Can big IT companies compete in cloud?

The big IT vendors are IBM, HP, Dell, Oracle, EMC, Netapp, Cisco. These companies are now struggling to maintain revenue growth as customers are moving to cloud.

Can the traditional IT companies also move their business model to cloud and be competitive with the big three leading cloud vendors Amazon, Google and Microsoft?

Here are the significant challenges that face the traditional IT vendors:

1. Cloud is not new and the big three have almost two decades of experience running massive cloud environments:

• Amazon:  is the world’s largest online retailer and started in 1995, they run massive infrastructure in many geographies to support their online business and since 2002 they have offered their cloud to the public as AWS
• Google:  is the world’s largest cloud company and started in 1998. It’s cloud power the world’s most popular search engine, mail service (Gmail), mobile platform (Android), video platform (YouTube), blogging platform (Blogger), navigation platform (Google Maps) and many other cloud services. Google in the undisputed leader in cloud, yet it only started offering their cloud to the public since 2012 as Google Compute Engine.
• Microsoft:  have significant experience with running large cloud environments in many geographies and started with leading mail service hotmail.com 1996. Since then they have uses cloud to power search engine Bing.com, Office 365, Skype, Xbox live gaming, Windows mobile platform, Skydrive. They started offering their cloud to the public since 2010 as Microsoft Azure.
• As part of running these massive clouds the big three have almost two decades of experience running huge environments and have tens of thousands of cloud specialists with extensive experience.  The big IT vendors including IBM, HP, Dell, Oracle, EMC, Netapp, Cisco have no direct experience running massive cloud of this scale and it will take years to build up this level of experience.

2. The big three have massive scale and leverage their online clouds to drive much higher levels of efficiency and lower cost:

• http://www.datacenterknowledge.com/archives/2013/07/15/ballmer-microsoft-has-1-million-servers/

At Microsoft’s 2013 Worldwide Partner Conference, CEO Steve Ballmer gave a very interesting tidbit about the scale of Microsoft’s server operations. “We have something over a million servers in our data center infrastructure. Google is bigger than we are. Amazon is a little bit smaller. You get Yahoo! and Facebook, and then everybody else is 100,000 units probably or less. So the number of companies that really understand the network topology, the data center construction, the server requirements to build this public cloud infrastructure is very, very small.”

• http://www.theregister.co.uk/2013/06/05/ibm_softlayer_cloud/

The big three have huge investment over many years in data centers, servers, storage, networking in many geographies to create their clouds.

  Unless the boards of traditional IT companies agree to have huge capital investments in the order of tens of billions of dollars they can not compete on the same level of scale and cost.

3. The big three design cloud from the ground up using commodity low cost infrastructure and open source in energy efficient data centres.

a) Data Centres

Cloud companies design massive data centres of unprecedented scale that are extremely power efficient . Data-center experts gauge efficiency using a statistic called power usage effectiveness (PUE), which is computed by dividing all the electrical power used in a facility by the power delivered to just the computers and related networking equipment. Google's data centers achieved an overall PUE of 1.13, Facesbook's Open Compute project is aiming for PUE of 1.07.

The data centres are very different to enterprise data centres, at times designed around modular giant shipping containers allowing the cloud provider to pack up and move all these servers to another location at a moment's notice.

http://spectrum.ieee.org/telecom/internet/under-the-hood-at-google-and-facebook

b) Low cost commodity hardware

Cloud companies like  Google, Amazon, Microsoft and Facebook  buy their hardware directly from low-cost asian original design manufacturers like Huawei, Quanta, Compal Electronics and Wistron. This cuts out traditional IT vendors from these server sales. Facebook's Open Compute project is open sourced hardware, it lays out the blueprint how to blueprint for low cost infrastructure and energy efficient design

• Modern cloud providers have completely redesigned IT from the ground up including:
• Servers based on low cost commodity servers
• Storage based on low cost commodity storage
• Networking based on low cost commodity networking
• Software based on free and open-source software

• http://www.wired.com/wiredenterprise/2012/11/amazon-google-secret-servers/

Amazon followed Google and started using low cost commodity servers. Google goes straight to Asia. So does Facebook. And according to a former Quanta employee who spoke to Wired earlier this year, even Microsoft is purchasing data-center hardware straight from Asian companies that the average American has never heard of.

In operating at such an enormous scale and cutting hardware costs in this way, Amazon can improve its bottom line, but it can also drive down the cost of Amazon Web Services.

