Coming Soon – Amazon EC2 G4ad Instances Featuring AMD GPUs for Graphics Workloads

Customers with high performance graphic workloads, such as game streaming, animation, and video rendering for example, are always looking for higher performance with less cost. Today, I’m happy to announce new Amazon Elastic Compute Cloud (EC2) instances in the G4 instance family are in the works and will be available soon, to improve performance and reduce cost for graphics-intensive workloads. The new G4ad instances feature AMD’s latest Radeon Pro V520 GPUs and 2nd generation EPYC processors, and are the first in EC2 to feature AMD GPUs.

G4dn instances, released in 2019 and featuring NVIDIA T4 GPUs, were previously the most cost-effective GPU-based instances in EC2. G4dn instances are ideal for deploying machine learning models in production and also graphics-intensive applications. However, when compared to G4dn the new G4ad instances enable up to 45% better price performance for graphics-intensive workloads, including the aforementioned game streaming, remote graphics workstations, and rendering scenarios. Compared to an equally-sized G4dn instance, G4ad instances offer up to 40% improvement in performance.

G4dn instances will continue to be the best option for small-scale machine learning (ML) training and GPU-based ML inference due to included hardware optimizations like Tensor Cores. Additionally, G4dn instances are still best suited for graphics applications that need access to NVIDIA libraries such as CUDA, CuDNN, and NVENC. However, when there is no dependency on NVIDIA’s libraries, we recommend customers try the G4ad instances to benefit from the improved price and performance.

AMD Radeon Pro V520 GPUs in G4ad instances support DirectX 11/12, Vulkan 1.1, and OpenGL 4.5 APIs. For operating systems, customers can choose from Windows Server 2016/2019, Amazon Linux 2, Ubuntu 18.04.3, and CentOS 7.7. Instances using G4ad can be purchased as On-Demand, Savings Plan, Reserved Instances, or Spot Instances. Three instance sizes are available, from G4ad.4xlarge, with 1 GPU, to G4ad.16xlarge with 4 GPUs, as shown below.

Instance Size	GPUs	GPU Memory (GB)	vCPUs	Memory (GB)	Storage	EBS Bandwidth (Gbps)	Network Bandwidth (Gbps)
g4ad.4xlarge	1	8	16	64	600	Up to 3	Up to 10
g4ad.8xlarge	2	16	32	128	1200	3	15
g4ad.16xlarge	4	32	64	256	2400	6	25

The new G4ad instances will be available soon in US East (N. Virginia), US West (Oregon), and Europe (Ireland).

Learn more about G4ad instances.

— Steve Via AWS News Blog https://ift.tt/1EusYcK

New EC2 M5zn Instances – Fastest Intel Xeon Scalable CPU in the Cloud

We launched the compute-intensive z1d instances in mid-2018 for customers who asked us for extremely high per-core performance and a high memory-to-core ratio to power their front-end Electronic Design Automation (EDA), actuarial, and CPU-bound relational database workloads.

In order to address a complementary set of use cases, customers have asked us for an EC2 instance that will give them high per-core performance like z1d, with no local NVMe storage, higher networking throughput, and a reduced memory-to-vCPU ratio. They have indicated if we built an instance with this set of attributes, it would be an excellent fit for workloads such as gaming, financial applications, simulation modeling applications such as those used in the automobile, aerospace, energy and telecommunication industries, and High Performance Computing (HPC).

Introducing M5zn
Building on the success of the z1d instances, we are launching M5zn instances in seven sizes today. These instances use 2nd generation custom Intel^® Xeon^® Scalable (Cascade Lake) processors with a sustained all-core turbo clock frequency of up to 4.5 GHz. M5zn instances feature high frequency processing, are a variant of the general-purpose M5 instances, and are built on the AWS Nitro System. These instances also feature low latency 100 Gbps networking and the Elastic Fabric Adapter (EFA), in order to improve performance for HPC and communication-intensive applications.

