Optical Cables Light Up InfiniBand

By Michael Feldman

August 17, 2007

As InfiniBand data rates rise and clusters grow in size, copper cable technology is getting stretched to its limits. The current InfiniBand Double Data Rate (DDR) of 20 Gigabits per second (Gbps) means that conventional copper cable is limited to an 8 to 10 meter reach. While this might be fine for moderate sized clusters, as systems expand, the physical layout of the cluster becomes a real challenge.

The problems of copper as an interconnect media are numerous. It’s heavy and bulky, making it difficult to physically manage in the datacenter. The bulkiness is a particular problem when it comes to maintaining airflow around the systems to maintain proper cooling. And because of the nature of electrical signals, electromagnetic interference limits copper’s performance and reliability. These problems really start to add up when you have hundreds or thousands of connections in a scaled-out high performance cluster.

Fortunately, optical fibers provide a capable alternative to copper for higher data rates and longer distances. The problem has always been cost. The optical fiber itself is less expensive than copper, but the transceivers required to convert the optical stream to electrical signals makes these solutions prohibitively expensive at shorter distances. That’s why optical is the solution of choice for long-haul networks. At the level of the computer this model is inverted. The highest data rates are on the processor, which represent the shortest communications distances. As you move away from the CPU, onto the board, the backplane, and then out of the box, distance increases and data rates drop. To get optical connections at the volumes required for computer systems, low-cost solutions will be necessary.

That process has begun. Two companies, Intel and Luxtera, have recently unveiled optical cable products aimed at HPC clusters that are using the higher data rates of InfiniBand. The cables will also support 10GbE traffic, but until 40GbE gets established, InfiniBand will be the principle driver for these products. Intel unveiled its optical cables (“Intel Connects Cables”) at the International Supercomputing Conference in June. Luxtera announced its “Blazar” optical offering earlier this week.

Intel Outside

While Intel is mainly known for its microprocessors, the company became aware several years ago that copper-based communication was going to become an issue as CPU performance outpaced communication bandwidth at all levels of the system. The company’s silicon photonics research effort is one instance of their desire to move beyond traditional electrical interconnects. While a complete silicon photonics solution is probably years away, Intel’s optical cables are a commercial product today and represent the company’s answer to the copper problem for cluster interconnects.

Intel’s optical cable offering is aimed squarely at the rapidly expanding HPC cluster computing market, which has the most pressing need for high bandwidth, low latency communication. The cables support SDR and DDR InfiniBand (and 10GbE ) and boasts bit error rates (BER) of at least 10^-15, 1000 times better than top-of-the-line copper. Due to the nature of the optical fiber, they are 84 percent lighter than their copper counterparts, take up 83 percent less space and have a 40 percent better bend radius. The cable assemblies can be built in lengths ranging from 1 to 100 meters, enabling a lot more flexibility in system layout — even allowing for configurations that extend into multiple rooms and multiple floors.

Tom Willis, general manager of Intel Connects Cables, thinks if people are going to build large (1000-plus node) clusters using InfiniBand, they’re going to need this product. He says their business model is based on the assumption that all clusters requiring DDR interconnects of 10 meters and above would go optical, while 10 meters and below would stay as copper. But Willis says they’re actually seeing some orders for sub-10-meter cable assemblies, where people are interested in the better BER and easier manageability of optical.

The product is currently in production release, but deliveries won’t be take place until the end of the year. “This is just the first product,” notes Willis. “Over time, all the high performance links out of your computer are going to be optical.”

Luxtera Raises the Stakes

Luxtera, a Carlsbad, California-based startup, announced their optical cable offering on August 13 and, like Intel, is targeting InfiniBand-enabled HPC clusters. Their optical solution has the same basic advantages over copper as Intel’s offering. Here though, there’s a twist. Luxtera’s product, named “Blazar,” can support all InfiniBand speeds — SDR, DDR and QDR — that is, up to 40 Gbps, and in lengths of up to 300 meters. Blazar will also support four lanes of 10GbE. At 40 Gbps, conventional copper cabling would be limited to 7 meters, essentially making it impractical to go beyond the rack.

“We’ve talked with a number of datacenter managers and none of them expect copper to be a viable media for QDR InfiniBand deployment,” says Luxtera’s Marketing VP, Marek Tlalka.

Although both companies used optical fiber media, the Luxtera technology is quite a bit different from that of Intel’s. Using their CMOS photonics technology, Luxtera was able to integrate most of the transceiver functionality onto a silicon chip. Only the laser and the photodetectors are discrete components (the photodetectors will be integrated in the next-generation products). Intel’s transceiver is based on Vertical-Cavity Surface-Emitting Lasers (VCSELs) and uses an unspecified collection of discrete components.

Because Luxtera catches light from the waveguide right onto the chip, the company is able to use single-mode optical fiber. By contrast, since Intel uses VCSEL technology, multimode fiber is required in their offering. Since the cost of single-mode fiber is less than that of multimode fiber, this may afford Luxtera some price advantage, especially for longer reaches.

The Blazar product will be ready for sampling by development partners in Q4 2007 and generally available by the middle of 2008. This should coincide with the first wave of QDR InfiniBand gear expected in the second half of next year. Tlalka says he’s expecting to attract DDR InfiniBand users to their solution as well. But here, they’ll be going head to head against Intel, a company known to get rather aggressive with pricing.

“Our objective right now as the first product comes to the market is to sell our QDR solutions at price points comparable to Intel’s [DDR] solution,” says Tlalka. “So you get twice as much speed, with three times the reach for the same price.”

Luxtera’s long-term goal is to get prices down to those of copper even for short reaches. Tlalka claims this is mainly a function of volume economics. He says if they were building a million of these products every month, they would probably be there today.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire