Last week at IDF, two Intel technologists spoke about different fixes to the problem of compute capacity outpacing the typical server's ability to handle it.
For the past 5 years, x86 CPU makers have boosted performance by adding more cores within the processor. That's enabled servers with ever-increasing CPU horsepower. RK Hiremane (speaking on "I/O Innovations for the Enterprise Cloud") says that that I/O subsystems haven't kept pace with this processor capacity, moving the bottleneck for most applications from the CPU to the network and storage subsystems.
He gives the example of virtualized workloads. Quad-core processors can support the compute demands for a bunch of virtual machines. However, the typical server I/O subsystem (based on 1Gb Ethernet and SAS hard drives) gets overburdened by the I/O demand of all those virtual machines. He predicts an immindent evolution (or revolution) in server I/O to fix this problem.
Among other things, he suggests solid-state drives (SSDs) and 10 gigabit Ethernet will be elements of that (r)evolution. So will new virtualization techniques for network devices. (BTW, some of the changes he predicts are already being adopted on ProLiant server blades, like embedded 10GbE controllers with "carvable" Flex-10 NICs. Others, like solid-state drives, are now being widely adopted by many server makers.)
Hold on, said Anwar Ghuloum. The revolution that's needed is actually in programming, not hardware. There are still processor bottlenecks holding back performance; they stem from not making the shift in software to parallelism that x86 multi-core requires.
He cites five challenges to mastering parallel programming for x86 multi-core:
* Learning Curve (programmer skill sets)
* Readability (ability for one programmer to read & maintain other programmer's parallel code)
* Correctness (ability to prove a parallel algorithm generates the right results)
* Scalability (ability to scale beyond 2 and 4 cores to 16+)
* Portability (ability to run code on multiple processor families)
Anwar showed off one upcoming C++ library called Ct from RapidMind (now part of Intel) that's being built to help programmers solve these challenges. (Intel has a Beta program for this software, if you're interested.)
To me, it's obvious that the "solution" is a mix of both. Server I/O subsystems must (and are) improving, and ISVs are getting better at porting applications to scale with core count.
Today, we launched some new and updated technologies that I thought you might find useful. Here's the news in language as plain as it gets in IT.
Our Virtual Connect interconnect porfolio was updated in two ways.
- First, the Fibre Channel module was updated to 8Gb performance. That means you get more connections at higher speeds. The server side NPIV feature also increased support for up to 255 WWNs (world wide names) so you can support more VM's per module. This also means lower network costs per virtual machine. We also published a new Virtual Connect Fibre Channel cookbook that most folks find very useful. Get all the technical tips, tricks and best practices to help you set it up right and get the most out of Virtual Connect.
- Second, the the Virtual Connect Flex-10 module now supports multi-enclosure stacking which means you can connect up to 4 enclosures into one group and cut down the number of Ethernet cables to as few as two per rack. The number of total domains that can be managed together was also increased to 200 (or 800 with stacking). This makes it easier to manage more connections across a big environment in a single view.
The other news today was the update of our 4 processor, ProLiant BL685c blade server. Basically, the new G6 version dobules the supported memory per blade versus the older, G5 version. That's 32 DIMM sockets and 256Gb of memory per blade. We made this move to remove a key bottleneck -- memory performance, to btake advantage of the new quad-core AMD processors and to support more VM's or big applications per blade.
If you have more questions, leave them below and our experts will fill you in on all the technical details. In the future, visit this site to keep up to date with other blade news.
As many of our readers know, HP Virtual Connect is not only a technology that enables them to connect more while spending less, but that it is also an industry award winning technology.
One of our newest members of the Virtual Connect family is HP Virtual Connect Enterprise Manager. This fine piece of engineering allows IT teams to simplify the set-up and ongoing management of server I/O for up to 800 BladeSystem enclosures or put another way, up to 12,800 servers! Enabling system administrators the ability to manage the connectivity of up to 12,800 servers goes a long way to making life simpler and less expensive for many of our customers.
Apparently we’re not the only ones thinking that Virtual Connect Enterprise Manager, or VCEM, is pretty spiffy. It turns out that VCEM was just named as a first place winner of the Virtualization Journal Readers' Choice Awards. Read on for details!SYS-CON Media, the world's leading i-technology media and events company, has announced the winners of its Virtualization Journal Readers' Choice Awards. SYS-CON's Readers' Choice Awards, also known as the “Oscars of the Software Industry,” has been one of the most prestigious industry award programs for over a decade. The Virtualization Journal Readers' Choice Awards recognize the best tools, solutions, and platform offerings in 18 categories. Winners were selected through reader-submitted nominations, followed by online voting at SYS-CON Media's world-leading Virtualization Journal online magazine site.