• http://www.wired.com/wiredenterprise/2014/01/microsoft-open-compute-servers/

http://www.globalfoundationservices.com/posts/2014/january/27/microsoft-contributes-cloud-server-specification-to-open-compute-project.aspx

For nearly two years, tech insiders whispered that Microsoft was designing its own computer servers. It only made sense. Typically, when you run a web service the size of Bing, needing tens of thousands of machines to keep the thing going, traditional server hardware becomes far too expensive.

Microsoft “open sourced” its server designs, sharing them with the world at large so that other online outfits can use them inside their own data centers. “We’re trying to drive hardware innovation in cloud computing,” It’s yet another sign that the worldwide market for data center hardware is changing in enormous ways. In the past, if you needed servers or data storage gear or networking hardware, you simply bought what was available from American hardware vendors like Dell and HP and Cisco. Now, massive web outfits like Google and Facebook and Amazon and even Microsoft are designing their own hardware, partnering with manufacturers in Asia and other foreign locales to build this hardware on the cheap.

• http://www.wired.com/wiredenterprise/2012/03/google-microsoft-network-gear/

Google, Amazon, Microsoft, and Facebook buy more networking hardware than practically anyone else on earth. After all, these are the giants of the internet, yet they are quietly moving away from U.S.-based sellers like Cisco and Juniper to buy cheaper gear in bulk straight from “original design manufacturers,” or ODMs from China and Taiwan.

Howard Wu — the president of greater China for Joyent, an Amazon-like cloud computing outfit based in San Francisco — agrees. “If you’re a small business and you’re going to buy five servers, you’re going to Dell or HP, because of the support services. But if you’re a data center operator and you’re going to buy 10,000 servers, you’re going straight [to the ODMs],” he says. “It’s kind like buying couches. If you buy one, you go to a retail store. If you buy 10,000 couches, you go straight to the factory.”

Cloud provider they use low-cost devices from ODMs. Most of the  networking   equipment uses automated software defined networking where the networking equipment has APIs and is far easier to manage through software.

                                                                         

• http://mvdirona.com/jrh/TalksAndPapers/JamesHamilton_Reinvent20131115.pdf             

Cloud provider have custom storage design based on low cost commodity infrastructure

• http://www.cs.cornell.edu/projects/ladis2009/talks/dean-keynote-ladis2009.pdf             

Google's Jeff Dean presentation "Designs, Lessons and Advice from Building Large Distributed Systems"

• http://mvdirona.com/jrh/TalksAndPapers/JamesHamilton_Reinvent20131115.pdf

Amazon's  James Hamilton presentation " Why Scale Matters and how the Cloud Really is Different" (Best insight on Amazon AWS infrastructure!!)

• http://www.opencompute.com

The Open Compute Project initiative was announced in 2011 by Facebook to openly share designs of data center products. It includes designs for: data centre, racks, servers, storage, networking,  hardware management. These design are free and open sourced and are available for purchase from Asian ODMs.

• Open Source Hardware + Open Source Software + Asian ODMs = Disruption

    4.  Heavy use of automation - Software Define Datacentre

The big three cloud vendors have hundreds of thousands of servers to manage. Installing, configuring and operating with the traditional IT model would require a huge number of resources and hundreds of thousands of manual tasks which is expensive, slow and error prone. So they developed heavy use of automation where the majority of the tasks is done by software rather than by administrators.

The software defined data center encompasses all the concepts of software defined networking, software defined storage, cloud computing, automation, management and security. Every low-level infrastructure component in a data center can be provisioned, operated, and managed through an API. Not only are there tenant-facing APIs, but operator-facing APIs which help the operator automate tasks which were previously manual. The software defined datacentre is managed using APIs or a user friendly portal that encompasses all aspects that are done in the traditional IT environment.

The big three cloud providers already use software defined data center where every component can be accessed and manipulated through an API. The proliferation of APIs will change the way people work. Programmers who have never formatted a hard drive will now be able to provision terabytes of data. A web application developer will be able to set up complex load balancing rules without ever logging into a router. IT organizations will start automating the most mundane tasks. Applications can be created to mimic the organization’s process and workflow and automate infrastructure management.

Organisations that do not use highly integrated software define datacentre portal and APIs will not be competitive as they will significantly more staff and will be more more expensive and much slower to deploy.

