You just gotta love a Cinderella story. Advanced Micro Devices is the hardscrabble kid who came to Silicon Valley with a dollar and a pack of Luckies and ended up in a building with its name on top. AMD's rapid rise from startup to US$5 billion semiconductor powerhouse is, as Humphrey Bogart's English teacher once said, the stuff of which dreams are made.
AMD went after those dreams with an unabashed moxie that would always be the company's trademark, from its first clone of Intel's 8080 to supercharged chips that enjoy a cult following among hard-core gamers. In the process, AMD has become known as the company that kept Intel honest, the Linux of the semiconductor world. Competition from AMD has reversed the trend of rising prices and stagnant innovation that characterize a controlled market. AMD is responsible for US$500 desktops, US$1,200 rack servers, and multigigahertz mainstream microprocessors, despite the fact that most of them have Intel's logo on them.
Today, AMD's pluck is paying off bigger than ever before. After decades of aping Intel architectures, the AMD64 architecture, rooted in Opteron and Athlon 64 processors, has actually been imitated by Intel in the form of Nocona, Intel's 64-bit version of Xeon. In a stunning reversal of fortune, Intel was forced to build that chip because Opteron was invading a server market that the Intel Itanium was supposed to dominate.
Suddenly, Intel is feeling a breeze where its pants used to be. But with Intel mad as hell and hot on AMD's heels, can AMD grab enough sales traction to hold up to the punishing onslaught everyone knows is coming?
Anyone shopping for servers needs to consider that question seriously because Opteron creates a new 64-bit path to follow -- one that continues the x86 tradition rather than, as Itanium does, consigning that architecture to the dustbin of history. To understand the crossroads at which AMD finds itself and what the implications are, shrewd observers must take a hard look at the company's technologies and market position now and in the past.
My friend, my enemy
AMD has had a massive impact on the market, but prior to Opteron, it showed a disappointing lack of ingenuity in its PC processors. That copycat tendency stretches all the way back to the early '70s, when AMD built the 8080A, a knockoff of Intel's 8080. Shortly thereafter, Intel and AMD struck up cross-licensing deals spanning 17 years, knighting AMD as Intel's chipmaker-in-waiting when demand exceeded supply.
Despite highly successful semiconductor product lines and plenty of engineering wins of its own, AMD came to be known as just another maker of CPUs based on Intel's blueprints. AMD was one of a pile of x86 licensees that included Cyrix, NEC, and Nexgen. There were glints of cleverness such as AMD's short-lived RISC approach to an Intel-compatible CPU. But overall, Intel held the reins on its licensees, and they were all content with that.
Then things took a horrible turn for AMD: Intel pulled the plug on its AMD agreements before they expired. Intel no longer needed or wanted a second source, and AMD was getting a little too creative with its variations on Intel's designs. AMD was forced to persevere without Intel's blessing.
With the introduction of Athlon in 1999, AMD showed real promise that it could make it on its own. Prior to Athlon, AMD's processor line was technologically suspended somewhere between the 80486 and early Pentiums. Athlon busted down the door with the intent to go straight after Pentium II. It was a huge risk. Athlon was not a drop-in replacement for any Intel processor. It required its own motherboard and chip set, which, although software-compatible with Intel's, forced manufacturers to alter their production lines and testing procedures.
That gave Intel an easy way to keep AMD in its cage. Intel went to its OEMs -- which it controlled through allocated distribution of parts -- to discourage them from making the investment in retooling for Athlon-compatible products. Rather than risk supply problems with Intel, the majority of manufacturers effectively froze AMD out. And when Microsoft got serious about its enterprise push with the release of Windows 2000 Server, Intel had Pentium II, Pentium III, and Xeon ready to rush the entry-level server space that Unix players Hewlett-Packard, IBM, and Sun Microsystems had neglected. Intel was madly grabbing IT unit share, and although AMD wanted a piece of that, there was no way to blunt Intel's advance.
