Amazon EC2 provides a broad selection of instance types to accommodate a diverse mix of workloads. In this session, we provide an overview of the Amazon EC2 instance platform, key platform features, and the concept of instance generations. We dive into the current generation design choices of the different instance families, including the General Purpose, Compute Optimized, Storage Optimized, Memory Optimized, and GPU instance families. We will also provide an overview of the newest instances announced at re:Invent, including the latest generation of Memory and Compute Optimized Instances R4 and C5 instances, new Storage Optimized High I/O I3 instances, and new larger T2 instances. We also detail best practices and share performance tips for getting the most out of your Amazon EC2 instances. Learning Objectives: • Get an overview of the EC2 instance platform, key platform features, and the concept of instance generations • Learn about the latest generation of Amazon EC2 Instances • Learn best practices around instance selection to optimize performance

1. © 2015, Amazon Web Services, Inc. or its Affiliates. All rights reserved.© 2016, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Adam Boeglin, HPC Solutions Architect
Thursday, January 19, 2017
Choosing the Right EC2 Instance
and Applicable Use Cases

2.  Understanding the factors that going into choosing an EC2 instance
 Defining system performance and how it is characterized for
different workloads
 A look into our current generation instances and their features
 How Amazon EC2 instances deliver performance while providing
flexibility and agility
 How to make the most of your EC2 instance experience through the
lens of several instance types
What to Expect from the Session

3. Host Server
Hypervisor
Guest 1 Guest 2 Guest n
Amazon EC2 Instances

4. In the past
 First launched in August 2006
 M1 Instance
 “One size fits all”
M1

5. Amazon EC2 Instances History
2006 2008 2010 2012 2014
2016
m1.small
m1.large
m1.xlarge
c1.medium
c1.xlarge
m2.xlarge
m2.4xlarge
m2.2xlarge
cc1.4xlarge
t1.micro
cg1.4xlarge
cc2.8xlarge
m1.medium
hi1.4xlarge
m3.xlarge
m3.2xlarge
hs1.8xlarge
cr1.8xlarge
c3.large
c3.xlarge
c3.2xlarge
c3.4xlarge
c3.8xlarge
g2.2xlarge
i2.xlarge
i2.2xlarge
i2.4xlarge
i2.4xlarge
m3.medium
m3.large
r3.large
r3.xlarge
r3.2xlarge
r3.4xlarge
r3.8xlarge
t2.micro
t2.small
t2.med
c4.large
c4.xlarge
c4.2xlarge
c4.4xlarge
c4.8xlarge
d2.xlarge
d2.2xlarge
d2.4xlarge
d2.8xlarge
g2.8xlarge
t2.large
m4.large
m4.xlarge
m4.2xlarge
m4.4xlarge
m4.10xlarge
x1.32xlarge
t2.nano
m4.16xlarge
p2.xlarge
p2.8xlarge
p2.16xlarge
x1.16xlarge
r4.large
r4.xlarge
r4.2xlarge
r4.4xlarge
r4.8xlarge
r4.16xlarge
t2.xlarge
t2.2xlarge
2015

6. Instance generation
c4.large
Instance family Instance size

7. Choices and Flexibility
 Choice of Processor
 Memory
 Storage Options
 Accelerated Graphics
 Burstable Performance

8.  Servers are hired to do jobs
 Performance is measured differently depending on the job
Hiring a Server
?

9. Performance Factors
Resource Performance factors Key indicators
CPU Sockets, number of cores, clock
frequency, bursting capability
CPU utilization, run queue length
Memory Memory capacity Free memory, anonymous paging,
thread swapping
Network
interface
Max bandwidth, packet rate Receive throughput, transmit throughput
over max bandwidth
Disks Input / output operations per
second, throughput
Wait queue length, device utilization,
device errors
Acceleration FPGA or GPU offloading from CPU Parallelism and Code Design

10. Broad Set of Compute Instance Types
M4
General
purpose
Compute
optimized
C4
C3
Storage and I/O
optimized
I3
G2
GPU or FPGA
enabled
Memory
optimized
X1
P2
F1
R4
R3
C5
I2
D2

11. Resource Utilization
 For given performance, how efficiently are
resources being used
 Something at 100% utilization can’t
accept any more work
 Low utilization can indicate more resource
is being purchased than needed

12. Example: Web Application
 MediaWiki installed on Apache with 140 pages of content
 Load increased in intervals over time