Here are the M5zn instances (all VPC-only, HVM-only, and EBS-Optimized, with support for Optimize vCPU). As you can see, the memory-to-vCPU ratio on these instances is half that of the existing z1d instances:

Instance Name	vCPUs	RAM	Network Bandwidth	EBS-Optimized Bandwidth
m5zn.large	2	8 GiB	Up to 25 Gbps	Up to 3.170 Gbps
m5zn.xlarge	4	16 GiB	Up to 25 Gbps	Up to 3.170 Gbps
m5zn.2xlarge	8	32 GiB	Up to 25 Gbps	3.170 Gbps
m5zn.3xlarge	12	48 GiB	Up to 25 Gbps	4.750 Gbps
m5zn.6xlarge	24	96 GiB	50 Gbps	9.500 Gbps
m5zn.12xlarge	48	192 GiB	100 Gbps	19 Gbps
m5zn.metal	48	192 GiB	100 Gbps	19 Gbps

The Nitro Hypervisor allows M5zn instances to deliver performance that is just about indistinguishable from bare metal. Other AWS Nitro System components such as the Nitro Security Chip and hardware-based processing for EBS increase performance, while VPC encryption provides greater security.

Things To Know
Here are a couple of “fun facts” about the M5zn instances:

Placement Groups – M5zn instances can be used in Cluster (for low latency and high network throughput), Spread (to keep critical instances separate from each other), and Partition (to reduce correlated failures) placement groups.

Networking – M5zn instances support the Elastic Network Adapter (ENA) with dedicated 100 Gbps network connections and a dedicated 19 Gbps connection to EBS. If you are building distributed ML or HPC applications for use on a cluster of M5zn instances, be sure to take a look at the Elastic Fabric Adapter (EFA). Your HPC applications can use the Message Passing Interface (MPI) to communicate efficiently at high speed while scaling to thousands of nodes.

C-State Control – You can configure CPU Power Management on m5zn.6xlarge and m5zn.12xlarge instances. This is definitely an advanced feature, but one worth exploring in those situations where you need to squeeze every possible cycle of available performance from the instance.

NUMA – You can make use of Non-Uniform Memory Access on m5zn.12xlarge instances. This is also an advanced feature, but worth exploring in situations where you have an in-depth understanding of your application’s memory access patterns.

To learn more about these and other features, visit the EC2 M5 Instances page.

Available Now
As you can see, the M5zn instances are a great fit for gaming, HPC and simulation modeling workloads such as those used by the financial, automobile, aerospace, energy, and telecommunications industries.

You can launch M5zn instances today in the US East (N. Virginia), US East (Ohio), US West (N. California), US West (Oregon), Europe (Ireland), Europe (Frankfurt), and Asia Pacific (Tokyo) Regions in On-Demand, Reserved Instance, Savings Plan, and Spot form. Dedicated Instances and Dedicated Hosts are also available.

Support is available in the EC2 Forum or via your usual AWS Support contact. The EC2 team is interested in your feedback and you can contact them at ec2-m5zn-customer-feedback@amazon.com.

— Jeff;

 

 

Via AWS News Blog https://ift.tt/1EusYcK

Coming Soon – EC2 C6gn Instances – 100 Gbps Networking with AWS Graviton2 Processors

Based on the amazing feedback from customers such as Snap, NextRoll, Intuit, SmugMug, and Honeycomb who are running their workloads on Amazon Elastic Compute Cloud (EC2) instances powered by AWS Graviton2, today we are announcing an addition to our broad Arm-based Graviton2 portfolio with C6gn instances that deliver up to 100 Gbps network bandwidth, up to 38 Gbps Amazon Elastic Block Store (EBS) bandwidth, up to 40% higher packet processing performance, and up to 40% better price/performance versus comparable current generation x86-based network optimized instances.

Compared to C6g instances, this new instance type provides 4x higher network bandwidth, 4x higher packet processing performance, and 2x higher EBS bandwidth. This means that customers with workloads that need high networking bandwidth such as high performance computing (HPC), network appliance, real-time video communications, and data analytics, will be able to bring their biggest and most challenging applications to Arm and take advantage of the performance and cost-optimization.

C6gn instances will be available in 8 sizes:

Name	vCPUs	Memory (GiB)	Network Bandwidth (Gbps)	EBS Throughput (Gbps)
c6gn.medium	1	2	Up to 25	Up to 9.5
c6gn.large	2	4	Up to 25	Up to 9.5
c6gn.xlarge	4	8	Up to 25	Up to 9.5
c6gn.2xlarge	8	16	Up to 25	Up to 9.5
c6gn.4xlarge	16	32	25	9.5
c6gn.8xlarge	32	64	50	19
c6gn.12xlarge	48	96	75	28.5
c6gn.16xlarge	64	128	100	38

The new instances are built on the AWS Nitro System, a collection of AWS-designed hardware and software innovations that maximize resource efficiency. C6gn instances support Elastic Fabric Adapter (EFA) on the c6gn.16xlarge sizes for workloads that can take advantage of lower network latency (such as HPC and video processing) and use Message Passing Interface (MPI) for highly scalable clusters. These new instances also fully support network frameworks like Data Plane Development Kit (DPDK), making it easier to migrate network appliance workloads.

