Oracle kills AMD Opteron on Sun iron

Sun volume server master plan gutted


Oracle is abandoning AMD's Opteron processors, according to a person familiar with the company's server plans.

The new owner of Sun Microsystems will not use the new Opteron 6100 and impending Opteron 4100 processors in future Sun Fire x64 servers, and all existing Opteron servers will be discontinued.

Thus ends an Opteron era that began with an impressive bang back in 2003 when IBM and then Sun enthusiastically embraced the Opteron because — at the time — Intel's 32-bit, frontside-bussed Xeons were just plain awful.

Oracle laid out a rough sketch of its server plans back in January, after closing the Sun Microsystems acquisition. But additional details have be scant.

In January, Charles Phillips, Oracle's co-president, said the company is not interested in being in the commodity x64 server racket. John Fowler — executive vice president of hardware engineering at Oracle who used to run the Systems Group at Sun — said that Oracle would focus x64 servers on enterprise-class products used in clusters. In other words — though Oracle never came right out and said this — the company was not going to have the plethora of different rack and blade machines in various shapes and sizes, but would streamline its products. This is understandable considering that Sun's volume approach had generated some revenues but probably lost Sun a fair amount of money.

Oracle has not been precise about exactly what it would do to streamline its x64 server lineup, but circumstantial evidence indicates that Oracle isn't exactly an enthusiastic x64 supporter. Sun was the most gung-ho of the Opteron server makers, but Oracle was quiet as a dead mouse when the twelve-core "Magny-Cours" Opteron 6100s and their related chipsets were announced by Advanced Micro Devices at the end of March. And it was equally quiet when Intel announced its six-core "Westmere-EP" Xeon 5600 chips two weeks earlier.

On March 30, at the eight-core "Nehalem-EX" Xeon 7500 launch in San Francisco, Oracle showed off an unnamed box presumably supporting the Xeon 7500s. As The Register previously reported, Oracle has been tweaking Solaris 10 to run well on eight-socket Xeon machines (which can only be built with Xeon 7500-class processors). Oracle did sneak out a Sun Fire X2270 M2 blade server sporting the new Xeon 5600s, and Intel's presentations about who is doing what with the Xeon 7500s show that Oracle is working on an eight-socket box.

The lack of Opteron machines was telling. And a source familiar with Oracle's plans tells The Reg that all Opteron servers will be end-of-lifed.

Oracle did not respond to repeated requests for comment, saying it is still in its quiet period ahead of reporting its next financial results. And AMD referred us back to Oracle. "With respect to your question on Oracle's future plans for hardware, specific details about any server vendor's overall strategy are best addressed by that vendor," AMD told us. "In the server space, we don't believe that a credible hardware vendor looking to satisfy business customers' needs today can do so without a leading industry-standard x86 solution. The datacenter is in transition now. Customers are facing a delayed refresh cycle and they are looking for maximum value."

But Oracle did discuss its x64 plans with server analysts at a recent event in Australia. A May 5 blog post from Ideas International analyst Gary Burgess discussed a speech from Shane Sigler, Oracle's director of strategic engagement for systems, that he delivered as part of an Oracle's "Welcome Event" for Sun customers:

The biggest news for Ideas at the event was Mr Sigler's announcement that Oracle intends to go forward with just a single x86 processor architecture. The company will bring to market new Sun x86 server using the Intel processor architecture, and has no plans to develop any new servers with the AMD processors, including Magny-Cours. In fact, Mr. Sigler said that the company is in the process of EOL'ing the current family of AMD x86 servers.

That seems like a pretty clear statement. But when we asked for Fowler or Sigler or anyone else at Oracle to get on the phone and chat about this, the company did not respond.

