At the ongoing MWC Shanghai 2018, Qualcomm announced three new chips called the Snapdragon 632, 439 and 429. Qualcomm states that the new chips have been designed to outperform its predecessors by 40 per cent and supports AI capabilities. The company also confirmed that the new processors would be made available to the smartphone manufacturers in the H2 2018.
Taking a detailed look at the processors, the new Snapdragon 632 SoC is built on a 14-nm FinFET process that provides better power efficiency and performance. The new chipset will have Kryo 250 semi-custom cores, where the first four cores are based on Cortex-A73 clocked at 1.8GHz, while the remaining four cores will have Cortex-A53 with the same frequencies.
As for the graphics unit, this new processor features Adreno 506 GPU which is claimed to be 10 per cent faster and is expected to support the latest APIs including Vulkan. Other features include a Hexagon 546 DSP, supports ML frameworks and the Qualcomm Neural Processing SDK, which will help the mid-range devices to run AI features smoothly. The 632 SoC also offers the X9 LTE modem, which allows users to get download speed up to 300Mbps and 150Mbps for upload.
On the other hand, the new arrivals in the 400 series (439 & 429 SoC) are built on the 12-nm FinFET process giving it better power usage. The Snapdragon 439 SoC has an octa-core processor of A53 Cortex cores, which is split into two groups. The first four cores designed for high performance are clocked at 1.95GHz whereas the rest four cores have a clock speed of 1.45GHz for power efficiency purposes. Coming to the second processor – Snapdragon 429 SoC comes with a quad-core CPU of A53 Cortex cores clocked at 1.95GHz.
Both the processors will get an Adreno 505 and 504 GPUs respectively. Considering the performance boost, Snapdragon 439 & 429 SoCs are claimed to get up to 25 per cent more than the predecessors. Both the chipsets feature X6 LTE modem which supports download and upload speed up to 150Mbps and 75 Mbps respectively. The 439 and 429 processors have Hexagon 536 DSP and support ML frameworks like Caffe/2, TensorFlow/Lite, ONNX.
(The above story first appeared on LatestLY on Jun 27, 2018 04:49 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).