MediaTek's 10-core Helio X30 now official, debuting next quarter

At Mobile World Congress, MediaTek has officially kicked off availability of its high-end SoC for 2017, the Helio X30. The chipset is built on a 10nm node, and will be making its debut in phones next quarter.

Like its predecessor, the Helio X30 offers a tri-cluster CPU design that sees two Cortex A73 cores clocked at 2.5GHz to form the high-performance cluster, complemented by four Cortex A53 cores at 2.2GHz and four Cortex A53 at 1.9GHz. MediaTek is touting its intelligent task switcher, CorePilot 4.0, to deliver up to 25% power savings. CorePilot 4.0 pulls data from the device's thermals and relies on "user experience monitoring technology" to predict which CPU cores should be used for a particular task. The idea is to maximize battery life while ensuring that users don't notice a drop in performance.

On the GPU front, the Helio X30 fields Imagination's PowerVR Series7XT Plus clocked at 800MHz, which according to MediaTek is 2.4 times faster than last year's Helio X20 while consuming up to 60% less power. The SoC accommodates displays up to 2560x1600 in resolution, up to 8GB of LPDDR4X RAM, and eMMC 5.1 or UFS 2.1 flash memory.

Other improvements include a vision processing unit for processing images that uses 10% of the power a regular CPU would for undertaking the same task. The 14-bit Imagiq 2.0 ISP supports 28MP sensors, dual 16MP+16MP sensors, improved EIS for smooth videos and ultra-fast autofocus.

The SoC is also the first to include a hardware-based 4K 10-bit HDR10 decode for 4K video playback at 30fps. Also included is a Category 10 LTE modem with 3x carrier aggregation that has a theoretical peak of 450Mbps.

MediaTek mentions that the first phones powered by the Helio X30 will be available in the second quarter of 2017. Chinese manufacturer Vernee has already announced that its upcoming Apollo 2 will feature the Helio X30, and we'll likely hear more about other handsets that will be powered by the SoC in the coming weeks.

Harish Jonnalagadda
Senior Editor - Asia

Harish Jonnalagadda is a Senior Editor overseeing Asia at Android Central. He leads the site's coverage of Chinese phone brands, contributing to reviews, features, and buying guides. He also writes about storage servers, audio products, and the semiconductor industry. Contact him on Twitter at @chunkynerd.

  • I wonder....Are all these cores really necessary?
  • Yes. No. Maybe. More CPU cores (theoretically) mean more software threads can run. You can run more than one thread on a single core but the core itself has to switch between them. Android has a single thread (a very complex thread) running for the user interface, This doesn't mean the UI itself, but anything and everything that the user sees, taps, enters or requests. Background threads run, too. Potentially a lot of them. When the software performs a task --asking a program to start by tapping an icon creates a task (the request) that starts the task (starting the application) as an object on a background thread the data needs to be handed off to an object on the UI thread. Yeah, this is a complex idea and hard to describe. Think of it this way maybe. You have one stream of data coordinating with the user. You have many streams of data doing all kinds of things in the background. All the background streams of data need to communicate with the one stream fo data that communicates with the user. Ideally. each stream of data would have its own CPU core. That can;t happen, so the tasks on the streams of data need to be scheduled on the fly. Properly programmed — both in the kernel and for each application — more cores = better. Poorly programmed (using what Google wrote for their Android to schedule and run tasks on YOUR version of Android) more cores = cores using power and doing nothing that helps. Unfortunately, most companies do the latter. That's why people building custom kernels can make such a drastic difference in how fast an Android OS can perform these tasks.
  • I completely understand what you are saying but I'm still shaking my head at this chip. Arm stated that two A73 cores could fit in the same footprint as an 4 A53 use the same amount of power as 4 A53 all while matching or exceeding the performance of the 4 A53 cores. This soc makes no sense from a performance stand point. Shoot seems to be geared to the Asian market where core count is more important then performance.