Cloudera — Series A Pitch Deck (September 2008)
15 slides from Cloudera's pitch deck, September 2008
Slide 1: Cloudera: Hadoop for the Enterprise
Cloudera: _ Hadoop for the Enterprise = ss September 2008
Slide 2: Data Growing Much Faster than Moore's Law
Data Growing Much Fastertian Woore Ss Law —— : ‘Source: Richard Winter, - Why Are Data 2008 «©2010 2012 Warehouses Growing Figure 1: Exponential Data Warehouse Growth so Fast?, April 2008 (size in terabytes of user data) 04/21/17 Cloudera Confidential
Slide 3: Processor Performance
Dinprocessor Performance - a 10000 +— 53x] From Hennessy and Patterson, Computer Architecture: A ) Quantitative Approach, 4th edition, Sept. 15, 2006 8 So Performance (vs. VAX- 11/780) S =) processors per chip 1¢ 5 = T T T T T T T T T t t T 1 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 20
Slide 4: Founding Team
Es nding Team — — Created Google/NSF Hadoop cluster and A, MS CS, Berkeley eae aa Wwadallah CTO. VP — BACS, U Washington eee Watalan, e Jeff Hammerbacher, VP |] = Engineering => Product = Founder Aptivia/VivaSmart : — Ran world’s largest = 8 years at Yahoo! running ; fo) tional BI rt BI infrastructu
Slide 5: Parallelize a data analysis task
== Parallelize a data analysis task * Interfaces built on top of MapReduce * Storage layer beneath (HDFS) e Doug Cutting, Mike Cafarella are advisors 04/21/17 Cloudera Confidential
Slide 6: Open source licensing benefits
mm laliae= REGUCE —— Third-party inspection of source code provides = assurances on security, product quality * Business-friendly license encourages commercial development — “Open core” licensing — Closed-source components, applications 04/21/17 Cloudera Confidential 6
Slide 7: Hadoop Users
meagdoop Users CARRIER [[@) mim. Autodesk G O ( gle sn Powerset @ | invent “New York Times AOL & quantcast 7
Slide 8: Competitors and Google Trends
Womenwim=s Google Irends @teradata @ netezza Netezza: $127M in FY08, $79M in FY07 Teradata: $830M in 1H08, $1.7B in FY07 04/21/17 Cloudera Confidential
Slide 9: Google Insights search volume for Hadoop
Source: -Google Insights world map for searches on “hadoop”, Search volume index Sept 2008. 0 CE 0 04/21/17 Cloudera Confidential
Slide 10: Operates on unstructured and structured data
= Bperates on unstructured data, and structured data (HBASE, HIVE) e Prescriptive development, grows with you without needing to re-architect e Procedural language offers power 04/21/17 Cloudera Confidential uy
Slide 11: Data Mining and Live Web Site architecture
Data Mining Live Web Site R, Weka, MemCached SAS, SPSS lion MySQL ETL |] ETL |} ETL ry Expensivq ETL Grids File Server Farm for Warehouse ()0)-Gucryabic) Warehouse Pre-Processing Log Collection Instrumentation 04/21/17 Cloudera Confidential 11
Slide 12: Smart Storage
—_ mart Storag nil Data Mining Live Web Site R, Weka, MemCached SAS, SPSS MySQL A able Constimption mia io Smart Storage: Grid For File Storage & Data Processing Warehouse Pre-Processing Log Collection Instrumentation 04/21/17 Cloudera Confidential 12
Slide 13: Data Volume and complexity
at Volume _ 10078 Qu Batch Data Processing Schema Read/Write Appen Only /) Stryctured 100TB > Unsirucwures Complexity Per Job Data Volume OLAP/OLTP y interactive Responsiveness + 100 Tables Table Join Complexity 04/21/17 Cloudera Confidential 13
Slide 14: The Cloud Wars: $100+ billion at stake
The Cloud Wars: $100+ billion at stake Wi The Cloud - A multi-year shift in the computing paradigm We are in the midst of a pronounced shift from client-server to Cloud computing, which is more analogous to centralized mainframe computing. Quantum improvements in Internet bandwidth, computing power
Slide 15: Enterprise Hadoop platform
Ena, Hado) e}| platfonm with Wcal multip exi xing over many customers venant Support: Concurrency, Priority, Namespace eet Performance Isolation. Beilience and Fast Recovery: Anon-sexy problem Siiiatis critical to enterprises, no time to restart ETL job = from scratch, otherwise misses SLA. “== IDE