13. Example: Web Application
 Memory stats

14. Example: Web Application
 Disk stats

15. Example: Web Application
 Network stats

16. Example: Web Application
 CPU stats

17.  Give back instances as easily as you can acquire new ones
 Find an ideal instance type and workload combination
 EC2 Instance Pages provide “Use Case” Guidance
 With EBS, storage and instance size don’t need to be coupled
Instance Selection = Performance Tuning

18. “Launching new instances and running tests
in parallel is easy…[when choosing an
instance] there is no substitute for measuring
the performance of your full application.”
- EC2 Documentation

19. How not to choose an EC2 instance
 Brute Force Testing
 Ignoring Metrics
 Favoring old generation instances
 Guessing based on what you already have

20. Instance sizing
c4.8xlarge 2 - c4.4xlarge
≈
4 - c4.2xlarge
≈
8 - c4.xlarge
≈

21. Choosing the right size
Understand your unit of work
 Web request
 Database / Table
 Batch Process
What is that unit’s requirements?
 CPU threads
 Memory Constraints
 Disk & Network
What are it’s availability requirements?

22. Utilization & Auto Scaling Granularity

23. Utilization & Auto Scaling Granularity

24. Utilization & Auto Scaling Granularity

25. General Purpose Instances

26. Review: M4 Instances
 General Purpose Instance Family
 Balance of Compute, Memory, and Network Resources
 2.3 GHz Intel Xeon® E5-2686 v4 (Broadwell) processors or
2.4 GHz Intel Xeon® E5-2676 v3 (Haswell) processors
Model vCPU Memory
(GiB)
Storage EBS
Bandwidth
(Mbps)
m4.large 2 8 EBS Only 450
m4.xlarge 4 16 EBS Only 750
m4.2xlarge 8 32 EBS Only 1,000
m4.4xlarge 16 64 EBS Only 2,000
m4.10xlarge 40 160 EBS Only 4,000
m4.16xlarge 64 256 EBS Only 10,000
Databases, Data Processing, Caching, SAP, SharePoint, and other enterprise applications.

27. Review: T2 Instances
 Lowest cost EC2 instance at $0.0065 per hour
 Burstable performance
 Fixed allocation enforced with CPU credits
Model vCPU Baseline CPU Credits
/ Hour
Memory
(GiB)
Storage
t2.nano 1 5% 3 .5 EBS Only
t2.micro 1 10% 6 1 EBS Only
t2.small 1 20% 12 2 EBS Only
t2.medium 2 40%** 24 4 EBS Only
t2.large 2 60%** 36 8 EBS Only
t2.xlarge 4 90%** 54 16 EBS Only
t2.2xlarge 8 135%** 81 32 EBS Only
General Purpose, Web Serving, Developer Environments, Small Databases

28. How Credits Work
 A CPU credit provides the performance of a
full CPU core for one minute
 An instance earns CPU credits at a steady rate
 An instance consumes credits when active
 Credits expire (leak) after 24 hours
Baseline rate
Credit
balance
Burst
rate

29. Tip: Monitor CPU credit balance

30. Compute Optimized Instances

31. Review: C4 Instances – “Compute”
 Custom Intel E5-2666 v3 at 2.9 GHz Turbo to 3.5 Ghz
 P-state and C-state controls
Model vCPU Memory (GiB) EBS (Mbps)
c4.large 2 3.75 500
c4.xlarge 4 7.5 750
c4.2xlarge 8 15 1,000
c4.4xlarge 16 30 2,000
c4.8xlarge 36 60 4,000
Batch & HPC workloads, Game Servers, Ad Serving, & High Traffic Web Servers

32. C5 Instance Preview
• Next Generation “Skylake” Intel® Xeon® Processor
family
• AVX 512 Instruction Set
• Up to 72 vCPUs in a single instance
• 144 GB of RAM
• Coming early 2017

33. Memory Optimized Instances

34. Review: R4 Instances – “RAM”
 Intel Xeon E5-2686 v4 (Broadwell) Processors @ 2.3 Ghz
 DDR4 Memory
Model vCPU Memory (GiB) Network
Performance
r4.large 2 15.25 Up to 10 Gigabit
r4.xlarge 4 30.5 Up to 10 Gigabit
r4.2xlarge 8 61 Up to 10 Gigabit
r4.4xlarge 16 122 Up to 10 Gigabit
r4.8xlarge 32 244 10 Gigabit
r4.16xlarge 64 488 20 Gigabit
In-memory databases, big data processing, HPC workloads