Best Grid Virtualization: Winner - HP Virtual Connect Enterprise Manager, Hewlett-Packard
On Monday, April 20th, we announced a new Virtual Connect family member and expanded capabilities for all Virtual Connect products. We’ve see a great deal of momentum building behind virtualization and infrastructure convergence - and these enhancements will help our customers better meet their goals.
When customers put applications onto fewer servers with virtualization, they increase the needed density of both data and storage networking. So, customers not only need server virtualization, but they also need to virtualize and converge server I/O. Last November, we introduced the HP Virtual Connect Flex-10 technology that divides a dual-port network interface controller into 8 FlexNIC ports. This technology reduces the cost associated with data networking in a virtualization environment by greatly reducing the number of cables, switches and additional NICs needed.
Now we just announced a new Virtual Connect 8 Gb Fibre Channel module to support the heavy SAN needs of virtual servers. The HP Virtual Connect 8 Gb 24-port Fibre Channel Module has twice the bandwidth of our 4 Gb Fibre Channel module running at up to 8 Gb on all downlinks and uplinks. Second, it has a total of 8 uplink connections, which is double our current module. Third, it features support for increased server side NPIV support with 255 World Wide Names available per server. So all together more Virtual Machines can be hosted per server and per set of Virtual Connect Ethernet and Fibre Channel modules. The result is needing fewer servers AND fewer interconnect modules. Fewer servers and interconnect modules mean a lower purchase cost, simpler set-up and ongoing management, and fewer cables, all able to host more application workloads.. More for less works well for everyone.
We added a new Virtual Connect multi-enclosure stacking feature. Multi-enclosure stacking allows up to 4 BladeSystem enclosures to be connected together into one Virtual Connect Domain. This provides two big benefits. One, it creates a single simple server connection management domain for up to four enclosures, or up to 64 servers. Second, it also means fewer uplink cables to top of rack or core network ports, further reducing cable and expensive core port costs.
We’ve also enhanced Virtual Connect Enterprise Manager. The new 1.30 release supports our new Virtual Connect 8Gb Fibre Channel Module, our latest G6 server blades announced last month, and extends the number of supported Virtual Connect domains to 200. When combined with multi-enclosure stacking, this means that Virtual Connect Enterprise Manager can simplify the set-up and ongoing management of server I/O for up to 800 BladeSystem enclosures or put another way, up to 12,800 servers! Enabling system administrators the ability to manage the connectivity of up to 12,800 servers will go a long way to making life simpler and less expensive for many of our customers.
So for customers looking to converge infrastructure or increase benefits from virtualization, we hope you let HP and our resellers help you save money, reduce your network complexity, and simplify your IT environment with Virtual Connect Ethernet and Virtual Connect Fibre Channel.
“Connect More - Spend Less!”
ESS Virtual Connect
Every time a competitor introduces a new product, we can't help but notice they suddenly get very interested in what HP is blogging during the weeks prior to their announcement. Then when the competitor announces, the story is very self-congratulatory "we've figured out what the problem is with existing server and blade architectures". The implication being that blades volume adoption is somehow being constrained by the very thing they have and everyone else is really stupid.
HP BladeSystem growth has hardly been constrained; with quarterly growth rates of 60% or 80% and over a million BladeSystem servers sold. So I have to wonder if maybe we already have figured out what many customers want - save time, power, and money in an integrated infrastructure that is easy to use, simple to implement changes, and can run nearly any workload.
Someone asked me today "will your strategy change?" I guess given the success we've had, we'll keep focusing on the big problems of customers - time, cost, change and energy. It sounds boring, it doesn't get a lot of buzz and twitter traffic, but it's why customers are moving to blade architectures.
Our platform was built and proven in a step-by-step approach: BladeSystem c-Class, Thermal Logic, Virtual Connect, Insight Dynamics, etc. Rather than proclaim at each step that we've solved all the industry's problems or have sparked a social movement in computing; we'll continue to focus on doing our job to provide solutions that simply work for customers and tackle their biggest business and data center issues.
data center 3.0
Dynamic Power Capping
x86 server market