Introduce myself
My role at HP
Thanks for giving me 15 minutes, I’m going to share some thoughts on Innovation in our industry specifically pace and rate of change
Review what innovation is “doing to us” and why we now need innovation just to accommodate or absorb innovation.
Necessity really the mother of invention with father being survival. In our world and our economy ultimately is comes down to $$ and we just can’t spend like we used to.
Many people know HP as an iconic American company started by Dave Packard and Bill Hewlett.
HP’s reputation is the result of 75 years of innovation and some 36,000 patents that make up HP’s legacy of IP.
This list shows many things: a great legacy, a strong track record, and hundreds of customer problems solved and needs fulfilled.
I’m glad to tell you nothing has changed, HP is still a source if not the source of discontinuous innovation in Information Technology and I’ll be highlighting some of what’s here now and an example of what’s to come.
One thing to notice – look at the timeline and the pace of change.
”Riskcycle”—Key to Lifecycle Management
This is how it started for me
10th grade 1975/76
I remember punch cards too – mostly because my dad would bring them home
Storage was a spool of yellow paper tape
Machine language
Had to wait your turn
GE Timeshare not= condos on Maui
Lee, want to show other computer devices, not just HP…In 1990 when I had a real job I had one of these compaq portables
“sewing machine”
Was very cool – my friends came to see it (remember my Dad)
20 years later I had one of these. A lot changed in 20 years. Or at least it felt like it.
Remember these?
8” floppy in early 70s, Wang drove 5.35” adoption in mid 70w. Then came the 3.5”, culminating in dual sided “HD” Wahoo. Then they disappeared.
This was 20 years of innovation in floppy disk drives.
My kids don’t even know what a floppy drive is.
70s, 80s and 90s – in this business we would have said that was an era of dramatic and unparalleled innovation – and we were right at the time.
1990 wasn’t so long ago - but really it’s been an eon in our business
Today we don’t have our Fathers innovation pace, that’s for sure. Let’s take a look …
In the 15 short minutes I’m speaking with you …
This is our reality
Social networking
Machine to machine, device to device, devices everywhere – the internet of things
Big Big Big Data (really big)
Cloud Computing
This is of course an uncontrolled runaway explosion, driving a very different reality which if left unchecked will consume us and every dollar we have.
And it’s only going to get worse…
Of the entire world population, only about 3 Billion are “in the game” today
I think these predictions – near term – are understated.
Especially the prediction about devices
Internet connected device proliferation – this is just one home and not all the devices …
Devices, things are going to drive this massive consumption of compute, consumption of power, and creation of data
A single transatlantic flight can create 2 to 4 TB of telemetry – mostly presented and discarded. But that data is too important, too valuable to discard
And all this drives the need for more and more compute, more and more data centers, more and more storage, and the need for more and more dollars
Just in the next year – one year
10 million new servers needed just for commerce, search and social networking
That translates to somewhere between $10 and $20 Billion in new data center capacity needed
That’s a lot of power and a lot of construction
Enough power for millions of homes.
Maybe 10 power plants – and that’s just this year
You are not spared in Government. You have data center sprawl too. And cost sprawl.
Don’t forget about the data. Where are you going to put it? A lot of you are moving to cloud-like storage services.
Something to think about though
Your cloud storage apps may be scooping up what seems like grains of sand – but this can grow to mountains
The convergence of cloud and big data brings opportunities and pitfalls.
If you are buying a cloud storage service and you are ingesting mass amounts of data – and keeping it.
Terabytes turn to Petabytes … and 3 years later you have 125 PB of data … and you need to move it …
Lee,
This can’t be a HP commercial…need to make point that technology companies are developing lower cost computing to meet customers needs…….perhaps show mainframes vs todays technology with lower cost and footprint….
HP Labs has been paying attention
HPs project moonshot conceived a few years ago specifically to address the crisis in power consumption and density
An entire new form factor for servers
Using low power processors – cell phones – ARM and ATOM chips
Result = ultra high density, ultra high power.
What you see is a self contained server
Lee…ideas in telling the story in the next few slides without it being an HP commercial….this is a presentation on technology innovation, not hp………………This is the server chassis, or drawer
In this platform you can fit 45 cartridges in a 4.3u drawer
Everything unique to the server is on the cartridge, and everything that can be shared is in the drawer
- power, cooling networking, control
Several cartridge configurations available now, including 4 nodes on a single cartridge
That’s 160 servers in a single drawer. They consume 1/10 the power as an analogous traditional server, 20 times the density.
For hyperscale applications – migrating to this platform will save you $195 Million for every 100,000 servers.
$195 Million every year.
The first HP ProLiant Moonshot Server is available today with the Intel® Atom Processor S1260.
Trend is good, focus on power, workloads, maybe explain software servers and the value of that, not HP specific…
Aside from the power savings and innovation in density, the real leverage of the platform is in the system on a chip architecture
HP and other partners are building out an ecosystem of cartridges to deliver the first true software defined server.
Recap: SW defined datacenter and SW defined server
The server is configured to optimize particular workloads. This not only makes for more effieient data centers and lower datacenter costs but for more efficient code and SW development and maintenance costs as well.
Delete Proliant reference, can we tell the cable story without that?????Some guys in the test lab took these pictures for me.
This is 8 drawers, can get 10 in a full height rack.
That’s not a lot of cables and accompanying complexity for 1440 servers.
There is more innovation coming – let me tell you about Memristors. It takes time to get to market
Theorized in 1971 and only "discovered" in 2008 at HP Labs, Memristor exhibits characteristics of a both a resistor and a memory cell.
Unlike DRAM, memrister is nonvolatile
Unlike Flash, it doesn’t use charge to retain its state
Unlike magnetic storage devices, which are accessed mechanically, memrister is solid-state
It consumes zero power when not actively being written or read.
All of which means it could have a profound impact on computing and storage.
It feels as if we have seen as much in the last two years a we saw in the prior 20. I don’t think that’s changing any time soon.
More innovation is now required. Not just from HP.
Thank you.