Coming Soon
EC2 C6gn instances will be available later this month and make it easier to optimize costs for HPC and workloads that require high network bandwidth and low latency. Let me know what you are going to build with them!

To get practice with the AWS Graviton2 architecture, you can try t4g.micro instances for free for up to 750 hours per month until March 31st, 2021.

Learn more about EC2 C6gn instances today.

— Danilo

Via AWS News Blog https://ift.tt/1EusYcK

New – Amazon EC2 R5b Instances Provide 3x Higher EBS Performance

In July 2018, we announced memory-optimized R5 instances for the Amazon Elastic Compute Cloud (Amazon EC2). R5 instances are designed for memory-intensive applications such as high-performance databases, distributed web scale in-memory caches, in-memory databases, real time big data analytics, and other enterprise applications.

R5 instances offer two different block storage options. R5d instances offer up to 3.6TB of NMVe instance storage for applications that need access to high-speed, low latency local storage. In addition, all R5b instances work with Amazon Elastic Block Store. Amazon EBS is an easy-to-use, high-performance and highly available block storage service designed for use with Amazon EC2 for both throughput- and transaction-intensive workloads at any scale. A broad range of workloads, such as relational and non-relational databases, enterprise applications, containerized applications, big data analytics engines, file systems, and media workflows are widely deployed on Amazon EBS.

Today, we are happy to announce the availability of R5b, a new addition to the R5 instance family. The new R5b instance is powered by the AWS Nitro System to provide the best network-attached storage performance available on EC2. This new instance offers up to 60Gbps of EBS bandwidth and 260,000 I/O operations per second (IOPS).

Amazon EC2 R5b Instance
Many customers use R5 instances with EBS for large relational database workloads such as commerce platforms, ERP systems, and health record systems, and they rely on EBS to provide scalable, durable, and high availability block storage. These instances provide sufficient storage performance for many use cases, but some customers require higher EBS performance on EC2.

R5 instances provide bandwidth up to 19Gbps and maximum EBS performance of 80K IOPS, while the new R5b instances support bandwidth up to 60Gbps and EBS performance of 260K IOPS, providing 3x higher EBS-Optimized performance compared to R5 instances, enabling customers to lift and shift large relational databases applications to AWS. R5b and R5 vCPU to memory ratio and network performance are the same.

Instance Name	vCPUs	Memory	EBS Optimized Bandwidth (Mbps)	EBS Optimized IOPs@16KiB (IO/s)
r5b.large	2	16 GiB	Up to 10,000	Up to 43,333
r5b.xlarge	4	32 GiB	Up to 10,000	Up to 43,333
r5b.2xlarge	8	64 GiB	Up to 10,000	Up to 43,333
r5b.4xlarge	16	128 GiB	10,000	43,333
r5b.8xlarge	32	256 GiB	20,000	86,667
r5b.12xlarge	48	384 GiB	30,000	130,000
r5b.16xlarge	64	512 GiB	40,000	173,333
r5b.24xlarge	96	768 GiB	60,000	260,000
r5b.metal	96	768 GiB	60,000	260,000

Customers operating storage performance sensitive workloads can migrate from R5 to R5b to consolidate their existing workloads into fewer or smaller instances. This can reduce the cost of both infrastructure and licensed commercial software working on those instances. R5b instances are supported by Amazon RDS for Oracle and Amazon RDS for SQL Server, simplifying the migration path for large commercial database applications and improving storage performance for current RDS customers by up to 3x.

All Nitro compatible AMIs support R5b instances, and the EBS-backed HVM AMI must have NVMe 1.0e and ENA drivers installed at R5b instance launch. R5b supports io1, io2 Block Express (in preview), gp2, gp3, sc1, st1 and standard volumes. R5b does not support io2 volumes and io1 volumes that have multi-attach enabled, which are coming soon.

Available Today

R5b instances are available in the following regions: US West (Oregon), Asia Pacific (Tokyo), US East (N. Virginia), US East (Ohio), Asia Pacific (Singapore), and Europe (Frankfurt). RDS on r5b is available in US East (Ohio), Asia Pacific (Singapore), and Europe (Frankfurt), and support in other regions is coming soon.

Learn more about EC2 R5 instances and get started with Amazon EC2 today.

