ARM Muscles In on Intel’s Dominance in Datacenters

By Richard L. Brandt

January 28, 2013

In late 2007, Robert Hormuth called a former colleague he had worked with when they were both at Intel: Ian Drew, now a marketing vice president at ARM Holdings in the UK. Hormuth had moved on to Dell Computer’s Server and Architecture team, and wanted to explore a new idea.

Instead of relying exclusively on Intel’s powerful microprocessors to build the servers that handle billions of bits of data flowing through datacenters every second, would it be possible to use smaller, less powerful and lower-power ARM chips, now popular for cell phones and other hand-held devices?

“We thought ARM would be interesting for a web server or home server,” explains Hormuth, Senior Distinguished Engineer at Dell. “So we started kicking some tires in late 2007, early 2008.”

It seemed counter-intuitive at the time. But Dell’s Server & Architecture group had been watching the incredible growth in data moving though the Internet and the growing cloud computing business. They conducted a study of computer infrastructure at Fortune 500 companies. “The results were surprising and shocking,” says Hormuth. “They don’t need all the horsepower being thrown at them. They were just sitting there using up watts.” The big promise, of course, is that ARM chips use much less power than Intel’s powerful Xeon line of processors – about five watts vs. several tens of watts for Intel’s traditional Xeon chips.

The folks at ARM Holdings also thought it was a promising idea. “Dell approached us and started talking about the power challenges they were facing,” says Ian Ferguson, vice-president of marketing at ARM Holdings. “On the strength of that we started to do some work,” exploring the idea that small-core processors could better handle the task of moving data through datacenters.

A couple years later the organization started hosting a website with a server run by an old, and definitely inadequate, ARM-based processor, just to see what problems would come up and what design changes might be needed to fix them.

In 2012, the ARM architecture has begun to show real promise – or at least interest from chipmakers, server manufacturers and datacenter operators. AMD, Applied Micro, Calxeda, Marvell and Qualcomm are moving into the server-class ARM chip or system-on-chip (SoC) business, and Samsung is rumored to be doing the same. Dell, Hewlett-Packard, and Penguin Computing are starting to build ARM-based servers. Facebook has professed a strong interest in low-power server chips and is keeping an eye on both ARM and coming Intel chips.

The race to build low-power processors for datacenter servers has officially begun. Intel is pitting its low-power Atom processors against an array of ARM competitors. Most server makers are likely to continue to use Intel chips as they wait for the market to sort itself out, but the future will depend on how well each player rises to the need for more efficient processors.

It’s an important race to Intel. Its processors are at the core of over 90 percent of datacenter servers today. That market is a small fraction of the PC business. Some 90 million PCs were shipped in the last three months of 2012, and Intel owns about 80 percent of that market. But that number also shrank by five percent compared to the same period in 2011. Standard PCs are being displaced by tablet computers, smart phones and other portable devices. Intel has not been very successful moving into the new consumer products market, losing out to ARM and other processors. It does not want to lose the growing server market as well.

For the last two years Intel has asserted that the microserver business that will use ARM and other low-power chips is likely to represent just 10% of the market in datacenters. Further, Raejeanne Skillern, director of the Cloud Computing Data Center Group at Intel, says that two-thirds of that market will go to “brawny” core chips, with higher compute power than the “wimpy” core chips that ARM largely represents. But a lot of people disagree with that assessment. “The microserver market is going to be more important than Intel would argue today,” says Nathan Brookwood, Research Fellow at semiconductor consulting firm Insight 64.

Datacenter servers are of huge importance today, because they run the Internet. Companies don’t just use information technology to run their own businesses, IT is their business. Servers are not just a back-office expense item, they are a big part of many companies’ product costs. Modern tech giants, including Amazon, Apple, eBay, Google, Facebook, Microsoft and Yahoo deliver their services through datacenter servers. The less expensive the server (and the datacenter,) the lower the cost of producing their products.

There are several factors that will enter into the race. Low-power is a huge issue. Another is whether there are 64-bit processors available, and how much memory the chips can handle. A third depends on which applications the servers will run and how easy it will be to port software to the different chips. One of the most critical technology issues is the on-chip “fabric” that will tie together different components of the SoC to make sure they run at peak efficiency. Chip manufacturing technology and the balance between open or proprietary products will also play a role.

Next: Power – Less is More >

 

Power: Less is More

“Among the customers that we serve, both HPC as well as large web farms, one of the common themes that comes up is that in the world of semiconductors today, it’s all about performance per watt,” says Charles Wuischpard, CEO of Penguin Computing.

Although supercomputers crunching complex floating-point instructions need powerful CPUs, they also need to lower their power consumption or their growth will be limited by the sheer cost of electricity to run them. Adding co-processors and ARM chips into the mix to handle appropriate functions can help. Penguin makes servers with X-86 processors from Intel, but last fall announced a new microserver based on an ARM SoC from Calxeda.

And demand from datacenters keeps increasing. “I’ve seen estimates that demand for cloud computing is increasing anywhere from 20% to 44% per year,” says Mike Major, vice president of corporate marketing for Applied Micro. If the datacenters don’t dramatically increase their efficiency, he says, in a few years “datacenters are going to account for the better part of seven percent of the world’s energy consumption.”

ARM chips, designed for small portable devices, were created to survive on little power, which is a big part of their appeal. “The energy efficiency story is clearly an important one,” says Drew Schulke, product marketing manager for Dell’s Data Center Solutions business.

That goes beyond just how much power the chips themselves draw. Low-power chips generate less heat, reducing the need for expensive cooling equipment. Smaller chips built into microservers can take up less space, leaving smaller areas to cool.

But competing claims are hard to decipher. Intel boasts that its 6 watt S1200 Atom chip is the lowest-power 64-bit processor on the market. That may be, but it doesn’t really match the power efficiency of today’s ARM chips. “The only way you can say the S1200 has low power is to say that it’s way less than Intel’s mainstream Xeon chips,” says Insight 64’s Brookwood.

Calxeda, for example, has circulated a table comparing its own ECX1000 SoC to Intel’s S1200. The ARM chip draws 3.8 watts, compared to 6.1 watts for the S1200. Further, it provides four cores, compared to Intel’s two, and has a cache memory of 4 gigabytes compared to Intel’s 1GB. Plus, the S1200 is not a full “server class” solution because it doesn’t include a Serial Advanced Technology Attachment (SATA) microcontroller, Ethernet, or a fabric switch or fabric ports. Adding in those features would require an extra 10 watts of power, agrees Brookwood.

Intel, however, points out that the S1200 is a 64-bit processor. “It’s almost impossible to compare,” notes Intel’s Skillern. Plus Intel continues to work on future products that will be more efficient, including the upcoming Avoton. And in the end it’s not all just about the chip. Virtualization will also decrease the energy sink of a datacenter. “We have not been sitting still in either technology or IP,” says Skillern. “We have a history of investing and innovation.”

Even Ferguson at ARM Holdings agrees with that. “Intel is a very, very competent organization,” he says. The market for datacenter servers, he adds, “is not a game they’ll just give up on and let ARM walk in.”

 

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