Table of Contents
Vivo V2 Chip New Features and Capabilities
When a factory shoved a camera lens onto a cell phone, I wonder how many people had the same thought as we did – “Mobile phone photography hardware went to the head.”
The ultimate hardware stack is the first option for cell phone manufacturers in the image quality competition. From the “bottom of the big level of pressure” to the periscope/camera lens on the phone, cell phone manufacturers have the physical means to improve the optical performance of almost all the means to make.
How can the optical hardware pile up to the extreme case further? The answer to this may have to be given to – computational photography. Is it SoC or ISP? There is no right or wrong route.
Whether it is the controversial Huawei “moon mode” or today’s release of Vivo V2 Chip, major cell phone manufacturers are beginning to plow the potential of computational photography, it has become the next breakthrough in the cell phone industry.
Computational photography sounds advanced, the actual is in addition to the lens and sensor optical performance, the captured picture information optimization, outputs higher quality or more in line with the needs of the picture.
So strictly speaking, the early beauty filters and HDR are part of computational photography. We are not new to computational photography. But to do more with computational photography, the extra processing of the image creates an exponential increase in hardware computing performance.
Apple recognized this problem early on. They chose to develop their own SoCs and adjusted and optimized them concerning the performance requirements and power consumption of specific applications. iPhone 13’s stunning Cinematic Mode is also based on AI analysis of the relationship between people in the picture, recording shallow depth-of-field video and softly switching focus so that the video has a movie-like quality.
Rely on self-developed SoCs to solve the computing photography software and hardware packages. So, Vivo just released its own Vivo V2 Chip, which is a low-power AI acceleration chip with AI-ISP architecture.
On November 10, Vivo held the “Dual Core x Imaging Technology Communication Meeting”. The meeting focused on imaging, and performance, and introduced the deep cooperation between Vivo and MediaTek, Vivo’s new generation of self-research chips and other new technologies, showing the latest achievements of vivo in independent research and development, and open cooperation.
1. Self-developed Vivo V2 Chip architecture iteration, bringing the ultimate arithmetic power and energy efficiency ratio
Since 2021, Vivo has gained market recognition through two generations of self-research chips, V1 and V1+. This time, the self-research chip V2 opens a new chapter of mobile imaging with a new iteration of AI-ISP architecture.
Vivo V2 ISP brings an overall improvement in compatibility and functionality, with significant upgrades to the on-chip memory unit, AI computing unit, and image processing unit.
The FIT dual-core interconnection technology is proposed to establish a new high-speed communication mechanism between the Vivo V2 chip and the flagship platform of Dimensity 9200, enabling two chips with completely different architectures and instruction sets to complete dual-core interconnection synchronization within 1/100th of a second, realizing optimal coordination and high-speed collaboration of data and arithmetic power.
Thanks to the near memory DLA (Vivo’s own AI deep learning accelerator) module and large-capacity dedicated on-chip SRAM (high-speed low-consumption cache unit), the self-research chip V2 re-matches the arithmetic capacity, arithmetic density, and data density to significantly increase the capacity and computing speed of on-chip cache.
Compared with the usual DDR external memory design used in NPUs, SRAM data throughput power consumption can be theoretically reduced by up to 99.2%, improving the energy efficiency ratio by 200% compared to traditional NPUs.
The FIT dual-core interconnect and near-storage DLA architecture design enables the AI-ISP architecture of the self-research chip V2 to be established and complement the ISP algorithm and arithmetic power of the platform chip NPU to achieve the ultimate image processing effect and the ultimate energy efficiency ratio.
2. Self-developed image algorithm iteration, cooperate with V2 image arithmetic to expand professional imaging in all scenes
In this communication meeting, Vivo brought advanced versions of self-research image algorithms such as telephoto images, motion capture, dark light capture, etc. With the support of a self-research chip, V2 curing algorithm capability and the addition of HDR, NR, and ProMEMC, Vivo’s mobile image technology will be pushed to a new height.
