Over the next few months, we are planning a series of events to demonstrate the HX network processor’s wirespeed capabilities. Stay tuned for more news and information on 100 Gbit/s wirespeed network processing.

Update:  Some online links to the news coverage on this story -

• Light Reading: Xelerated Finally Gets 100G
• Embedded.com: NPUs – Xelerated samples 100 Gbit/s wirespeed network processor
• Broadband Properties Blog: 100 Gbps Wirespeed Network Processor From Xelerated
• SOC Central: Xelerated Samples 100-Gbps Wirespeed Network Processor
• Converge! Network Digest: Xelerated Samples 100 Gbps Network Processor

That being said, a customer recently recognized us with one of the most prestigious awards available – Huawei’s Support Supplier Award.  We are especially humbled because out of 200+ companies, only 4 were recognized for “the most competent in supply.”  I am so happy and proud of this accomplishment for Xelerated.  It truly is the best token for customer appreciation a company can ask for.

Still, there are limitations to what network processors can do. They are developed for a special-purpose; to process packets very efficiently. What they do, they do very well. But if you want to accomplish anything other than process packets, you will need another type of processor. General-purpose multicore processors would be the choice. But can’t they be used for packet processing as well?

For a deep-dive to this subject, I recommend reading Håkan Zeffer’s and my recent article in Electronic Design. We compare the differences between the special-purpose dataflow architecture, which is found in all Xelerated NPUs, and general-purpose multicore architectures that are popular in today’s server architectures. Both have their merits, and a comparison should be made for the target application. When looking to the architectures, you can start to calculate how efficient the different approaches are for different types of applications. I hope you find the conclusions interesting.

And interestingly enough it is not a tenfold from very low levels, which tends to be the case when we see booming markets numbers being touted. No, the 10G and above market in the enterprise and service provider segments was already in 2009 north of $10 billion ($11.1 to be exact according to the report).

So what drives this insatiable demand for bandwidth?

For the Service Provider segment it is clear that the growth stems from the build-outs of new Fiber Access networks. They are architected for service delivery over Ethernet, and generate traffic in high volumes that is aggregated by switches and routers. In turn, they need interface speeds that are several magnitudes higher than the access side of the network. Already in this early phase of the market, the interest for Gigabit access is real, and growing through for instance the Googles initiative “Fiber for Communities”. Also new products are arriving to meet the new demands. The recently announced high-density gigabit access switch developed jointly by Accton and Xelerated, is perfectly aimed to this market.

The 10G growth is not limited to the service provider segment. Also enterprises are inevitably migrating to 10G interface connectivity. End-users consume more bandwidth, and require that data centers scale in parity with that growth.

Some philosopher said “the best way to predict the future is to create it”.  At Xelerated we make the fastest and most flexible chips for this new market. We are part of the creation. But does this mean we know where future networking will be in 20 years? Sure – let me just take out my crystal ball from the drawer…

At the Expo, the international pavilions compete for attention from the crowd. For example, the U.K. pavilion looks like a hedgehog. And the Danes made a brilliant move when they decided to transport the little Sea-Maid over to the Expo.  The Swedish pavilion has a bar on the top of its yellow painted roof.

Still, what attracts my attention the most are the Chinese pavilions. Yes, there are more than one. China Mobile and China Telecom have sponsored one pavilion in particular about the future of Telecoms, while one of the world’s largest ship manufacturers, CSSC, has another pavilion which looks like a complete power generation site.  And yet another pavilion is dedicated to Chinese Private Enterprises. Anyone in doubt of China’s commitment to entrepreneurship should pay a visit to this pavilion. Although we, who are already convinced, should go there anyway to enjoy the world-class piece of art that concludes the tour.

But of these Chinese pavilions, what catches my eye the most is the story told over at the Information and Communication Pavilion – about the bright future enabled by telecommunications.  At this pavilion, visitors are asked to make their wishes for the future by typing them onto a PSP-like device. We then enter the first of two circular cinemas, the first dedicated to the history of telecommunications, and the latter to the future. Glossy cartoon characters, JiLing and GuDu, guide us through the bright history of telecommunications, and to the even brighter future. All wishes which will one day become true. You can teleport yourself into the future, and back to history, have your favorite basketball player to teach you how to score, have the impressionist icon Vincent Van Gogh to teach you how to paint.

The conclusion isn’t hard to capture, despite technical translation issues. The future is bright. For Telecoms. And for China.

Projected spending in Carrier Ethernet equipment. Source: Infonetics Research.

Michael Howard’s latest research on the Carrier Ethernet market is yet another proof point that service providers continue to migrate their networks to Carrier Ethernet. According to the latest Infonetics Research report investments in Carrier Ethernet gear are expected to go from US$21.6 billion in 2009 to over US$32 billion in 2014.

