fully abstracted and truly software-defined infrastructure from top to bottom.
M5 servers come with the new UCS Director 6.5, which includes our new Workload Optimization Manager. These take automation beyond basic configuration and deployment of infrastructure components, enabling your IT team to complete 80% of operational tasks from a single console and match workload demand with the right amount of resources automatically.
Ultimately, you can deploy a multi-cloud strategy with uncompromising performance, while driving more efficiency in on-premises IT. And while components such as servers and networking, etc. can come and go, the operating model stays consistent over time.
Now that’s magic!Tags: