AMD claims that the microarchitectural improvements in Jaguar will yield 15% higher IPC and 10% higher clock frequency, versus the previous generation Bobcat. Given the comprehensive changes to the microarchitecture, shown in Figure 7, and the additional pipeline stages, this is a plausible claim. Jaguar is a much better fit than Bobcat for SoCs, given the shift to AVX compatibility, wider data paths, and the inclusive L2 cache, which simplifies system architecture considerably.
AMD is already shipping a wide variety of Jaguar-based products including Temash (tablets), Kyoto (microservers), Kabini (notebooks, desktops). However, it is difficult to judge the performance of the Jaguar core on the basis of these products. The performance of a modern processor core in a SoC is a function of many variables: the memory controller, the fabric, the power management, etc. The existing Jaguar SoCs lack an aggressive DVFS system that is capable of boosting the frequency based on operating conditions. AMD is eminently capable of implementing these techniques, which are used in higher performance parts based on the Steamroller core; but when it comes to Jaguar, the company was clearly constrained by a lack of budget and engineering resources.
Despite this technical uncertainty, it is clear that Jaguar has been a massive financial success for AMD. For the last several years, the company has been bleeding money slowly. Intel’s Sandy Bridge and Haswell architectures have largely relegated AMD to the value portion of the market, which is threatened by inroads from ARM-based tablets vendors.
The one bright spot for AMD is the Playstation 4 and Xbox One design wins, which has catapulted the company from a sea of red ink to profitability. While these design wins were largely driven by the graphics group, it is clear that Jaguar was developed so that AMD could offer a complete SoC solution, designing out the earlier generation of IBM processors.
One of the interesting side-effects of AMD’s success with Jaguar for consoles is the future trajectory of the core. Normally, an x86 microarchitecture stays relevant for 2-3 years before it is obsoleted by derivatives – but there are reasons to believe this will not be the case for Jaguar. Since all the high-performance consoles will rely on Jaguar, game developers will inevitably optimize their code to extract the maximum performance. The lifetime of these consoles is roughly 5-7 years, which means that game developers will be producing code optimized for Jaguar for the better part of the next decade. This gives AMD tremendous incentive to keep on shrinking and optimizing the Jaguar core, with minor revisions. In theory, this means that Jaguar derivatives should eventually break the 1W barrier and become suitable for extremely low power applications, but only time will tell.
Discuss (86 comments)