35. Review: X1 Instances
 Largest memory instance with 2 TB of DRAM
 Quad socket, Intel E7 processors with 128 vCPUs
Model vCPU Memory (GiB) Local Storage Network
x1.16xlarge 64 976 1x 1920GB SSD 10Gbps
x1.32xlarge 128 1952 2x 1920GB SSD 20Gbps
In-memory databases, big data processing, HPC workloads

36. Storage Optimized Instances

37. Review: D2 Instances – “Dense Storage”
 Intel Xeon E5-2676 v3 (Haswell) @ 2.4 GHz
 3.5 GBps read and 3.1 GBps write throughput w/ 2 MiB block size
Model vCPU Memory
(GiB)
Storage Read Throughput
(2 MiB Block size)
d2.xlarge 4 30.5 3 x 2 TB 438 MB/s
d2.2xlarge 8 61 6 x 2 TB 875 MB/s
d2.4xlarge 16 122 12 x 2 TB 1,750 MB/s
d2.8xlarge 36 244 24 x 2 TB 3,500 MB/s
MapReduce & Hadoop, Distributed File systems, Log and Data Processing

38. Review: I2 Instances – “IOPS”
 16 vCPU: 3.2 TB SSD; 32 vCPU: 6.4 TB SSD
 365K random read IOPS for 32 vCPU instance
Model vCPU Memory
(GiB)
Storage Read IOPS Write IOPS
i2.xlarge 4 30.5 1 x 800 SSD 35,000 35,000
i2.2xlarge 8 61 2 x 800 SSD 75,000 75,000
i2.4xlarge 16 122 4 x 800 SSD 175,000 155,000
i2.8xlarge 32 244 8 x 800 SSD 365,000 315,000
NoSQL Databases, Clustered Databases, Online Transaction Processing (OLTP)

39. Coming Soon: I3 Instance
• Up to 64vCPUs
• Intel® Xeon® E5-2686 v4 “Broadwell” @ 2.3 GHz
• 488 GB of Memory
• 15.2 TB of NVMe Based SSD’s
• 3.3 Million Random IOPS @ 4KB
• 16GB/s sequential throughput
• Early 2017

40. EBS Performance
 Instance Size Matters
 Match your volume size and
type to your instance
 Use EBS Optimization if EBS
performance is important

41. Accelerated Instances

42. Review: G2 Instances – “GPU”
• Up to 4 NVIDIA GRID K520 GPUs in a single instance
• Each with 1,536 CUDA cores and 4GB of video memory
• High-performance platform for graphics applications using
DirectX or OpenGL
G2
Instance
Size
GPUs vCPUs Memory
(GiB)
SSD Storage
g2.2xlarge 1 8 15 1 x 60GB
g2.8xlarge 4 32 60 2 x 120GB
Video creation services, 3D visualizations, streaming graphics, server-side graphics workloads

43. Review: P2 GPU Instances – “Parallel”
• Up to 16 K80 GPUs in a single instance
• Supports CUDA 7.5 and above, OpenCL 1.2, and the
GPU Compute APIs
• Including peer-to-peer PCIe GPU interconnect
P2
Model GPUs GPU Peer
to Peer
vCPUs Memory
(GiB)
GPU
Cores
GPU
Memory
Network
Bandwidth*
p2.xlarge 1 - 4 61 2,496 12 Gib High
p2.8xlarge 8 Y 32 488 19,968 96 Gib 10Gbps
p2.16xlarge 16 Y 64 732 39,936 192 Gib 20Gbps
*In a placement group
Deep learning, HPC simulations, and Batch Rendering

44. Review: F1 Instances – “FPGA”
• Up to 8 Xilinx Virtex UltraScale Plus VU9p FPGAs in a single instance
with four high-speed DDR-4 per FPGA
• Largest size includes high performance FPGA interconnects via PCIe
Gen3 (FPGA Direct), and bidirectional ring (FPGA Link)
• Designed for hardware-accelerated applications including financial
computing, genomics, accelerated search, and image processing
F1
Instance Size FPGAs FPGA
Link
FPGA
Direct
vCPUs Memory
(GiB)
NVMe
Instance
Storage
Network
Bandwidth*
f1.2xlarge 1 - 8 122 1 x 480 5 Gbps
f1.16xlarge 8 Y Y 64 976 4 x 960 30 Gbps
*In a placement group

46. Next steps
 Visit the Amazon EC2 documentation
 Launch an instance and try your app!

47. Thank you!