AMD is introducing “Seattle,” a 64-bit ARM-based server System-on a –Chip (SoC) built on the same technology that powers billions of today’s most popular mobile devices.
3. Overview:
The Push for Low-Power, Parallel Processing in Large Cloud Data Centers
The explosive demand for compute cycles generated by mobile devices, such
as tablets and smartphones, is driving fundamental shifts in server deployment
models. With much of the computing generated by these devices occurring
on the server side, operators of cloud data centers are facing huge pressures
to meet compute demands while minimizing their physical footprint or power
consumption. Data center managers are looking at opportunities to offload
certain parallel processing tasks onto new server and processor designs that
use numerous small, low-power processors working in unison in dense clusters.
AMD’s Data Center Strategy – Delivering Performance that Matters
We understand that compute requirements for many workloads are different
today. We also recognize that Total Cost of Ownership (TCO) has become an
essential measure of infrastructure deployments as the standard by which
server hardware is evaluated.
Our response to this impending shift in the data center has been to expand
processor choice and deliver platforms that are specifically tuned for
targeted workloads. We believe that by building server processors that
are purpose-designed for specific workloads, we can enable the optimum
amount of performance to efficiently carry out tasks while helping to provide
substantial cost improvements in terms of compute per-watt per-dollar.
Continuing on our long history of datacenter innovation, we have introduced
low-power, small footprint CPUs as well as Accelerated Processing Units
(APUs) for servers that help enable major efficiency gains in power use and
costs for multimedia-oriented workloads.
We have now added “Seattle,” a 64-bit ARM-based server processor, to
our server processor offering. Seattle combines the ARM A57 Core on a
System-on a–Chip (SoC) that contains all the standard server processor
elements you would expect from AMD. Seattle will be the industry’s only
64-bit ARM-based server SoC from a proven server processor supplier.
Introducing the industry’s only
64-bit ARM-based server SoC from
a proven server processor supplier.
Target Workloads for AMD ARM SoCs
With ARM and AMD, you can expect:
Application-optimized performance
Maximum choice and flexibility
Extreme compute, space, and energy efficiency
Why AMD + ARM?
AMD is exceptionally well-suited to bringing ARM to the data center. We
have the server architecture engineering expertise. We have the experience
in dealing with the multiple layers of software integration required for
excellence in server processor design. We have invested more than a decade
in becoming a great server processor vendor and building an IP portfolio for
server applications. We have a history of pioneering disruptive technologies
for the data center, most notably introducing the first 64-bit implementation,
the first native multi-core processor, and the first hardware enabled
virtualization. It’s not a turn-key proposition to build chips that power
mission-critical applications in the data center. Our hard-won expertise
gives us mastery over a myriad of details and puts us farther down the
learning curve to deliver a truly disruptive and meaningful datacenter
infrastructure solution.
As a building block architecture, it is much more possible with ARM to
innovate and develop custom software/chip architectures that can deliver
meaningful gains in the processor’s execution of code instructions at low
power. Such innovation and custom tuning are the ultimate drivers of great
performance per watt and low total cost of ownership (TCO).
ARM
Cloud Computing
AMD+ARM
Media
Web Hosting
Data Analytics
®
AMD’s ARM Ecosystem
We are engaged with the ARM server processor ecosystem that will lead
the way to innovation and expanded choice. Our partnerships with builders
of ultra-dense server platforms, such as HP with Moonshot, gives us the
delivery vehicle necessary to enable the optimal performance of ARM-based
SoCs. Our participation in the Linaro organization, an engineering consortium
that is committed to consolidating and optimizing open source software
for the ARM architecture, helps us deliver advances in running standardized
Linux-based server OS for ARM infrastructure. It is the scope of this type of
software innovation that will drive gains in workload-optimized and energy
efficient parallel processing.
For more information
about AMD’s ARM
SoC products, please
visit www.amd.com
4. Overview:
The Push for Low-Power, Parallel Processing in Large Cloud Data Centers
The explosive demand for compute cycles generated by mobile devices, such
as tablets and smartphones, is driving fundamental shifts in server deployment
models. With much of the computing generated by these devices occurring
on the server side, operators of cloud data centers are facing huge pressures
to meet compute demands while minimizing their physical footprint or power
consumption. Data center managers are looking at opportunities to offload
certain parallel processing tasks onto new server and processor designs that
use numerous small, low-power processors working in unison in dense clusters.
AMD’s Data Center Strategy – Delivering Performance that Matters
We understand that compute requirements for many workloads are different
today. We also recognize that Total Cost of Ownership (TCO) has become an
essential measure of infrastructure deployments as the standard by which
server hardware is evaluated.
Our response to this impending shift in the data center has been to expand
processor choice and deliver platforms that are specifically tuned for
targeted workloads. We believe that by building server processors that
are purpose-designed for specific workloads, we can enable the optimum
amount of performance to efficiently carry out tasks while helping to provide
substantial cost improvements in terms of compute per-watt per-dollar.
Continuing on our long history of datacenter innovation, we have introduced
low-power, small footprint CPUs as well as Accelerated Processing Units
(APUs) for servers that help enable major efficiency gains in power use and
costs for multimedia-oriented workloads.
We have now added “Seattle,” a 64-bit ARM-based server processor, to
our server processor offering. Seattle combines the ARM A57 Core on a
System-on a–Chip (SoC) that contains all the standard server processor
elements you would expect from AMD. Seattle will be the industry’s only
64-bit ARM-based server SoC from a proven server processor supplier.
Introducing the industry’s only
64-bit ARM-based server SoC from
a proven server processor supplier.