As previously noted here at Xelerated Xpress, the market for Carrier Ethernet gear continues to outpace investments in other technologies. We may debate over the exact numbers – future projections always need to be considered with a pinch of salt – but the trend is unquestionable; service provider networks are turning Ethernet.

So what implications will this have on 3G and 4G?  The most obvious is the demand for more bandwidth in the radio access network. It will have to be optimized for carrying data traffic. Voice services must be preserved, but data will dominate. And keep in mind that we are just in the early stages of the mobile data explosion.

Carriers have been trying to patch their networks in different ways to support more aggressive mobile data traffic volumes for some years now. Ethernet and circuit emulation in different variants have been introduced. Ethernet is a more high performance transport technology compared to ATM (AAL2 for voice and AAL5 for data). But carriers need rigorous synchronization schemes over Ethernet to make it work.  And there are other challenges like where to terminate legacy services and where to introduce the Ethernet ports? How can the new transport network comply to existing service provisioning schemes?

If the trends observed by Ericsson are correct, we are heading toward a data optimized radio access infrastructure faster than anyone expected. The Ethernet-based transport in the radio access network is needed to cope with the mobile data explosion. There are different attempts to solve this. We can learn from China Mobile’s PTN requirements which utilizes an Ethernet-based transport with synchronization support based on point-to-point (PTP) Ethernet. These networks are designed for data growth, but continue to support voice. And these requirements are here and now. They are designed for 2G, 3G and beyond.

Those who are seriously evaluating multi-core architectures for packet processing should be prepared for some surprises. Again. The same evolution happened several times before with separate processors brought together in a multi-processor architecture on the same die to scale processing performance.  Have people already forgot about why previous multi-core proposals for packet processing did not fly?

First, multi-core architectures consume a lot of power. Second, they are not designed for deterministic wirespeed performance. Third, they are difficult to program efficiently – making it hard to meet the performance requirements in modern packet processing applications.  And using ANSI-C does not help the inefficiency and performance challenges.

Multi-core processors are designed for general purposes, and they are therefore not optimized for packet processing.  They lack the necessary service density. Xelerated’s Dataflow Architecture, in contrast, was designed to solve the challenge of combining programmability and super-efficient packet processing. It is a linearly scalable wirespeed-by-design processing architecture with low power and a great amount of service density.

Support for 40 or 100 G interfaces does not say anything about the device’s ability to perform a meaningful application at these speeds. And when looking into the requirements in advanced Carrier Ethernet, Fiber Access or Mobile Backhaul applications, general-purpose multi-core designs continue to fall short. To give you an idea: Xelerated’s new HX330 has over 900 percent greater service density compared to the most high-end multi-core processor on the market. That is, it has 9 times the processing capacity for network and packet processing!

It is time to learn from history. Multi-core architectures have a bright future in general applications, for the server and consumer markets. Here is where they belong – processing applications, not processing packets.

There is a reason why 20+ NPU vendors that spent multi-million dollars in multi-core architectures failed to deliver a commercial and technically viable option to the networking industry. This history is just ten years away. I’m confident network equipment vendors have a longer memory than this.

According to Point Topic, Asia now accounts for nearly 40% of all broadband subscribers.  Not to mention, the region outperformed the sum of all other regions in terms of net subscriber additions last year. With China, Japan and South Korea leading the way, other Asian countries are growing fast, but from a smaller base. The Philippines grew 60% last year, and India 40%(!).

China passed the 100 million subscriber milestone in the fourth quarter last year, and the pace of growth is not slowing down. There is no doubt these numbers have a profound impact on how the industry is being shaped in support for the emerging fiber-based broadband market.

The Network Processor (NPU) industry is no different. A new generation comes with significant more processing power, and still our customers ask for more. There are more standards coming down the road, and there are more features requested by service providers around the world. Demand is always more, never less.

At the recent Linley tech seminar Xelerated’s Vice President of Business Development, Thomas Eklund, delivered a presentation – available for download on Slideshare – on the importance of service density in the NPU business. If your processing device can’t manage all the network services expected, then you will find your Research & Development department under hard stress to balance features and performance against available resources.  The need for greater service density is an important aspect for the whole service provider industry. If next generation routers and switching platforms can be built with greater headroom for new services, we can extend the lifetime of the equipment, and thereby strengthen the business case for the broadband services.

Service density can be a bit hard to measure, but doing some very basic calculations on the raw service processing capabilities of a particular chip is fairly straightforward and this will give you a rough idea on the capabilities of the chip. Failing to do this in an evaluation process for a new line card design often leads to unhappy surprises at a very late stage in the project. In turn this results in risk of missing an important market window, as well as signficantly increased engineering and product costs.  So take a look and find out – how much service density is your chip providing?

Update: this post was re-published at Advanced TCA Systems Web Site.