Rapid advances in processor technology, illustrated by the leap from single-core to dual-core and then to multicore processors, have made the task of picking a processor for your system more difficult than ever. In addition, there are different processor architectures to choose from. Understanding the differences in these technologies can help you select the right system and processor combination for your business.
The processor war between Intel and AMD is the force driving this rapid advance in technology. In the mid-to-late 1990s, it seemed that Intel had the processor race all locked up—until the company tried to push the market toward its new 64-bit Itanium processor in 2001. Intel stumbled, however, by making the Itanium binary-incompatible with x86 software, which opened the door for AMD's 2003 release of the AMD Opteron, a 64-bit, x86-compatible x64 processor. The Opteron supplanted the Itanium in the general-purpose server market and enabled AMD to gain significant server market share at Intel's expense.
Following the success of the Opteron and its Direct Connect Architecture, Intel revamped its system architecture, moving from its NetBurst microarchitecture to the higher-performing, more power-efficient Intel Core microarchitecture, and launched its own x64-compatible processors. Since then, the Itanium has been relegated to serving only very high-end, processor-intensive workloads, while x64 architecture has become the current server standard. On the surface, AMD's and Intel's multicore, x64-compatible processors might seem the same, but there are significant differences in how they operate that you should be aware of before you choose between them.
Ante Up with Dual-Core
On the heels of their move to x64 architecture, both AMD and Intel jumped into dual-core processing technologies. The ability to improve processing power simply by boosting processor speed had reached a plateau, and both chip manufacturers realized that the easiest path to more power was through parallelism. The ever-shrinking size of processors made it possible to produce dual-core chips, which combine two processors on one die. This approach had the added benefit of nearly doubling the available CPU power while using the same power envelope (i.e., the same wattage requirements) as a single processor.
In 2005, Intel became the first to enter the dual-core market with the release of the Pentium D processor, built using the Intel NetBurst microarchitecture. In January 2006, Intel switched to the Core microarchitecture. Based on Intel's Pentium M mobile processors, the Core microarchitecture uses a shorter instruction pipeline than does NetBurst, letting processors execute substantially more instructions per clock cycle and achieve higher levels of performance even though they run at a lower clock frequency than earlier Intel CPUs. AMD quickly countered Intel's move with its own 64-bit dual-core Athlon 64 X2.
The two manufacturers' dual-core designs use quite different architectures. Intel uses a shared front-side bus technology that gives each processor half the bandwidth of the frontside bus. Memory and I/O access operations also share the bus, making the bus speed a critical factor in overall system performance.
Intel's latest dual-core processor, the Core 2 Duo, is built using 65 nanometer (nm) technology. The Core 2 Duo and Core 2 Extreme CPUs integrate both cores on a single die. Each core has 64KB of dedicated L1 cache that consists of a 32KB instruction cache and 32KB data cache, and both cores share a 4MB L2 cache. The Core 2 Duo features one-cycle throughput for 128-bit floating-point Streaming Single-Instruction, Multiple-Data Extensions (SSE) instructions, which are used for complex math and graphical-display processing. It also has a new power-saving design and a front-side bus that runs at 1066MHz.
AMD uses a completely different design for its dual-core processors. In AMD's Direct Connect Architecture, each CPU has an integrated memory controller. The HyperTransport bus allows an 8GBps direct connection between the CPUs, I/O, and memory. In February 2007, AMD released new dual-core Opterons that offer higher clock speeds—up to 2.8GHz—and greater power efficiency than earlier models; the company will push its dual-core designs to 3GHz later this year.
Dual-core processors are both successful and popular. Today, most servers and many business desktops benefit from the speed and power efficiency of dual-core processors.
Double Down with Quad-Core
Following the same path that led to dual-core processors, Intel and AMD have doubled the stakes and jumped from dual-core to quadcore chip designs. With the release of the Quad-Core Intel Xeon processor 5300 series in November 2006, Intel is clearly ahead in the game. AMD won't have an entry until mid 2007, when it will enter the market with a quad-core chip code-named Barcelona. Just as with the dual-core processors, Intel and AMD offer significantly different quad-core designs.
The Intel quad-core design combines two dual-core processors onto a single chip. In other words, instead of being a "native" quad-core processor, Intel's quad-core Xeon is actually a dual dual-core chip. Although this architecture enabled Intel to get its quad-core chip to market early, the design isn't optimal. When processors that are on separate cores exchange data, the data must be sent over the front-side bus and through the memory controller, which isn't the most efficient mechanism. In addition, as with previous Intel designs, this approach makes the overall system speed dependent on the speed of the front-side bus. Despite these drawbacks, improvements in the Intel Core microarchitecture and the additional CPUs cumulate to make Intel's quad-core chips the fastest x64compatible processors available today. Figure 1 shows Intel's quad-core design. You can find more details about Intel's quad-core processors at http://www.intel.com/quad-core/index.htm?qc_tl+techresearch_promo&.
In contrast, AMD's upcoming Barcelona mounts four independent CPUs on a single die. Like earlier Opteron designs, the AMD quad-core chip will utilize AMD's Direct Connect Architecture. The Barcelona processor will be built using a 65 nm process technology and will have versions that utilize 68-, 95-, or 120-watt power envelopes. The native quad-core model enables all four cores to act independently. Theoretically, the true quadcore model also enables more efficient power consumption because each core can raise and lower its frequency according to the workload.
The Barcelona design incorporates a number of other important enhancements. It sports 128-bit floating-point processing and a new 2MB L3 cache that's shared between all processors. Because each processor performs more work per clock cycle, an estimated 15 percent efficiency improvement per core results in an improvement in processor performance of about 40 percent. One important note about the AMD quad-core design is that it's socket-compatible with existing Socket F dual-core processors. Consequently, existing dual-core systems built with the AMD Socket F can be upgraded to quad-core by performing a CPU swap and then upgrading the BIOS. The scalability of the Barcelona should also be greater than that of Intel's quad-core CPU. Each core on Barcelona's quadcore die could theoretically be upgraded to a dual-core chip in the future, essentially enabling a design that incorporates four dual-core CPUs on one quad-core die. Figure 2 shows AMD's quad-core architecture. For more information about AMD's multicore processors, refer to http://multicore.amd.com/us-en/AMDMulti-Core.aspx.
Letting the Applications Ride
So what are the biggest advantages of multicore processors? Multithreaded OSs such as Windows Vista, Windows Server 2003, and Windows XP can take advantage of multiple cores to run separate threads simultaneously. So, for example, programs such as your email or antivirus software can run on separate background threads while you perform interactive work at full speed on the foreground thread.
On the application side, games are always at the leading edge of technology, and many of today's games, such as Quake and Call of Duty, can take advantage of multiple processors. In addition, enterprise-level database servers such as Microsoft SQL Server 2005 can take full advantage of all CPUs that are present, as can virtualization software, such as VMware Server. Because these types of applications are designed for multiprocessor support, they can initiate separate threads on the individual processors of multicore systems.
All In for the Future
Intel announced its next line of multicore chips, code-named Penryn, last fall and expects to make those products available later this year. The Penryn line of processors will utilize a new 45 nm manufacturing technology, enabling Intel to increase processing speed while simultaneously reducing power requirements and heat generation. The move to 45 nm manufacturing will give Intel a temporary leg up on AMD in the game of processor leapfrog, but AMD plans to raise the stakes with its own line of 45 nm chips for 2008. Look for a second double-down move in late 2008 with Intel's rumored eight-core processor, code-named Dunnington.