The “ultra-clear image quality engine” with the optical super-resolution algorithm as the core can restore about 35% of the clarity information of the focal length above 5 times; Ultra Zoom EIS technology integrates the three modules of IMU, OIS, and EIS, in the process of high-magnification zoom shooting It can effectively offset the jitter and ensure the stability of the preview image. Even when holding the camera, the lens can be stably moved, which brings a qualitative leap to Vivo’s telephoto capability.
To solve the shutter delay problem and shorten the gap between cell phone shooting and professional cameras, Vivo has developed “zero latency” capture and a new generation of the motion detection algorithm. By optimizing the image processing process to enhance the sensor start-up speed, the shutter delay time is reduced to 30ms, reaching the level of professional cameras. It can quickly take pictures with a press of the shutter, which greatly improves Vivo’s ability to capture moving objects and helps users capture more fleeting moments in life.
To meet the full-scene image needs of professional users, Vivo also realizes the dark-light capture function through algorithm overlay and photography full-link optimization, which significantly improves the sensor’s light-sensitive capability in dark-light scenes; through multi-frame fusion technology and self-developed RawEnhance 2.0 algorithm to realize motion picture overlay and eliminate trailing shadows, enabling users to easily capture high-quality dynamic images in dark-light environments.
3. Five jointly developed features to unleash the potential of the flagship platform of Dimensity 9200
The latest generation of MediaTek’s flagship SoC platform, the Dimensity 9200, is a typical representative of this communication event. Vivo has been involved in the development of this platform at an extremely early stage, and through close cooperation between the two parties for 20 months, it has brought five major features, including MCQ multi-loop queue, Honor of King adaptive image quality mode, chip-level eye protection, APU framework fusion, and AI airport mode.
MCQ Multi-Cycle Queue is a new processing engine redefined to bring out the ultimate CPU performance, supporting up to 8 channels of data transfer between CPU and UFS, making application software switching and background download wakeup faster and smoother.
The chip’s eye protection technology reduces blue light in real-time by monitoring the percentage of blue light on the screen in real-time through an innovative algorithm and hardening into IP, making the percentage of high-energy visible blue light less than 5% and reducing the degree of color shift by 12%. At the same time, the screen color effect is dynamically adjusted according to the monitoring results to ensure that the screen does not deviate from the color while reducing blue light.
APU framework fusion is based on the hardware features of the sixth generation APU of Dimensity 9200 and encapsulates the underlying common capabilities of MediaTek into the self-researched VCAP heterogeneous computing acceleration platform.
Deep optimization is realized from the bottom of the platform to the framework layer, allowing algorithms to be collaboratively scheduled among multiple processors, bringing significant energy efficiency improvements, and in this way, camera ultra-clear documents, live text, and other features are introduced.
AI airport mode can achieve an average energy saving of 30% during users’ flight; after closing flight mode, the fastest network connection can be quickly restored within 1.52 seconds, making the re-networking speed one step faster.
4. System kernel-level joint tuning, multiple scenarios show flagship platform extreme performance
In addition to joint research and development, Vivo and MediaTek also jointly tuned several details to help Dimensity 9200 deliver the ultimate performance. After the joint research of Vivo and MediaTek, the performance of Dimensity 9200 has once again broken through the limit, with a runtime score of over 1.28 million in the AnTuTu Benchmark.
In the gaming scenario, Vivo brings black technologies such as game super score and multi-link and all-around optimization of a gaming experience through the speedy start engine and network acceleration engine; in the video scenario, for the first time, advanced cyclic parallax network is realized with extremely low power consumption, which significantly optimizes energy efficiency performance and reduces energy consumption by 15%. In particular, in the 4K 60fps extreme video scene, the power consumption of Dimensity 9200 is reduced by 25% compared to Dimensity 9000.
The self-research Vivo V2 Chip demonstrates Vivo’s determination and efforts to win users and the market through continuous iteration of the underlying core technology; the mobile imaging technology, which has been continuously strengthened and pushed up, helps Vivo further stabilize its high-end flagship positioning; the continuous in-depth cooperation with the world’s top SoC flagship platforms and the fruitful results show Vivo’s unswerving open attitude and profound achievements on the road of joint R&D.