A team of researchers at the University of California, Berkeley, plans to launch open source chips

"Editor's note" The current chip market is largely controlled by companies such as Intel, ARM and IBM, and a group of researchers at the University of California, Berkeley, are planning to launch open source chips, in frustration over existing computer chips. The risc-v instruction set architecture they used was originally used to teach computer architecture, but now its developers want to push it into emerging markets such as cloud computing and IoT. The project will bring the Gospel to small businesses and independent developers.

The following is the original text:

Tired of the limitations of current computer chips and their associated intellectual property restrictions, a research team at the University of California, Berkeley, is pushing an open source alternative. The risc-v instruction set architecture was originally developed to help teach the university's computer architecture, but now its developers want to mainstream it, helping to drive cloud computing and IoT development.

One of Risc-v's developers is David Patterson, who also participated in the creation of RISC instruction sets in the 80 's. The mainstream chip architecture (Intel and ARM) is protected by patents, and even if a licensing fee is paid, the instruction set is complex and requires a lot of work to adapt to specific job requirements.

David Patterson

However, if you are a big company, you can buy licenses from big chip vendors, invest a lot of time and money to develop some very good technology, which of course is very good. But Patterson seems to prefer small companies and independent researchers-small companies or researchers who want to develop chips for their professional applications, but don't have the wherewithal to do so. They need to understand the underlying instruction set, chip design, and open shared work without fear of violating the license terms.

"To achieve this, you have to have an unlimited set of instructions," Patterson said.

In fact, there are other open source instruction sets, including OpenRISC and SPARC V8, as well as IBM's OpenPOWER and mips-based prpl. Patterson that it is arbitrary to discuss the success of these projects prematurely, especially in the face of small companies, individuals or universities.

RISC-V Demo Board

Just a few months ago, Patterson and his colleagues realised that they should try to push risc-v out of the classroom, and that they felt they should make a difference with the desperation of the existing architecture and the interest in open source chips. The University of California, Berkeley, has created a number of risc-v based cores and has several other projects in place. Patterson and his colleague Krste Asanović recently published technical papers explaining RISC-V technical specifications and details.

Patterson says Risc-v will be better and more effective in many ways because it has a small code base and other features that make it more suitable for system-level chip design and dominate the current computer world, thanks in large part to arm. With the increasing demand for connectivity devices, the maturity of Raspberry Pi and the development of large-scale cloud architecture, a thriving risc-v community should be able to design the chips needed for it.

Patterson in answer to my question about how risc-v fits into an existing Open-source project, such as Facebook's Open Compute project, "We think this project is more effective for cloud design than standard processors," He also mentions that the UC Berkeley Amplab is working on a data processing and distributed system, suggesting that a simple customizable chip architecture can also help in some cases the problem of fault tolerance and 64-bit addressable memory space deficiencies.

"I think this is going to happen and hardware will become more professional for customers and cloud computing," Patterson said.

Source text link: RISC Creator is pushing open source chips for cloud computing and the Internet of Things (Zebian/Wei)

Free Subscription "CSDN cloud Computing (left) and csdn large data (right)" micro-letter public number, real-time grasp of first-hand cloud news, to understand the latest big data progress!

CSDN publishes related cloud computing information, such as virtualization, Docker, OpenStack, Cloudstack, and data centers, sharing Hadoop, Spark, Nosql/newsql, HBase, Impala, memory calculations, stream computing, Machine learning and intelligent algorithms and other related large data views, providing cloud computing and large data technology, platform, practice and industry information services.