– Kame;

Via AWS News Blog https://ift.tt/1EusYcK

EC2 Update – D3 / D3en Dense Storage Instances

We have launched several generations of EC2 instances with dense storage including the HS1 in 2012 and the D2 in 2015. As you can guess from the name, our customers use these instances when they need massive amounts of very economical on-instance storage for their data warehouses, data lakes, network file systems, Hadoop clusters, and the like. These workloads demand plenty of I/O and network throughput, but work fine with a high ratio of storage to compute power.

New D3 and D3en Instances
Today we are launching the D3 and D3en instances. Like their predecessors, they give you access to massive amounts of low-cost on-instance HDD storage. The D3 instances are available in four sizes, with up to 32 vCPUs and 48 TB of storage. Here are the specs:

Instance Name	vCPUs	RAM	HDD Storage	Aggregate Disk Throughput (128 KiB Blocks)	Network Bandwidth	EBS-Optimized Bandwidth
d3.xlarge	4	32 GiB	6 TB (3 x 2 TB)	580 MiBps	Up to 15 Gbps	850 Mbps
d3.2xlarge	8	64 GiB	12 TB (6 x 2 TB)	1,100 MiBps	Up to 15 Gbps	1,700 Mbps
d3.4xlarge	16	128 GiB	24 TB (12 x 2 TB)	2,300 MiBps	Up to 15 Gbps	2,800 Mbps
d3.8xlarge	32	256 GiB	48 TB (24 x 2 TB)	4,600 MiBps	25 Gbps	5,000 Mbps

As you can see from the table above, the D3 instances are available in the same configurations as the D2 instances for easy migration. You’ll get 5% more memory per vCPU, a 30% boost in compute power, and 2.5x higher network performance if you migrate from D2 to D3. The instances provide low-cost dense storage that delivers high performance sequential access to large data sets. They are perfect for distributed file systems such as HDFS and MapR FS, big data analytical workloads, data warehouses, log processing, and data processing.

The D3en instances are available in six sizes, with up to 48 vCPUs and 336 TB of storage. Here are the specs:

Instance Name	vCPUs	RAM	HDD Storage	Aggregate Disk Throughput (128 KiB Blocks)	Network Bandwidth	EBS-Optimized Bandwidth
d3en.xlarge	4	16 GiB	28 TB (2 x 14 TB)	500 MiBps	Up to 25 Gbps	850 Mbps
d3en.2xlarge	8	32 GiB	56 TB (4 x 14 TB)	1,000 MiBps	Up to 25 Gbps	1,700 Mbps
d3en.4xlarge	16	64 GiB	112 TB (8 x 14 TB)	2,000 MiBps	25 Gbps	2,800 Mbps
d3en.6xlarge	24	96 GiB	168 TB (12 x 14 TB)	3,100 MiBps	40 Gbps	4,000 Mbps
d3en.8xlarge	32	128 GiB	224 TB (16 x 14 TB)	4,100 MiBps	50 Gbps	5,000 Mbps
d3en.12xlarge	48	192 GiB	336 TB (24 x 14 TB)	6,200 MiBps	75 Gbps	7,000 Mbps

The D3en instances have a high ratio of storage to vCPU, and are optimized for high throughput and high sequential I/O to very large data sets, with a cost-per-TB that is 80% lower than on D2 instances. D3en instances can host Lustre, BeeGFS, GPFS, and other distributed file systems, they can store your data lakes, and they can run your Amazon EMR, Spark, and Hadoop analytical workloads.

Both of the instance types are built on the AWS Nitro System and are powered by custom Intel^® Second Generation Scalable Xeon^® (Cascade Lake) processors that can deliver all-core turbo performance of up to 3.1 GHz. The HDD storage is encrypted at rest using AES-256-XTS; traffic between D3 or D3en instances in the same VPC or within peered VPCs is encrypted using a 256-bit key.

Things to Know
Here are a couple of things that you should keep in mind regarding the D3 and D3en instances:

Regions – D3en instances are available in the US East (N. Virginia), US West (Oregon), and Europe (Ireland) Regions; D3en instances are available in all of those regions and also in the US East (Ohio) Region, with more regions coming soon.

Purchase Options – You can purchase D3 and D3 instances in On-Demand, Savings Plan, Reserved Instance, Spot, and Dedicated Instance form.

AMIs – You must use AMIs that include the Elastic Network Adapter (ENA) and NVMe drivers.

Now Available
D3 and D3en instances are available now and you can start using them today!

— Jeff;

Via AWS News Blog https://ift.tt/1EusYcK