Tech Hammer

Chapter 87 There are also different opinions

"Professors, teachers, hello everyone, thank you very much for joining our EDA project team with your trust in me. Today's meeting is to briefly talk about the design concept of our minimalist EDA and what we will do next. Make a division of labor.”

In the small conference room on the fourth floor of the laboratory, Ning Wei was giving a report to everyone in the project team.

There is still pressure.

After all, except for Yu Xingwei, the youngest one is also a round older than him.

"Let's put it this way, the EDA software our project team is about to design is not as complicated as everyone thinks. I call it object-oriented EDA software design."

"Here I first assume that every structure that can be produced by existing technology is an object. I divide these objects into four types."

"They are respectively Yuan, which is the smallest structural unit that can be produced by existing technology. This unit can be updated with technological breakthroughs."

"Group, a structure with a single function formed by several elements connected and arranged with each other.

"A group is a structure in which several groups with the same or different functions are connected and arranged to form a structure that can independently complete a specific function."

"Container: A structure used to hold elements, groups, or groups."

"Every object must contain two parts: properties and functions."

"Attributes not only include the physical attributes of the object, such as size, size, etc., but also include the index and presentation attributes of the object, such as representative symbols, unified serial numbers, etc., and also include the layout attributes of the object, such as connection methods and arrangements, etc. .The function must have input parameters and return value, which is a description of the object's function, such as single-item conduction, amplification, cut-off, etc."

Ning Wei explained his basic design concept to the project team members based on the PPT he produced yesterday.

“From this perspective, everyone should see that our EDA design idea is actually very simple. On the one hand, we want the operator to easily position, arrange and wire symbols representing different types of objects on the design interface; On the other hand, all the operations of the operator are actually the computer background calling the functions of the object in series with each other, ultimately forming a large function process."

"The above are the requirements for the design process. At the same time, our EDA software must have simulation functions. The EDA simulation process is the process of inputting parameters into the large function process and recording the feedback results. It also includes the parameters and returns passed between each sub-process. value monitoring.

"The EDA instantiation process requires the layout position and arrangement of objects to generate a running script in order to control the lithography machine for production and manufacturing."

“Based on the above requirements, we directly guide the software structure.”

"Object Library: Used to define existing objects and save new designs for the future."

"Function library: Functions used to save existing or future objects."

"Design interface: for laying out, arranging, and connecting objects."

"Function connection: used to concatenate and call the functions of arranged and connected objects in the way they are arranged and connected."

"Simulation and monitoring: used to send parameters to the structure and detect the return value of each function node."

"Production function: used to script the layout, arrangement and connection of objects to the lithography machine for production."

"In addition, because this is a brand-new EDA software, we cannot rely on the peripheral interface to introduce plug-in programs, so we also need to have a library that is different from other EDA. This library corresponds to the object library. Every time we design a structure, we need to update an instruction set and supporting functions for this structure."

"In other words, every chip we design for a specific problem must have all-round support. This is where our product differentiation lies. Our products are not just for professional EDA designers from specific major manufacturers. , my original intention is to enable an elementary school child to use our software to design a usable chip after reading the instructions."

"After our software is launched, the structure and supporting instruction set and functions designed by professionals can be synchronized into our library after being verified and reviewed."

"Here is another requirement for our products, which is to have a built-in logical judgment ability on the structure. Simply put, what you ask for is what you get. In other words, users only need to be able to describe their needs more accurately. Our software It will automatically recommend the architecture and supporting configuration files that can fulfill this requirement!"

"I think everyone already understands the design concept of this software. It is simple. It is simple enough that there is almost no promotion cost."

"Do you guys have any questions?"

The project team members looked at each other.

Because it is difficult for people in the game to imagine the significance of developing such an EDA software.

As a toy for children?

Yes, simplicity can save a lot of promotion costs, but it also means that the practical value of this thing is not high!

First, it is almost impossible to design and simulate large and complex general-purpose chips, such as CPUs and GPUs.

The other is the application layer.

What Ningwei means is to bypass the various existing architectures and instruction sets, which also involves compatibility issues.

In fact, there are ready-made solutions for AI chips at this stage, such as FPGA.

Different from Ning Wei's idea, FPGA can be directly understood as a universal chip with a unified structure.

Users burn in the FPGA configuration file to define the connections between these gate circuits and memories, and use hardware description language to design the FPGA hardware circuit. Every time programming is completed, the hardware circuit inside the FPGA has a certain connection method and a certain function. The input data only needs to pass through each gate circuit in sequence to get the output result.

This method is obviously more cost-effective than the ASIC model designed by Ningwei. Although FPGA also has its own shortcomings, such as its performance is not as good as that of ASIC chips, and its price is higher than that of ASIC chips.

The reason is that various artificial intelligence deep learning algorithms are now in a period of rapid development, with extremely fast iterations. If Ning Wei’s idea is to follow Ning Wei’s idea, each structure must be designed with a corresponding optimized programmable configuration file, which means that every Every algorithm change requires a new generation of dedicated chip structure to be updated in the object library.

Once updates are slow, you will fall behind...

Therefore, few people in the entire project team are optimistic about the prospects of this EDA software.

But everyone has different ideas.

Some people think it doesn't matter. After all, the funding has been allocated, and no matter what, everyone can make some money. As for whether this software has practical value, there is nothing to care about.

Anyway, it is a special appropriation, so it is not used in vain.

But some people think this is inappropriate.

Public money is also money. Nowadays, the laboratory's plate is so big. If more is used on their projects, funds for other projects will be cut.

If you spend this money on projects that are destined to be of little value, what about other valuable projects?

After the silence, someone made a different sound.

"Ning Wei, can you explain, what is the original intention of designing this software? What is its practical value?"

Tap the screen to use advanced tools Tip: You can use left and right keyboard keys to browse between chapters.

You'll Also Like