Breaking the stranglehold
IT shops were grateful for the downward pressure that AMD's competition exerted on desktop PC prices, but Windows 2000 was seen as an Intel OS. At that time, it was hard enough for IT to trust Microsoft as an enterprise player, let alone risk the compatibility and stability problems that a non-Intel CPU might bring to the mix.
AMD's strategy for penetrating Intel's blockade was an odd one. Small VARs and individual system builders were strong proponents of choice and were in dire need of a stable supply of chips. The Web sites that served the enthusiast communities, which included gamers and PC technophiles, were no fans of the one-vendor market.
These small-fry operators built their systems from components, and motherboards made by Asian manufacturers were standard fare in custom-built systems. AMD couldn't convince any major motherboard makers to break loose from the Intel clan, so AMD made a motherboard of its own as a reference implementation and shipped it to every online source that might give Athlon some exposure. Soon these sites were benchmarking Pentium II and Pentium III against Athlon, and as AMD knew it would, Athlon kept up with Intel's CPUs cycle for cycle and wouldn't quit.
In Athlon, the grassroots IT community saw an alternative to Intel's tightly controlled lineup. It pressured Asian motherboard manufacturers, and a crack in the wall appeared when a lone motherboard maker stepped forward and then quickly retreated after what was assumed to be a tap on the shoulder from Intel. But it was the grassroots, low-volume builders that kept Intel motherboard makers in business. One by one, companies such as MSI, Supermicro, and Tyan bowed to system builders' demand for choice.
As motherboards began to show up, system builders, including midvolume manufacturers, were overjoyed to see faster processors, faster memory, faster bus speed, and more solid overall design than they could afford from Intel. Athlon proved consistently fast and stable. Compatibility was eliminated as an issue. And AMD earned its entry server credentials with the solid performance of AMD's dual-processor CPU, Athlon MP, showing AMD could break through the ceiling Intel tried to impose.
Hammering out partnerships
To get out of the bargain server market, AMD bet its life on Hammer, its first total system architecture that wasn't a carbon copy of Intel's. The details of AMD's Hammer processor were published well in advance of its unveiling. But Intel did not suspect that the processor -- which it planned to bury under a mountain of hype -- was the tip of a wedge. AMD built a whole-market strategy around Hammer with high-powered partnerships, inventive marketing that ignored Intel's existence, strong engagement with grassroots and commercial developers, and the brass ring: Windows.
In April 2003, AMD threw a press conference to roll out its first Hammer processor, the 64-bit Opteron. At the very instant Hammer became Opteron, Microsoft lifted it into parity with Itanium on its 64-bit Windows road map. The commitment from Microsoft to build non-Intel architecture into 64-bit client and server editions of Windows was the grail that AMD had been chasing ever since Intel tore up its contract.
Whether dealing with Microsoft or with Taiwanese motherboard makers, AMD's approach to partnerships differs from Intel's. Intel has created an insular empire, whereas AMD is building its empire through open cooperation with partners. AMD's embracing of developers has been inclusive, not selective; a Hammer emulator was available to developers as a free download long before the technology became a chip. Whereas Intel might tap a third-party fab as a second source, IBM and AMD are working together to the benefit of both partners on cooler, denser, 65-nanometer process technology at IBM's fab in East Fishkill, N.Y. AMD's second lab in Dresden, Germany, is slated to start producing samples in 2005 and to be ready for full production in 2006. And AMD still has its original fabs in Dresden and Austin, Texas.
Assured of a stable supply and finally, albeit barely, in the black again, AMD is running for daylight. In just more than a year, Opteron grew from a new dual-processor technology with one first-tier OEM to a quad-processor technology with HP, IBM, and Sun signed on. Sun is committed to a 64-bit edition of its Solaris operating system for its growing lineup of Opteron systems. AMD is on target for eight-processor Opteron and has already extended its 64-bit technology down to notebooks, desktops, and workstations. AMD has raised the performance of and grown its 32-bit, Athlon-based processor line as well, creating new options at all levels from dual-processor servers down to value home PCs. Not only has AMD stopped waiting for Intel, it has created a broader product line and is now encroaching on Intel's unit sales.
