Nvidia Rebrands Tegra 3 Architecture into "4-Plus-1"

Nvidia Emphasizes Heterogeneous Nature of Tegra 3 SoCs

by Anton Shilov
02/22/2012 | 11:59 PM

Nvidia Corp. has decided to underline advantages that its latest Tegra 3 system-on-chip has over competitors and renamed the term "variable symmetric multiprocessing" it introduced last year into 4-Plus-1. The new name should tell users that Tegra 3 is not just a quad-core, but a five-core chip with one core being an ultra low-power one. The question is: do customers care about the number of processing cores inside tablets or smartphones?

 

"We have described Tegra 3’s unique architecture in a number of ways since its launch back in November 2011 – as variable symmetric multiprocessing, as having a “companion core” or as wielding a 'ninja core'. Today, we are officially calling it the 4-Plus-1 quad-core architecture and trademarking the name," said Mike Rayfield, the head of Tegra business unit at Nvidia, in a blog post.

Nvidia explains that as Tegra 3 gains adoption among manufacturers of smartphones and tablets, its customers want a flashy for Nvidia's  variable symmetric multiprocessing (VSMP) technology that could describe advantages of the SoC in a simple way. For Nvidia, it costs nothing to call the tech as 4-Plus-1, but the benefits the trademark will provide are uncertain.

The vast majority of customers looking for Nvidia-based gadgets are looking for premium gaming experience, which is something Nvidia is working very hard on. The better Nvidia Tegra SoC become and the more game developers take advantage of its peculiarities, the better that experience will be and more consumers will look forward Nvidia brand on consumer electronics.

Furthermore, as ARM is bringing in its Big.Little heterogeneous multi-core technology on the market, all the mobile devices will eventually have different types of processing engines inside their SoCs and Nvidia's 4-Plus-1 (or 6-Plus-1) will barely show any actual advantages that Nvidia may have by the time.

Nvidia Tegra 3 "Kal-El" system-on-chip (SoC) is based on four high-performance Cortex-A9 general-purpose cores with Neon accelerators (1.4GHz single-core or 1.30GHz quad-core mode) and one power-optimized Cortex-A9 core (up to 500MHz), features a GeForce graphics processing unit with twelve stream processors and will have a new display and video engines capable of supporting Blu-ray disc video playback and stereo-3D graphics output. The new device integrates single-channel DDR3L-1500/LPDDR2-1066 memory controller, e.MMC 4.41, SD 3.0, SATA II, 7.1-channel audio controller, faster image signal processor and so on. According to Nvidia's own performance benchmarks, Tegra 3 is 50% - 200%+ faster than Tegra 2 in various applications.

The Tegra 3 processor implements Nvidia's variable symmetric multiprocessing (vSMP) technology, which adds fifth "companion" core to high-speed Cortex-A9. The companion core is architecturally similar to other general-purpose cores, but it is made using LP (low-power) process technology (and thus can operate at significantly lowered voltages) and operates at clock-speeds in the range between 0MHz and 500MHz, whereas other cores are capable of working at over 1GHz frequencies.

During less power-hungry tasks like web reading, music playback and video playback, Kal-El completely powers down its four performance-tuned cores and instead uses its fifth companion core. For higher performance tasks, Kal-El disables its companion core and turns on its four performance cores, one at a time, as the work load increases. Patented hardware and software CPU management logic continuously monitors the SoC workload to automatically and dynamically enable and disable the companion core and the main general-purpose performance-optimized ARM cores.