Chip start-up Houmo Smart Lights up the industry's first large-scale computing power AI chip

Houmo smart chip development board

On May 23, Houmo Smart, a large computing power AI chip start-up based on storage and computing integration technology, announced that its self-developed industry's first storage and computing integrated high computing power AI chip successfully lit up and successfully ran the intelligent driving algorithm model. .
Based on architectural innovation, this chip uses SRAM (Static Random Access Memory) as the integrated storage and computing medium. Through the deep integration of storage units and computing units, it achieves high performance and low power consumption. The sample computing power reaches 20TOPS, which can be extended to 200TOPS , the computing unit energy efficiency ratio is as high as 20TOPS/W.
This is the first AI chip in the industry based on a strict in-memory computing architecture, with AI computing power reaching dozens of TOPS or higher, and supporting large-scale visual computing models. Compared with the large computing power chip under the traditional architecture, this chip has significant advantages in terms of computing power and energy efficiency ratio.

Houmo Smart's first chip lights up the scene

This chip adopts a mature 22nm process, which can effectively control the cost while improving the energy efficiency ratio. In terms of flexibility, this chip supports mainstream algorithms on the market, as well as operators customized by different customers, and is more suitable for high-speed iteration of algorithms.
In edge-side high-concurrency computing scenarios such as intelligent driving, in addition to high computing power requirements, there are also high requirements for chip power consumption and heat dissipation. In the current conventional architecture chip design, the performance improvement speed of the memory system greatly lags behind the performance improvement speed of the processor. The limited memory bandwidth cannot guarantee high-speed data transmission and cannot meet the computing needs of high-level intelligent driving. Second, the transmission of data back and forth will generate huge power consumption. Based on this chip, Houmo Intelligent has run through the multi-scene and multi-task algorithm model in the intelligent driving scene for the first time on the in-memory computing architecture, providing a new technical path for high-level intelligent driving.

Based on the self-developed storage and computing integrated large computing power AI chip, the intelligent driving algorithm model is run through

Houmo Intelligent was founded at the end of 2020 and has completed 3 rounds of financing. The chips and solutions it provides with high computing power, low power consumption and high energy efficiency can be applied to edge terminals such as intelligent driving, pan-robots, and cloud reasoning scenarios. Houmo Intelligent carries out the design of AI chips with large computing power from the innovation of the underlying architecture. R&D personnel need to redesign circuits, cell arrays, tool chains, etc. according to traditional storage devices, and must break through various physical and structural technical problems. The successful lighting of the chip marks a key breakthrough of Houmo Intelligent in the engineering implementation of large computing power storage and computing integration technology.