Where AMD stands with IT
Today, IT operations can safely and affordably purchase Opteron servers as upgrades to Windows Xeon servers. And as InfoWorld benchmarks have shown, 32-bit Windows applications that rely heavily on I/O and smooth scalability will get an immediate and substantial kick from Opteron. In addition, 64-bit Windows and the development tools that follow will raise 32-bit enterprise app performance and capacity up another notch, making way for 64-bit optimized server software. This is the first time IT has been offered a three-stage migration path -- 32-bit, 32-bit apps on a 64-bit OS, and pure 64-bit -- on a single architecture; none of these steps requires so much as pulling a server out of the rack.
Those companies, organizations, and technologists leery of buying on potential can wait -- Opteron will only get faster over time -- or they can climb into pure 64-bit Linux PC servers now. AMD64 support has been merged into the Linux kernel tree for some time, so most free and commercial Linux distributions, including Suse (now part of Novell), Mandrake, Gentoo, and TurboLinux already run on Opteron in 64-bit mode. Version 5.2.1 of FreeBSD is also running on Opteron. The full GNU suite of Linux development tools compile and debug Opteron code.
The toughest challenge AMD faces now is not surviving Intel's competition or ramping up supply; it's convincing businesses they want 64-bit capabilities anywhere but in high-performance servers. It's an easy sell for developers targeting Opteron, but beyond that, the benefits aren't obvious. Many will be enlightened by 64-bit Windows. On the desktop, gamers and enthusiasts will once again assume the role of convincers.
During 2005, AMD's strength and influence will grow, as new fabs and processes come on line. The 90-nanometer process that AMD already has on line will be used to boost the performance and reduce the power consumption of Opteron, Athlon 64 desktop and mobile CPUs, and the Athlon 64 FX performance desktop product lines. In the second half of 2005, AMD plans to leverage the new process to bring dual-core (two CPUs on one die) Opteron and Athlon 64 FX processors to market. With Opteron's direct CPU-to-CPU interconnects, servers built with dual-core AMD processors can house 16 CPUs in one chassis with no logic -- or associated overhead -- to slow traffic between processors.
As the one company delivering stable, affordable processors for everything from home computers to eight-way servers, AMD looks like a modern-day, more enlightened Intel. Intel has remade itself more than once and won't allow itself to stay, if only in perception, a step behind AMD. But AMD won't ever kowtow to Intel again. And IT will never again be forced to submit to the price and technological power that Intel unilaterally decides the market needs.
Opteron vs. Nocona: It's the system, stupid
If you think AMD's Opteron and Intel's Nocona -- or more formally, "Xeon Processor with 800MHz System Bus" -- are cut from the same 64-bit cloth, look closer. Yes, they're compatible at the instruction-set and register levels; they should be because they're both based on AMD's x86-64 specification. But the total system architecture surrounding these chips -- which includes pathways to other CPUs, memory, and peripherals -- exhibits several differences that factor into buying decisions and developers' platform targeting.
At its core, Nocona is a NetBurst Xeon DP, a Pentium 4 equipped for dual-processor operation. It has 1MB of Level 2 cache and a top clock speed of 3.6GHz. All memory and I/O data, interrupts, interprocessor communication, and address requests flow over a fast shared bus with a maximum bandwidth of 6.4GBps. It's a highly evolved design, on the leading edge while remaining faithful to the legacy design principles that Intel is expected to maintain.
The 64-bit technology common to both processors is easy to explain: more memory and more registers. When you're running a 64-bit OS, standard PC caps on physical and virtual memory go away. (Well almost: Opteron has a larger total address space than Nocona, but Nocona can accommodate twice as much physical memory as a current dual-CPU Opteron system: 32GB vs. 16GB.) Registers are the fastest type of storage a CPU has. The more registers you have and the more bits each register holds, the more compilers can optimize application performance. Having more registers and using them well can also improve the speed and smoothness of task switching, which has an effect similar to that of Intel's Hyper-Threading technology.
Beyond the instruction set and address space, however, these two processors have nothing in common. And where they diverge most is in their total system architectures.
