Packet Launches Edge Compute Service to Power Distributed Workloads

July 26, 2017 Off By Hoofer
Grazed from Packet

Packet, the leading bare metal cloud for developers, today launched a new Edge Compute service with 11 additional global locations and a dynamic spot-market based pricing feature. The service expands upon Packet’s existing bare metal cloud and is targeted at latency specific workloads and software innovators that require global access to un-opinionated infrastructure without the use of virtualization or multi-tenancy.

New locations include Los Angeles, Seattle, Dallas, Chicago, Ashburn, Atlanta, Toronto, Frankfurt, Singapore, Hong Kong, and Sydney – with Paris, London, Sao Paulo, and Mumbai coming online by October. This expands upon Packet’s existing footprints in New York Metro, Sunnyvale, Amsterdam, and Tokyo.

The new Edge Compute locations feature a single powerful server configuration – a “Type 1E” instance based on an Intel SkyLake processor – as well as Packet’s full networking and infrastructure automation suite. Instances are deployable within minutes via API, standard DevOps tools or the Packet portal, and take advantage of a new spot market feature for demand-driven pricing. Customers can leverage their own IP space, deploy global anycast networks and build custom Layer 2 or SD-WAN networks across Packet’s scalable transit and transport network.


Infrastructure for the Internet of Tomorrow

The next wave of internet infrastructure, coupled with low latency, high speed 5G connectivity, will power everything from connected IoT devices to self-driving cars and augmented reality experiences.

While there are a plethora of Content Delivery Networks (CDNs) serving up cached content at the edge, the heavy lifting needed for compute often requires a trip of hundreds or thousands of miles to reach centralized public clouds such as Amazon Web Services or Google’s Cloud Platform. This distance – paired with the inconsistency and abstraction created by multi-tenant virtualization – adds latency and hinders the development of new services and experiences for customers.   

“While edge compute is still in its infancy, new experiences are driving demand for distributed infrastructure, especially as software continues its relentless pursuit down the stack,” said Zachary Smith, a co-founder and CEO of Packet. “We believe that the developers building these new experiences are hungry for distributed, unopinionated and yet fully automated compute infrastructure and that’s what we’re bringing to the market today.”

Innovating on Physical Hardware

Packet’s combination of dedicated physical hardware, uncongested network, and now 15 global locations provides a unique infrastructure underlay that until now has been accessible only to internet giants or those willing to stitch together an expensive network of colocation facilities and providers.

“What they say in real estate is even more relevant when it comes to network performance: location, location, location,” added Adam Rothschild, SVP of Network at Packet. “But more than just reducing latency, this expansion is about providing developer-style access to dedicated hardware at a global scale.”

The key is Packet’s proprietary technology for automating the lifecycle of infrastructure at the lowest layer, which is achieved without multi-tenancy, virtualization, or other abstractions. This means developers can leverage physical machines in the same way as virtual machines, enabling them to quickly develop new applications that take advantage of advances in hardware and specialty silicon.

Edge Compute locations will feature a new general purpose “Type 1E” configuration with additional locations and systems featuring ARMv8 processors, GPU / FPGA’s and other options to come.

Details and New Features

The Edge Compute service is available today by request at and provides users with a powerful dedicated bare metal server configuration in less than 8 minutes. The Type 1E features an Intel E3-1578L v5 processor, Intel IRIS GPU, 32GB RAM, 240GB SSD, and 10 Gbps Network interfaces.

While the “Type 1E” lists at $0.50/hr, they are also available via Packet’s new spot market, which offers marketplace-style pricing for infrastructure in all of Packet’s global 15 locations. Packet’s bare metal spot market functions similarly to the AWS – allowing users to take advantage of dramatically reduced pricing in exchange for revocable status.

For users requiring customized hardware at the edge, Packet is extending its popular “Private Deployment” to all global locations. With Private Deployments, users can deploy custom configurations while benefiting from Packet’s platform and network automation.

As with Packet’s existing bare metal cloud, the new Edge Compute locations feature Packet’s robust API, language-based libraries and a wide list of integrations including Terraform, Ansible, Canonical’s Juju, libCloud, Docker’s LinuxKit and jClouds enable developers to deploy bare metal resources around the globe as easily as a managing resources at AWS or Digital Ocean.

“The Packet platform is designed to automate infrastructure in hundreds or thousands of locations,” noted Nathan Goulding, SVP of Engineering. “Today’s announcement is a major first step in making scalable, non-opinionated infrastructure available to developers on a truly global basis.”