What does history say about the last time AMD defeated Intel?

Looking at how quickly retail shelves are emptied of AMD Ryzen processors, it is not difficult to see how popular AMD desktop processors are among consumers. With outstanding single-core performance, along with options that come with 6, 8, 12, or even 16 cores, the Raizen 5000 series processors have gone one step further than their Intel counterparts in terms of pure performance. All this makes us want to look at history and see when we have seen such conditions in the market.

AMD has already outperformed Intel products, but previous victories against this chip giant have been rare. In addition, each time Intel seemed weaker than its rival, it responded quickly and effectively.

Last time the story

If you want to see the last time AMD was able to completely surpass Intel in terms of pure performance, you need to ride with us in the time machine and go back to 2005. At the time, the dual-core Athlon 64 X2 processor had won critical acclaim and basically made the Intel Pentium 4 and Pentium D series look quite old.

The AMD Athlon 64 X2 processor was another new and equally acclaimed version of the processor called the Athlon 64 that defeated Intel in the 64-bit world, but was eventually overshadowed by the Pentium 4 Extreme Edition. The Pentium needed skyrocketing speed to get attention from AMD, and it seemed like a trick compared to AMD’s innovations. With this 64-bit dual-core processor, Intel went around for almost a year before finally being able to build and market competitive products.

Implement the map
The Netburst architecture used by Intel in the Pentium 4 was obsolete, and even the implementation of dual-core technology within these products did not allow Intel to achieve the same performance as AMD. The time for innovation had just arrived, and this mindset led to the development of Core architecture.

To achieve the Core architecture, Intel revised the Pentium III and Tualatin. Improvements to the design of the Pentium 3 led to the invention of the Pentium M, a mobile product that was both fast and efficient, with features such as SpeedStep for variable voltage and clock speeds that increased battery life.

This laptop-optimized chip was able to play a key role in Intel’s new strategy in 2006 when Intel needed something to impress fans of the AMD Athlon 64 X2. The Intel Core architecture also hit the market in the middle of the same year with the Core 2 Duo lineup, and Core 2 Extreme models were able to establish Intel’s superiority over AMD. Then came the quad-core products, again in the same year, and Intel products were clearly better, faster, and even more affordable than AMD could offer customers.

Core improvements

How did Intel do that? At the same time, the Core 2 Duo used a smaller 65-nanometer manufacturing process. AMD was making its own chips to compare it to the 90nm process. The Intel product also had more instructions per bit, slightly higher bass and clock speeds, and more L2 cache, and operated at a lower thermal voltage and power.

All of these capabilities directly led to improved performance. For example, these chips achieved higher optimization by pairing instructions during implementation, and this was a feature that Intel chose the name Macro-Fusion. In addition, the two processor cores shared L2 cache, and no specific amount of cache was allocated per core. Eventually, all the lessons Intel had learned in power management for Pentium M processors were added to the equation, and more optimization was achieved in processors based on the Core architecture.

AMD stumble and Intel’s long-term thinking

Entering the 45-nanometer manufacturing process, Intel continued to put more pressure on AMD, achieving lower power consumption and higher clock speeds. The move was part of Intel’s Ticketack production model. All changes made to the microarchitecture were considered “ticks”. Then we saw a reduction in the dimensions of the die in the production process, which played the role of “vine”. As Intel relentlessly strived to excel in performance racing, AMD made a number of important business decisions that had a long-term impact on the company’s operations.

In late 2006, AMD bought ATI, a graphics card maker, for a staggering $ 5.4 billion. Their next desktop processor could not compete with the Intel Core 2 Quad processor, and subsequent performance hiccups further damaged the company’s reputation. Meanwhile, Intel was experiencing a glorious comeback.

A few years later, a serious financial crisis gripped AMD, putting it in an even more difficult position. Finally, with Graphics Core Next in 2012, work on the Wii U and PlayStation 4 and Xbox One consoles in 2013, and the introduction of the Raizen series processors in 2017, there was evidence of a revival of AMD.

Change what answered

TickTock’s strategy worked a miracle for Intel, and led to the consistent release of a series of admirable processors that had a terrible performance difference with AMD products for about 10 years. When 2016 came, Intel changed its production model, moved to the “processing-architecture-optimization” approach, and essentially added another step to its strategy. Intel said the main reason for the decision was its greater financial feasibility and costly production of smaller nannies.

However, the acquisition of the 7-nanometer manufacturing process played a key role in AMD re-dominating Intel, leading to the formation of torque that brought them to their current successful state.

What can we expect from the future?

As Intel lags behind AMD in recent benchmarks, its next step will be crucial. So far we know very little about the new Rocket Lake processors. These processors will benefit from a completely new architecture and thus end the life of Skylake models.

These Rocket Lake processors seem to be making relative improvements, and some speculate that we will see a 10 to 18 percent improvement in performance compared to Intel’s current-generation processors. Other exciting improvements include PCIe 4.0 support, faster memory support, and integrated Intel Xe graphics. The twelfth-generation Alder Lake, or Intel Core, is expected to launch after Rocket Lake and be more like what Intel was doing before.

The Alder Lake will not be available until late 2021, but it will be manufactured with a 10-nanometer process, so Intel will have a better chance of competing with AMD in terms of power consumption and optimization. In these processors, we have to wait for more cores and support for DDR5 memory, which, if everything goes according to plan, will lead to the release of truly admirable products.

Given what has been said in history, it seems that Intel can regain its position as the king of CPU performance, but consumers will have to be patient until the release of Rocket Lake and then Alder Lake. Of course, we should not underestimate AMD (or Apple) in the meantime, which definitely have their own surprises in Austin and try to prevent Intel from growing again. If this old rivalry has shown us only one thing, it is that Intel and AMD will deal fatal blows to each other in the coming years.