5.   Features

Many organisation claim to have a competitive cloud offering yet they are missing many features that are true cloud innovation including the following:

• On demand, self service without requiring human interaction with service provider
• Pay as you use - no upfront cost, no long term contracts
• VMs from 1 vCPU to 32+ vCPUs, 1 GB to 200+ GB memory
• Unlimited storage on demand, options for fast flash storage with volumes over 3000+ IOPS
• Data centres in many geographies all around the world
• Multiple data centres in each geography to provide high availability and DR capability
• Dark fibre, low latency networking between data centres in each geography
• Database PaaS for major databases including options for high availability
• Security Compliance with major security standards: ISO/IEC 27001, SOC 1/SSAE 16/ISAE 3402, SOC 2, SOC 3, PCI DSS Level 1
• Easy to use web portal and software defined computing with APIs for all services:

• compute
• storage
• networking
• firewalls
• routers
• load balancing
• VPN
• DNS
• monitoring
• backup
• CDN
• database services
• message queuing service
• payment services
• identity and access management services
• automated provisioning services
• billing

    6.   Traditional IT vendors entering into low cost, low margin cloud business completely changes existing business model

• Pay as you use, no capital upfront - OPEX rather than CAPEX
• No long term contracts, much lesser vendor lock in
• Low margin business, where previously IT companies made 50%+ gross profit margin, they will be lucky to make 25% gross profit margin
• Cloud will canibalise existing high margin business, hence big IT vendors will be reluctant to lead sales with low cost, commodity cloud
• Profit margin is too thin to engage resellers and for resellers to still make good profit margin, unless resellers can compete directly with big IT vendors on services business which in turn would cannibalise IT vendors services business
• If customers choose cloud traditional IT vendors will face decline of software revenue as the need for operating system, virtualisation, backup, monitoring, system management, database and middleware will decline as competitive cloud offering are built on free and open software
• Cloud is highly automated and easy to manage through portal or API, hence significantly fewer resources are required to manage infrastructure compared to traditional enterprise IT which is heavily reliant on manual tasks by many resources which are time consuming and need to be coordinated. Cloud would lead to significantly less service revenue than traditional manual IT services
• IaaS/PaaS cloud platforms need far less management, hence traditional IT hardware, software and infrastructure services revenue will decline significantly
• If customers choose SaaS cloud platform traditional IT revenue will decline
• If traditional vendors move aggressively from high profit margin products and services to low cost, lower profit margin cloud their quarterly profit will decline. Will the boards and shareholders allow this transition or will they resist change?

                    Traditional IT Vendors Gross Profit Margin (2013):

	Revenue (Billion)	Gross Profit (Billion)	Gross Profit Margin (%)
IBM http://goo.gl/DVjJn	103.2	50.3	48.7%
HP http://goo.gl/y6OBj	118.68	27.97	23.6%
Cisco http://goo.gl/FhjYv	47.25	28.21	59.7%
Dell http://goo.gl/ARBIU	56.94	12.19	21.4%
EMC http://goo.gl/6Selz	22.01	13.64	62.0%
VMware http://goo.gl/EkBWd	4.74	3.88	81.9%
NetApp http://goo.gl/1n5Qa	6.32	3.71	58.7%
Oracle http://goo.gl/zZbYc	37.15	29.26	78.8%

    7. Cloud is agile, traditional IT is not

• No need to plan for data centre racks, power, cooling, capacity
• No need to purchase hardware infrastructure: servers, storage, networking
• No need to carefully plan infrastructure capacity
• No need to purchase software: virtualisation, operating system, database, middle ware, monitoring, backup, system management
• No need to carefully plan software capacity
• Enterprise apps store, just install best practice software configuration with one click.
• No cost upfront - Just pay for what you use
• Much faster delivery for typical enterprise applications as no need to:
• Purchase servers, storage, networking hardware
• Purchase expensive software licences
• Negotiate pricing
• Get approvals to spend expensive OPEX
• Wait for hardware to arrive
• Rack and stack hardware into datacentre
• Cable to LAN, SAN and configure networking
• Install operating systems
• Install database and middleware
• Install applications stack
• Test applications stack
• Once above base install has been done then application be customised
• No long term contracts, easy to experiment and trial new ideas without expensive lock-in
• With cloud the business can focus on key business objectives while the cloud platform does the undifferentiated heavy lifting of traditional IT.

8.   Enterprise apps store

Prior to apps stores we had to manually download software, check hardware and software compatibility lists, install any prerequisite software and then finally install the required software. Even a simple software installation was daunting and time consuming, yet with the introduction of the apps store application installation has now been greatly simplified. Apple introduced the apps store on their iPhone and the concept became a huge success and now we have apps stores everywhere. Apps stores are available on all iOS and Android mobile platforms even on Macs and Windows we now have the convenience of apps stores.