Target Workloads for AMD ARM SoCs
With ARM and AMD, you can expect:
Application-optimized performance
Maximum choice and flexibility
Extreme compute, space, and energy efficiency
Why AMD + ARM?
AMD is exceptionally well-suited to bringing ARM to the data center. We
have the server architecture engineering expertise. We have the experience
in dealing with the multiple layers of software integration required for
excellence in server processor design. We have invested more than a decade
in becoming a great server processor vendor and building an IP portfolio for
server applications. We have a history of pioneering disruptive technologies
for the data center, most notably introducing the first 64-bit implementation,
the first native multi-core processor, and the first hardware enabled
virtualization. It’s not a turn-key proposition to build chips that power
mission-critical applications in the data center. Our hard-won expertise
gives us mastery over a myriad of details and puts us farther down the
learning curve to deliver a truly disruptive and meaningful datacenter
infrastructure solution.
As a building block architecture, it is much more possible with ARM to
innovate and develop custom software/chip architectures that can deliver
meaningful gains in the processor’s execution of code instructions at low
power. Such innovation and custom tuning are the ultimate drivers of great
performance per watt and low total cost of ownership (TCO).
ARM
Cloud Computing
AMD+ARM
Media
Web Hosting
Data Analytics
®
AMD’s ARM Ecosystem
We are engaged with the ARM server processor ecosystem that will lead
the way to innovation and expanded choice. Our partnerships with builders
of ultra-dense server platforms, such as HP with Moonshot, gives us the
delivery vehicle necessary to enable the optimal performance of ARM-based
SoCs. Our participation in the Linaro organization, an engineering consortium
that is committed to consolidating and optimizing open source software
for the ARM architecture, helps us deliver advances in running standardized
Linux-based server OS for ARM infrastructure. It is the scope of this type of
software innovation that will drive gains in workload-optimized and energy
efficient parallel processing.
For more information
about AMD’s ARM
SoC products, please
visit www.amd.com
5. Overview:
The Push for Low-Power, Parallel Processing in Large Cloud Data Centers
The explosive demand for compute cycles generated by mobile devices, such
as tablets and smartphones, is driving fundamental shifts in server deployment
models. With much of the computing generated by these devices occurring
on the server side, operators of cloud data centers are facing huge pressures
to meet compute demands while minimizing their physical footprint or power
consumption. Data center managers are looking at opportunities to offload
certain parallel processing tasks onto new server and processor designs that
use numerous small, low-power processors working in unison in dense clusters.
AMD’s Data Center Strategy – Delivering Performance that Matters
We understand that compute requirements for many workloads are different
today. We also recognize that Total Cost of Ownership (TCO) has become an
essential measure of infrastructure deployments as the standard by which
server hardware is evaluated.
Our response to this impending shift in the data center has been to expand
processor choice and deliver platforms that are specifically tuned for
targeted workloads. We believe that by building server processors that
are purpose-designed for specific workloads, we can enable the optimum
amount of performance to efficiently carry out tasks while helping to provide
substantial cost improvements in terms of compute per-watt per-dollar.
Continuing on our long history of datacenter innovation, we have introduced
low-power, small footprint CPUs as well as Accelerated Processing Units
(APUs) for servers that help enable major efficiency gains in power use and
costs for multimedia-oriented workloads.
We have now added “Seattle,” a 64-bit ARM-based server processor, to
our server processor offering. Seattle combines the ARM A57 Core on a
System-on a–Chip (SoC) that contains all the standard server processor
elements you would expect from AMD. Seattle will be the industry’s only
64-bit ARM-based server SoC from a proven server processor supplier.
Introducing the industry’s only
64-bit ARM-based server SoC from
a proven server processor supplier.
Target Workloads for AMD ARM SoCs
With ARM and AMD, you can expect:
Application-optimized performance
Maximum choice and flexibility
Extreme compute, space, and energy efficiency
Why AMD + ARM?
AMD is exceptionally well-suited to bringing ARM to the data center. We
have the server architecture engineering expertise. We have the experience
in dealing with the multiple layers of software integration required for
excellence in server processor design. We have invested more than a decade
in becoming a great server processor vendor and building an IP portfolio for
server applications. We have a history of pioneering disruptive technologies
for the data center, most notably introducing the first 64-bit implementation,
the first native multi-core processor, and the first hardware enabled
virtualization. It’s not a turn-key proposition to build chips that power
mission-critical applications in the data center. Our hard-won expertise
gives us mastery over a myriad of details and puts us farther down the
learning curve to deliver a truly disruptive and meaningful datacenter
infrastructure solution.
As a building block architecture, it is much more possible with ARM to
innovate and develop custom software/chip architectures that can deliver
meaningful gains in the processor’s execution of code instructions at low
power. Such innovation and custom tuning are the ultimate drivers of great
performance per watt and low total cost of ownership (TCO).
ARM
Cloud Computing
AMD+ARM
Media
Web Hosting
Data Analytics
®
AMD’s ARM Ecosystem
We are engaged with the ARM server processor ecosystem that will lead
the way to innovation and expanded choice. Our partnerships with builders
of ultra-dense server platforms, such as HP with Moonshot, gives us the
delivery vehicle necessary to enable the optimal performance of ARM-based
SoCs. Our participation in the Linaro organization, an engineering consortium
that is committed to consolidating and optimizing open source software
for the ARM architecture, helps us deliver advances in running standardized
Linux-based server OS for ARM infrastructure. It is the scope of this type of
software innovation that will drive gains in workload-optimized and energy
efficient parallel processing.
For more information
about AMD’s ARM
SoC products, please
visit www.amd.com