AMD Weighs In on Nehalem

By Michael Feldman

April 1, 2009

Intel’s launch of its Nehalem server chips (Xeon 5500) into the market represents not only a new direction for the company’s Xeon processor line, but also a new phase of competition with arch-rival AMD. For the first time, the competing Intel Xeon and AMD Opteron product lines will be based on essentially the same fundamental architecture — a multicore design with on-chip memory controllers and a point-to-point processor interconnect.

Although AMD has used this basic design in its silicon for years, with Nehalem, Intel has in many ways improved upon the model. Based on current benchmarks, most analysts (and even AMD) would agree that the new 2P quad-core Nehalems outperform their AMD “Shanghai” counterparts. (For now, Shanghai still rules in the 4P and 8P space since Intel has yet to release corresponding Nehalem parts.) The new Intel products have accomplished this with a more aggressive QuickPath Interconnect implementation, DDR3 memory, a larger L3 cache, and a hyper-threading capability that can simultaneously execute two threads per core.

I spoke with John Fruehe, AMD’s worldwide business development manager, to get his take on how Nehalem will challenge AMD’s Opteron product line and what AMD has in store over the next 12 months. In general, Fruehe maintains his company is sticking with its game plan that focuses on energy efficiency, price, and platform stability. From his perspective, this strategy plays directly into Nehalem’s weaknesses and will be especially important in an economic climate in which cash-strapped customers are looking for ways to reduce capital expenditures.

In the next few weeks AMD is planning to announce some new higher performing Opterons in the low-power (55W) HE product line. Fruehe says they will also introduce some even lower-power server chips than the HE parts. All of these products are geared toward customers looking to minimize their energy footprint, which these days is basically everyone.

In the second half of 2009, AMD is scheduled to launch “Istanbul,” its six-core version of Shanghai. Istanbul ostensibly competes against Intel’s six-core Dunnington chip released last year, but with the Nehalems now running loose, AMD will be likely be positioning its six-shooter against Intel’s latest and greatest. Istanbul will also include “HT Assist,” a probe filtering technology that improves memory throughput by optimizing cache access. For HPC applications, especially, the improved six-core performance should keep AMD in the game for the remainder of 2009.

“But the real product that will compete against Nehalem is when we get into next year with the G34 platform,” says Fruehe. According to him, that platform will outgun the Xeon competition. The first processors to land on the G34 platform will be the six-core Sao Paolo and 12-core Magny-Cours. Like Nehalem, AMD’s next generation platform will support DDR3 memory. Fruehe pointed out the G34 design supports four-channel DDR3 compared to Nehalem’s three-channel setup. G34 will also include four HyperTransport 3.0 links to help cope with the increasing I/O traffic resulting from more cores.

On paper, G34 looks like it competes well against the Xeon 5500 design, but by the time it’s launched in 2010, Intel may have already moved on to “Westmere,” the 32nm process shrink of Nehalem. Because AMD lags Intel by 6 to 12 months on a number of technology fronts, it will have trouble playing leapfrog with its larger rival. But being on the bleeding edge can cut both ways and Fruehe was quick to point that out.

For example, forcing your customers to do a forklift upgrade every couple of years because you’ve switched chipsets may be a way to eke out the best performance, but it becomes an expensive proposition for customers. Fruehe noted that AMD has been much more focused on pricing and platform longevity than its rival. For example, the Socket F design introduced in 2006 provided the platform for multiple generations of Opterons, from the dual-core and quad-core chips through the upcoming six-core Istanbul.

Maintaining platform compatibility over a multi-year timeframe means customers can upgrade systems relatively inexpensively by plugging in new chips and upgrading the BIOS. Being able slip the latest Opterons into existing servers has been an especially useful way to extend the lives of multi-million dollar supercomputers. For example, last July TACC replaced all 15,577 quad-core Opterons on its half-petaflop “Ranger” cluster with a slightly faster processor, effectively adding 75.4 teraflops to the system’s peak performance rating.

“It’s pretty clear that there’s plenty of innovation that can happen within a platform without having to change everything,” says Fruehe. “We don’t buy into the argument that you need to rip everything up or push your customers through a lot of churn just to get an advantage over your competitor.”

Fruehe also believes Intel’s early bet on DDR3 memory is going to work against them. From his perspective, although DDR3 offers higher capacities in terms of total addressable memory, the hardware costs more, draws more power and has higher latency than DDR2. Fruehe thinks Intel miscalculated where the DRAM industry was going to be in 2009 in regard to DDR3, mainly because of the unforseen consequences of the economic downturn on the major Taiwanese memory manufacturers.

“[Intel] put their money on red and the Roulette wheel came up black on this one,” says Fruehe. According to him, the shortcomings of DDR3 should be addressed over the next year as the market matures. Fruehe admits that DDR3 memory will be “awesome,” but not until 2010, when, coincidentally, AMD jumps onto the technology.

Of course, the major architectural changes implemented in Nehalem didn’t leave Intel much room for backward compatibility. But the fact that it now has to sell such chips into an economic headwind is going to make that transition harder. Certainly, customers are going to have to balance gains in performance or energy efficiency against the cost of new Nehalem-based gear.

“We did a little math on one of the OEM’s Web sites yesterday, and the lowest-priced Nehalem configuration is higher than the most expensive Shanghai configuration,” notes Fruehe. “So there’s obviously a price premium the customer is going to pay. Some of that comes from the DDR3 memory and some of that is just a price premium on the platform itself.”

He estimates that customers that buy Nehalems this year will pay an extra 50 percent or more compared to similarly configured systems. Intel’s marketing juggernaut may be more than enough to overcome any such sticker shock, especially if it can make a case for a short-term payback on new equipment. For now, AMD seems content to make its big platform shift next year and hope that 2010 brings a return to prosperity in the IT sector.

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