As with all Xeons, Nocona's shared bus is the Achilles' heel of Intel's architecture. That only gets worse with SMP systems in which multiple CPUs must funnel their data, I/O, addresses, memory access, and interprocessor communication through a single bus and compete for access to a single pool of memory.
There are two ways to improve a shared bus: Make it faster or divide it up into independent buses. Intel sped things up, raising the bus speed from 533MHz to 800MHz, and tossed out a hint that it's going after the independent bus design. The new touch is PCI Express. The chip that directs traffic on Xeon's shared bus now has three onboard serial communications channels, each of which has a theoretical maximum throughput of 4GBps. Nocona can't touch the channels' aggregate potential of 12GBps with a 6.4GBps shared bus, but faster buses will inch it closer to that limit.
By contrast, Opteron implements as many as four independent high-speed buses on each processor, depending on the model of the CPU. One bus on each processor is dedicated to memory traffic, with a maximum bandwidth of 6.4GBps. The Opteron architecture gives each processor its own bank of memory, so theoretically, bandwidth rises and contention decreases as more processors are added to a server.
Communication with nonmemory system components, including other processors and peripherals, is handled by HyperTransport bus controllers built into the CPU. This parallel bus, developed by AMD and licensed by others, including Apple and Transmeta, has a bandwidth of 6.4GBps (3.2GBps each way), for a total potential system bandwidth of 19.2GBps, independent of memory traffic. Direct HyperTransport links between CPUs allow all processors to share all the system's memory, split though it is across processors, at full speed.
The bottom line is that the Opteron architecture with HyperTransport set the stage for blazing multiprocessing performance. And at this stage, Intel's Xeon line has nothing to match that.
How AMD stayed in the game
In 1999, while AMD was suffering through one of the darkest periods in its history, the financially strapped semiconductor maker needed to get the word out about its new Pentium II-compatible processor, Athlon. So it did what any serious company would do: It enlisted the aid of PC gamers, overclockers, and build-it-yourself enthusiasts.
AMD established an inclusive policy of supplying Web publications with processors and systems. It reached out to selected small and startup sites that were snubbed by other hardware vendors. For many of these sites, it was AMD's willingness to provide review hardware, no strings attached, that encouraged Intel, Nvidia, and other vendors to join the party.
Web publishing pioneers Anand Lal Shimpi and Thomas Pabst were already one-upping print publications with timely and painstakingly thorough benchmarks and reviews. There was a fair amount of religious fervor on the Web sites -- those with large followings and those just starting out, where hope for AMD's success colored test results and analysis. But over time, especially when Intel started matching AMD's level of cooperation with hardware sites, balance was restored, and races were usually called fairly.
Being friendly with these sites wasn't always a picnic for Intel or AMD. Enthusiast sites revealed the significant design trade-offs Intel accepted to raise Pentium 4's clock speed. They described how to turbocharge Intel's dirt-cheap Celeron processor and how to unlock its concealed capability of operating in a dual-processor configuration. AMD took its share of hits as well. It got badly dinged for its persistent heat problems -- a vendor called KryoTech sold an Athlon system that was cooled by the guts of a refrigerator -- and AMD couldn't keep photos and write-ups of its unreleased CPUs off the Web.
Now all the respected hardware sites have ads, staffs, blogs, and affiliate programs. And they never miss a trade show no matter what country it's in. AMD still works closely with Web publications on 32-bit Athlon XP and desktop Athlon 64 and Athlon 64 FX-series processors. The company will always sell the processors that, with other vendors' motherboards, power custom-built systems that outperform the fastest desktop in Dell's catalog and cost what Dell charges for last year's model.
Hewlett-Packard, IBM, and Sun Microsystems give AMD the clout it needs to sell Opterons, so AMD's top-end server chips rarely get reviewed on hardware sites as components for enthusiasts. It's the first time AMD has dialed down its relationships with hardware sites. But those sites may enter a new era, as IT and professional technologists look for fresh approaches to evaluating servers. If that happens, AMD will be there.