AMD first started dishing out details on Carrizo, the successor to Kaveri, during the closing months of 2014. AMD is claiming that Carrizo, which is still built on Global Foundries’ 28nm Super High Performance (28SHP) process node like its predecessor, will nonetheless deliver huge advances in both performance and efficiency.
When it was first announced, AMD detailed support for next generation Radeon Graphics (DX12, Mantle, and Dual Graphics support), H.265 decoding, full HSA 1.0 support, and ARM Trustzone compatibility. But perhaps one of the biggest advantages of Carrizo is the fact that the APU and Southbridge are now incorporated into the same die; not two separates dies built into the same package. This not only improves performance, but also allows the Southbridge to take advantage of the 28SHP process rather than older, more power-hungry 45nm or 65nm process nodes.
In addition, the Excavator cores used in Carrizo have switched from a High Performance Library (HPL) to a High Density Library (HDL) design. As you can see in the diagram below (left half) this allows for a dramatic reduction in the die area taken up by the processing cores (23 percent reduction according to AMD).
This allows for Carrizo to fit in 29 percent more transistors (3.1 billion versus 2.3 billion in Kaveri) in a die size that is only marginally larger (250mm2 for Carrizo versus 245mm2 for Kaveri). When all is said and done, AMD claims a 5 percent IPC boost (this comes mainly from doubling L1 data cache from 64KB to 128KB) and a 40 percent overall reduction in power usage.
AMD has also worked its magic on the graphics core, achieving an 18 percent reduction in leakage that allows for a 20 percent reduction in power usage at the same clock frequency as the previous generation graphics core. AMD can also opt for a 10 percent clock frequency boost at the same power levels.
Further efficiencies were found by optimizing voltage levels to prevent “droop”, also known as a transient drop in voltage. Since AMD is now able to compensate for voltage droops within nanoseconds, it can do so without [power] costly over-voltage measures, resulting in a 10 percent and 19 percent reduction in GPU and CPU power consumption respectively.
Further power saving measures have been taken in the form of Adaptive Voltage and Frequency Scaling (AVFS). AVFS uses a combination of silicon speed, voltage, power, and temperature sensors to “enable each individual APU to adapt to its particular silicon characteristics, platform behavior, and operating environment.” AMD says that AVFS can contribute to an impressive 30 percent power savings.
“As a part of our continued focus on building great products, the advanced power and performance optimizations we have designed into our upcoming ‘Carrizo’ APU will deliver the largest generational performance-per-watt gain ever for a mainstream AMD APU,” said AMD Corporate Fellow Sam Naffziger. “AMD has been pursuing Heterogeneous System Architecture (HSA) and proprietary power management technologies to make continued gains. The upcoming ’Carrizo’ APU takes a big step toward the AMD 25x20 energy efficiency goal and incorporates a wealth of new features that will be adopted across our full product line going forward.”
AMD says that its Carrizo APUs will make their way into notebooks and low-power desktop computers during the first half of 2015. AMD is targeting "double digit" increases in both performance and battery life for notebooks, so we're definitely looking forward to see this APUs cross our test benches.