Inside Nvidia’s NVSwitch GPU Interconnect

At the GPU Technology Conference last week, we told you all about the new NVSwitch memory fabric interconnect that Nvidia has created to link multiple “Volta” GPUs together and that is at the heart of the DGX-2 system that the company has created to demonstrate its capabilities and to use on its internal Saturn V supercomputer at some point in the future.

Since the initial announcements, more details have been revealed by Nvidia about NVSwitch, including details of the chip itself and how it helps applications wring a lot more performance from the GPU accelerators.

Our first observation upon looking …

Inside Nvidia’s NVSwitch GPU Interconnect was written by Timothy Prickett Morgan at The Next Platform.

Where will Microsoft spend $5B on IoT?

Microsoft’s announcement of $5 billion in new IoT investment is a public demonstration of the company’s commitment to the internet of things, but it's not immediatly clear what it will spend all that money on.In a statement announcing the new spending – to be spread out over the next four years – Microsoft cited research from A.T. Kearney that said IoT will bring a nearly $2 trillion productivity increase to the global economy and a $177 billion reduction in business costs by the end of the decade.[ For more on IoT see tips for securing IoT on your network, our list of the most powerful internet of things companies and learn about the industrial internet of things. | Get regularly scheduled insights by signing up for Network World newsletters. ] There are a lot of use cases for IoT in the enterprise, and Microsoft listed several customer wins in its announcement, including Johnson Controls, Kohler and  the Alaska Department of Transportation. Gartner research vice president for AI and IoT Mark Hung said that it’s possible to divide those cases into internal and external uses.To read this article in full, please click here

IDG Contributor Network: Intent-based networking can turn your data center into a digital crime scene

An important side effect of digital transformation is that your network is likely to become a digital crime scene. As such, it needs a systematic approach to identify the culprit. In this analogy, a crime can be equivalent to a network outage or gray failure. And this is where intent-based networking (IBN) can help.The general approach in solving a crime like this is to collect as much information as possible, as soon as possible, and to narrow down the pool of suspects. So, let’s see via an example what role IBN plays in all this.Digital crime scene profiling Without “intent” you don’t even know that a crime has been committed. Finding traces of blood in a room in a blood bank or hospital are expected. Finding traces of blood in a room of a home of a missing person is a different matter. But without intent it’s hard to distinguish a blood bank from a home. In a similar manner, dropping a packet of an intruder or forbidden traffic source is a good thing. Dropping a packet of a customer because of a misconfigured ACL is a bad thing. Intent helps you differentiate the two.To Continue reading

IDG Contributor Network: Intent-based networking can turn your data center into a digital crime scene

An important side effect of digital transformation is that your network is likely to become a digital crime scene. As such, it needs a systematic approach to identify the culprit. In this analogy, a crime can be equivalent to a network outage or gray failure. And this is where intent-based networking (IBN) can help.The general approach in solving a crime like this is to collect as much information as possible, as soon as possible, and to narrow down the pool of suspects. So, let’s see via an example what role IBN plays in all this.Digital crime scene profiling Without “intent” you don’t even know that a crime has been committed. Finding traces of blood in a room in a blood bank or hospital are expected. Finding traces of blood in a room of a home of a missing person is a different matter. But without intent it’s hard to distinguish a blood bank from a home. In a similar manner, dropping a packet of an intruder or forbidden traffic source is a good thing. Dropping a packet of a customer because of a misconfigured ACL is a bad thing. Intent helps you differentiate the two.To Continue reading

Why and how to deploy Voyager

In Part I of this blog series, “What is the open packet optical switch, Voyager?”, we discussed the challenges and remedies for providing additional bandwidth for intra and inter data center connections. DWDM is a powerful technology that provides hundreds of gigabits of bandwidth over hundreds or thousands of kilometers using just a fiber pair. We also reviewed some information about DWDM networks and transponder functionality. Voyager provides all the functionality of Cumulus Linux running on a Broadcom Tomahawk based switch and integrates the transponders into the switch itself, all in 1RU. This makes it the first open and fully integrated box operating at DWDM, Layer 2 and Layer 3 in 1RU, making it extremely flexible.

Incorporating routing, switching and DWDM in one node could mean fewer boxes needed for the network, since DWDM functionality could be incorporated directly into border leafs. Because it runs Cumulus Linux (CL), all CL data center functionality, such as VXLAN Routing with EVPN, is also supported. For example, a pair of Voyager nodes can be used as VXLAN routing centralized routers with EVPN, hosting VXLAN VTEPs, running MLAG, and provide the long distance DWDM connectivity all in one box!

Voyager also Continue reading

Another Step Toward FPGAs in Supercomputing

There has been plenty of talk about where FPGA acceleration might fit into high performance computing but there are only a few testbeds and purpose-built clusters pushing this vision forward for scientific applications.

While we do not necessarily expect supercomputing centers to turn their backs on GPUs as the accelerator of choice in favor of FPGAs anytime in the foreseeable future. there is some smart investment happening in Europe and to a lesser extent, in the U.S. that takes advantage of recent hardware additions and high-level tool development that put field programmable devices within closer reach–even for centers whose users …

Another Step Toward FPGAs in Supercomputing was written by Nicole Hemsoth at The Next Platform.

ONOS traffic analytics

Open Network Operating System (ONOS) is "a software defined networking (SDN) OS for service providers that has scalability, high availability, high performance, and abstractions to make it easy to create applications and services." The open source project is hosted by the Linux Foundation.

Mininet and onos.py workflow describes how to run ONOS using the Mininet network emulator. Mininet allows virtual networks to be quickly constructed and is a simple way to experiment with ONOS. In addition, Mininet flow analytics describes how to enable industry standard sFlow streaming telemetry in Mininet, proving a simple way monitor traffic in the ONOS controlled network.

For example, the following command creates a Mininet network, controlled by ONOS, and monitored using sFlow:
sudo mn --custom ~/onos/tools/dev/mininet/onos.py,sflow-rt/extras/sflow.py \
--link tc,bw=10 --controller onos,1 --topo tree,2,2
The screen capture above shows the network topology in the ONOS web user interface.
Install Mininet dashboard to visualize the network traffic. The screen capture above shows a large flow over the same topology being displayed by ONOS, see Mininet weathermap for more examples.

In this case, the traffic was created by the following Mininet command: 
mininet-onos> iperf h1 h3
The screen capture above shows top flows, busiest Continue reading

Reaction: The NRE as the new architect

Over at the Packet Pushers, Anthony Miloslavsky suggests that network architects have outlived their usefulness, so it is time to think of a new role. He describes a role called the “NRE” to replace the architect; the NRE would—

…spend no less than 50% of their time focusing on automation, while spending the other 50% deeply embedded in the operations/engineering/architecture realms of networking. They participate in an on-call rotation to stay in touch with the ops side of the house, with a focus on “treating operations as if it’s a software problem” in response. NREs would provide a expert big picture view of BOTH the development/automation and network operation/design sides of the house.

The author goes on to argue that we need someone who will do operations, engineering, architecture, and development because “pure architecture” folks tend to “lose touch” with the operations side of things. It is too easy to “throw a solution over the cubicle wall” without considering the implementation and operational problems. But, as a friend used to ask of everything when I was still in electronics, will it work? I suspect the answer is no for several reasons.

First, there is no such person as described, and Continue reading

1 … 1,712 1,713 1,714 1,715 1,716 … 3,894