get more from big data and your budget

12
GET MORE FROM BIG DATA AND YOUR BUDGET. with Hadoop and IBM POWER8 ® . [ 1 ]

Upload: ibm-power-systems

Post on 30-Jul-2015

643 views

Category:

Technology


0 download

TRANSCRIPT

GET MORE FROM BIG DATA AND YOUR BUDGET.

with Hadoop and IBM POWER8®.

[ 1 ]

McKinsey estimates that if

the U.S. health-care system

used big data analytics

creatively and effectively,

they could potentially find

$300 billion in value annually.

[ 2 ]

A healthierBottom Line.

60%60%A retailer has the

potential to increase

its operating margin by

60% by leveraging big

data to the fullest.

http://public.dhe.ibm.com/common/ssi/ecm/po/en/pol03216usen/POL03216USEN.PDF

[ 3 ]

Big data can help you make the sale.

$149Billion

New Technology helpsold world economies.

[ 4 ]

Big data analytics can help the developed economies of

Europe save $149 billion in operational efficiency.

http://public.dhe.ibm.com/common/ssi/ecm/po/en/pol03216usen/POL03216USEN.PDF

What kind of data are you managing?There are two types of data:Structured Data is what exists in your enterprise IT systems. They include CRM, inventory, billing and more.

Unstructured Data is everything else. 85% of it is estimated to come from sources like audio, documents, emails, images, RFID data, social media video and more.

[ 5 ]

To succeed, you need to be able to take the deluge of big data and turn it into actionable insights.

The convergence of mobile, social and cloud gives you the open, f lexible and agile environment you need.

THE

new age of computing.

[ 6 ]

DAT

DATA

[ 7 ]

DAT

DATA

[ 7 ]

iT Infrastructure is crucial.Traditional approaches to offline analysis of

siloed data can limit you. They can’t keep up

with the volume, variety and velocity of data

coming in due to next-gen network rollouts,

smart-phones, social media and the deluge of

unstructured data.

Emerging open source technologies like Hadoop

can help cut the processing time—especially in

distributed environments.

Unleash the value of data and deliver insights where they’re needed:

• High Performance processor designed for big data

• 4X more threads per core to process more concurrent queries vs x86

• 4X memory bandwidth and 5X cache for effi cient processing of big data vs 86

• 2.4X I/O bandwidth to ingest and move data faster than vs POWER7

• Excellent reliability, availability and serviceablity

• Unique innovation such as coherent accelerator processor interface

Drive faster insight with POWER8® for big data analytics.

[ 8 ]

The IBM Data Engine for Analytics is a customizable solution that’s built to meet your business needs:

• It addresses your challenges by connecting POWER8®-based compute servers to the IBM Elastic Storage Server, which is based on the IBM Spectrum Scale with a high-performance interconnect

• The Elastic Storage Server used by the IBM Data Engine for Analytics offers shared storage at lower cost than SAN and includes RAS and management functionality

[ 9 ]

Address the challenges inherent in Hadoop solutions.

Data volumes are growing at rates as high as

50% per year. Be ready:

• Get an expertly designed, tightly integrated and performance-optimized reference architecture for big data workloads

• Tailor it to meet big data workloads with a simple building block approach to match the appropriate mix of CPU and storage to application requirements—unlike traditional x86 solutions

• Optimize at a price competitive to x86 solutions

• Deploy clusters in a day or two thanks to automated software deployments

• Deliver with minimal staff training

Derive insight from all data with IBM Data Engine for Analytics — Power SystemsTM Edition.

[ 10 ]

The IBM Data Engine for Analytics offers many added benefits, including

significant advantages from POWER8® plus:

• Faster MapReduce engine

• Minimization of infrastructure duplication and reduced cost

• Sharing of infrastructure with non-Hadoop workloads

• Security isolation between tenants

• Allow tenants to expand to use available infrastructure

• Scale compute and storage independently with flexible building blocks as

your needs grow

[ 11 ]

Advantages over x86 standard Hadoop configurations.