At Intel’s Architecture Day in 2018, the company talked about design changes it’s making to future products. It is said going forward, Intel will design CPUs in a way that allows it to be used flexibly, rather than strictly attaching a given CPU design to a process node. Although the company did not say this publicly, the idea that Intel would take a 10 nm processor and send it back to 14 nm has already drifted as a solution to its misery.
During an AMA for the Rocket Lake platform, Intel revealed that the CPU backport project created by Cypress Cove was completed in the first quarter of 2019. The reason why Intel targeted Ice Lake’s Ice Cove processor, as opposed to the more advanced Willow Cove chip within Tiger Lake, is that TGL design has not yet been finalized. Rocket Lake will ship in late Q1 2021, indicating that Intel took about two years to build the new floor plan and replace the core.
This timeline provides a window into how slides are designed. In the first quarter of 2019, Intel’s top-end chip was the 9900K and the company still had its own solid stack of performance gains at the top of the CPU market, especially in the game. Intel has started working on Greenpress Cypress Cove to support Ice Lake at 14 nm, while the Core i9-9900KS (launched in October 2019) and the 10th Gen Core i9-10900K (starting 2020).
This is Intel’s roadmap from its previous architecture day in 2020. Rocket Lake is not even shown in this document, which focuses on advances in Intel core technology.
This is how semiconductor manufacturing tends to work. AMD is currently releasing Zen 3, finalizing Zen 4 and working on Zen 5. Intel has just launched Rocket Lake with Alder Lake in 2021 and is said to have a 7 nm successor intended for 2023. This means that Intel and AMD both design features and select targets. based on what they think the competitive situation will be 1-2 years later.
The AMA states that an eight-core die represents the largest D + UHD graphics that Intel has been able to produce without explaining whether it is related to the design of the LGA1200 socket or any other product limitation. The years since Ryzen’s launch have, we think, illustrated factual differences in how AMD approaches product design versus Intel.
While not every AM4 motherboard supports every Ryzen CPU, AMD has shown that it designs its platforms for a longer lifecycle and better upgrade paths than Intel does, despite the huge differences in their respective revenue and resources. If AMD were able to design AM4 to support both a switch from monolithic mattresses to chipsets with a central I / O die and a doubling of the CPU core counts, then there is no reason for Intel to continually discover that its products are so easily out of the headroom – unless intended. We still suspect that the limiting factors on Rocket Lake were thermal and power related. Intel is absolutely capable of producing larger chips than an eight-core RKL + integrated Xe GPU. To claim that it can not, without reference to what is the limiting factor, sends an erroneous message about Intel’s manufacturing capability, unless you think that the same company that succeeds in building a 28-core monolithic Xeon does not ‘ a 10-core can not manage desktop CPU with an IGP.
Pat Gelsinger has promised to bring Alder Lake to the desk only before the end of 2021. If Intel sticks to this schedule, it will be the fastest phase-out phase-out in history. Rocket Lake compares better at the bottom of the stack, but the top slide is competitively fast and non-competitively thirsty in terms of power.
Part of what we see here is an interesting example of how a company can make one set of reasonable decisions in 2019 and two years later it is difficult to find the final product. But the fact that Alder Lake is coming to Rocket Lake so fast also speaks to the size and strength of Intel. Intel has maintained multiple engineering teams working on multiple chips so that they can (perhaps) launch a new computer replacement CPU less than a year after the launch of the first one. This kind of turnaround represents its own assurance of a poor launch, and it’s not something we see happening regularly.
The closest analogy I can think of is in June and August 2002. In June, AMD introduced a 130 nm full-blooded Athlon XP, the so-called ‘T-wide A.’ The chip got hot and AMD did not get much of the clock. Two months later, AMD unveiled Thoroughbred B, a new twist of the chip with a 9th metal layer and significantly higher frequencies. On top of that, T-Broad A processors were 1.8 GHz, but full-blooded B was able to reach 2.25 GHz, which put AMD on a stronger footing against Intel, which was on its way in the middle of the year. Anyway, Rocket Lake occupies a strange place in Intel’s pantheon of desktop CPUs, and perhaps a bit of an object reading in the trouble of adapting a CPU design to a process node not built for.
Read now: