A person appearing on Twitter as RetiredEngineer, who occasionally comments on upcoming news in the world of processors and translates related texts from Chinese into English, pointed out the statement of a former Intel employee employed in China. After numerous questions, he decided to publish what is behind the situation at Intel in one of the discussion forums. RetiredEngineer then translated and published his (very extensive) statement in parts (over two weeks).
A key part of the communication concerns the situation with production processes. Everything went well until the completion of the 22nm process, the preparation of which was led by a person named Kaizad. Within the Intel Technology Manufacturing Group (TMG), there are (or have been) two teams that have taken turns. The development of the original 14nm process was led by Sanjay. During the preparation of this generation, it became clear that the preparation of new processes would not go nearly as quickly and smoothly as before. However, Sanjay, who had experience solving problems that arose during the development of the 14nm process, left because TMG became stuffy (more details below).
Kaizad was then in charge of the 10nm process. We know what problems this generation has been facing. One of the reasons for the problems (probably meant with the second generation of the process) was that the team emphasized the feedback on the development of graphics chips. For graphics cores that are large, one of the main criteria is density, ie how many transistors can be obtained per square millimeter of silicon. For smaller processors, the priority is what frequency can be achieved and how much the process will consume. However, the development of GPUs preferred high density (logically he did not care if the process handles 3 or 5 GHz, or what power consumption at the same time), which led to the emergence of technology almost unusable for mass production of CPUs. The following is a description of the (known) 7nm generation problem.
However, the orientation of one version of the process on graphics chips, but the personnel situation described by the ex-employee, will not be a somewhat more serious problem and the reason for the delays. After the 22nm process, TMG began to tighten because the management wanted to keep up the pace of development and the result was a military hierarchy and a workplace atmosphere that was no longer tolerable for many employees. Since then, TMG has suffered from a high circulation of employees who do not last long in the department and leave. If much of the remaining hard-core workforce is still being used to teach and train newcomers, it’s no wonder Intel isn’t being able to meet the announced deadlines and there are delays.
Jim Keller’s era was described by a former worker as a gust of fresh wind and an improvement in the working atmosphere. After about two years, however, it ended, Jim Keller resigned on June 11 and left the company. He promised to provide consultations until about Christmas.
Source: Diit.cz by diit.cz.
*The article has been translated based on the content of Diit.cz by diit.cz. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!
*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.
*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!