With cloud the apps store is now available to enterprises. Want a to install best practice CRM software? Simply use 1-click installer and it will do all the heavy lifting for you and install the virtual machines, storage, networking, best practice security and configure the applications stack so that different application components are ready. This type of task in a traditional enterprise IT environment would take months as it would require applications design, infrastructure design, capacity planned, financial approvals for expensive long term commitment for purchase of servers, storage and networking, applications software purchase, middleware and databases licences purchase, waiting for all the components to arrive, installation of hardware in data centre, installation of operating systems, installation of middleware. installation of databases, installation of application by various SMEs, testing of applications stack. Only then is the application ready to be customised to suite the customers business.

With common enterprise applications it is not unusual for this cycle to take anywhere from one month to over a year from initial concept to implementation. Furthermore this is expensive and all the hardware and software must be purchased up front. With cloud you pay for what you use and if the application does not suit the business you simply stop using it and pay more more without being locked into a solution that is a poor fit for the business. Hence cloud represent far less risk as the business can experiment and be far more agile as there are no long term, expensive lock ins.

Traditional IT without enterprise apps store in the cloud is not agile and deployments would take significantly longer. A project team spending many months deploying a new application carries significant cost including and a large part of overall costs is resource costs for project team, other costs include: data centre, hardware infrastructure, software licensing, applications licensing and external application consultants or contractors to get everything running.

With enterprise cloud apps store the basic applications framework with best practice implementation can be ready in hours at very low cost as you only pay for what you use, then if the application does not suit the business it can be abandoned easily without significant cost impact. This allows businesses to to experiment and trial applications quickly, business can focus on their core business rather than the expensive and convoluted world of traditional enterprise IT.

IT without enterprise apps store is not competitive.

Cloud is not new and the big three (Amazon, Google, Microsoft) have almost two decades of experience running massive cloud environments, traditional IT companies have little experience in this space and it take many years to build up thousand of resources and gain experience to compete, they will need to spend tens of billions to create the massive infrastructure required to be able to compete at competitive scale and price. Furthermore traditional IT companies would need to completely change their existing models and in the process will cannibalise their existing user base to move a much less profitable business model.

To summarise:

• Big three cloud vendors have almost two decades of experience running massive cloud environments.
• Big three cloud vendors have huge scale with over a million servers in their clouds based on low cost commodity servers/storage/networking and heavy use of free and open source software.
• Big three cloud providers are available in many geographies and have multiple data centres to provide high availability to applications.
• No upfront cost to get started, low cost to implement. No long term contracts – no lock in.
• Lower cost than most companies due to economies of scale to pass on to customers, regular price decreases every few months.
• Dramatic increase in speed and agility.
• Self service. Users are able to provision, monitor and manage computing resources as needed without the need for manual, error prone and expensive  administrators. Completely integrated with compute, storage, networking, security, backup, monitoring, database and many other services.
• Enterprise apps store which allows 1-click best practice installation of enterprise applications.
• Elastic infrastructure allow you to easily scale up or down, pay only for what you use. Eliminate the need for significant capacity planning
• Far less maintenance and employees. Datacentre/hardware/software is managed by cloud provider.
• Measured service – IT resource utilization is tracked for each application and tenant, typically for cloud billing
• For traditional IT companies to adapt low cost commodity cloud based their profit margin would require complete change in business model. Cloud will canibalise existing high margin business, hence big IT vendors will be reluctant to lead sales with low cost, commodity cloud.

Will traditional IT companies embrace change and adapt or will they resist and diminish?

Update 21/10/2014

https://gigaom.com/2014/10/20/microsoft-ceo-in-public-cloud-its-go-big-or-get-the-heck-out/

Building a global cloud is a huge opportunity but also a money pit. So if you’re a tech provider and haven’t already invested multiple billions in data centers and other infrastructure to support your cloud, you might as well use someone else’s. Microsoft CEO Satya Nadella didn’t say those words exactly, but it’s the gist of what he told CNBC in an interview on Monday.

“I think that if you’re not already spending a lot of capital in the order of four or five billion dollars each year to just grow your cloud, probably it’s a little too late to enter the market. I mean, that’s the entry barrier, and there are a few of us who are in that mega-scale of cloud,” Nadella noted in advance of a Microsoft cloud event to be held in San Francisco Monday.