And if you visit Burgess's blog post today, it reads a bit differently:

It appears that the company may focus on Sun x86 servers using only the Intel processor architecture. If true, (and there has been no formal announcement yet) this would be a big turnaround from the birth of Sun's "Galaxy" x86 offerings, which at that time ran exclusively on AMD Opteron processors. IDEAS speculates that is probably a reflection of how Intel has been able to turn things around of late, especially with its latest generation Nehalem family of server processors it has brought to market in phases over the past 12 months.

Both the past and current version of the blog written by Burgess says that Sigler told attendees at the Sun Welcome Event in Sydney that the Fire X4100 and X4200 machines are integral to the Exadata OLTP/data warehousing appliances and that the X2200 blades are key to Sun's HPC supercomputers and will see continued development.

You can't blame Oracle for wanting to simplify the Sun server lineup and to forge a tight partnership with Intel. The chip maker has come a long way in seven years with its Xeon lineup, and Oracle wants to make money more than it wants to make boxes with the hope of maybe making money.

But Oracle has a lot to learn about being clear to customers and channel partners about what its roadmaps for both x64 and Sparc servers are, which should be a source of comfort, not concern. It's doing a worse job than Sun in this regard, and that is not a good thing.

A few pesky journalists and Wall Street analysts asking questions and being uncertain is one thing. Thousands of Sun customers not being able to plan their server acquisitions is quite another. Maybe Oracle will be a little more forthcoming when it announces some new Xeon servers in June, as it is rumored to be doing. ®


Other stories you might like

  • Talos names eight deadly sins in widely used industrial software
    Entire swaths of gear relies on vulnerability-laden Open Automation Software (OAS)

    A researcher at Cisco's Talos threat intelligence team found eight vulnerabilities in the Open Automation Software (OAS) platform that, if exploited, could enable a bad actor to access a device and run code on a targeted system.

    The OAS platform is widely used by a range of industrial enterprises, essentially facilitating the transfer of data within an IT environment between hardware and software and playing a central role in organizations' industrial Internet of Things (IIoT) efforts. It touches a range of devices, including PLCs and OPCs and IoT devices, as well as custom applications and APIs, databases and edge systems.

    Companies like Volvo, General Dynamics, JBT Aerotech and wind-turbine maker AES are among the users of the OAS platform.

    Continue reading
  • Despite global uncertainty, $500m hit doesn't rattle Nvidia execs
    CEO acknowledges impact of war, pandemic but says fundamentals ‘are really good’

    Nvidia is expecting a $500 million hit to its global datacenter and consumer business in the second quarter due to COVID lockdowns in China and Russia's invasion of Ukraine. Despite those and other macroeconomic concerns, executives are still optimistic about future prospects.

    "The full impact and duration of the war in Ukraine and COVID lockdowns in China is difficult to predict. However, the impact of our technology and our market opportunities remain unchanged," said Jensen Huang, Nvidia's CEO and co-founder, during the company's first-quarter earnings call.

    Those two statements might sound a little contradictory, including to some investors, particularly following the stock selloff yesterday after concerns over Russia and China prompted Nvidia to issue lower-than-expected guidance for second-quarter revenue.

    Continue reading
  • Another AI supercomputer from HPE: Champollion lands in France
    That's the second in a week following similar system in Munich also aimed at researchers

    HPE is lifting the lid on a new AI supercomputer – the second this week – aimed at building and training larger machine learning models to underpin research.

    Based at HPE's Center of Excellence in Grenoble, France, the new supercomputer is to be named Champollion after the French scholar who made advances in deciphering Egyptian hieroglyphs in the 19th century. It was built in partnership with Nvidia using AMD-based Apollo computer nodes fitted with Nvidia's A100 GPUs.

    Champollion brings together HPC and purpose-built AI technologies to train machine learning models at scale and unlock results faster, HPE said. HPE already provides HPC and AI resources from its Grenoble facilities for customers, and the broader research community to access, and said it plans to provide access to Champollion for scientists and engineers globally to accelerate testing of their AI models and research.

    Continue reading

Biting the hand that feeds IT © 1998–2022