Posts: 194   +4
Staff
Why it matters: Arm and x86 processor architectures made a fair number of mistakes when first introduced to the market. You would think that developers working with the open-source RISC-V would have learned by now, but you would be wrong – at least according to Linux creator Linus Torvalds. He believes RISC-V is fated to repeat the same mistakes that Arm and x86 did as they gained traction. One reason for this is the wide gulf between software and hardware developers.

Linus Torvalds, the creator of Linux, believes that RISC-V will likely repeat the same mistakes made by Arm and x86 architectures as it evolves to become more complex to meet market demands. In a larger sense, this is a common point of debate when discussing the future of processor architectures and the trade-offs between simplicity and feature-rich designs.

Torvalds said in an interview that this is a trend that seems to happen every time new technology is introduced. "One frustration when Arm became a server platform is that they redid all the mistakes I had already seen a decade or two earlier on x86," he explained.

RISC-V is a relatively new open-source instruction set architecture (ISA) based on reduced instruction set computing (RISC) principles. Developed at the University of California, Berkeley in 2010, it has gained considerable momentum with over 10 billion chips containing RISC-V cores having shipped by the end of 2022.

It is making inroads particularly in China, where some tech companies are using it to bypass US sanctions. Just last month, Hong Kong manufacturer DeepComputing introduced a laptop built around the RISC-V ISA. However, it's still not at the performance level to compete against current generation x86 and Arm processors.

One reason why developers seem fated to reinvent the wheel is the significant gap between hardware and software developers, making it challenging to coordinate efforts effectively, Torvalds claimed.

"Even when you do hardware design in a more open manner, hardware people are different enough from software people [that] there's a fairly big gulf between the Verilog and even the kernel, much less higher up the stack where you are working in what [is] so far away from the hardware that you really have no idea how the hardware works," he said.

Also, a new generation of people involved in RISC-V development may not be aware of past issues, leading to a "we didn't think about that" scenario for some features, he added.

For these reasons, Torvalds suggests it will take a few generations of RISC-V processors to identify and address unforeseen issues.

Still, RISC-V will likely have an easier time making inroads in the market compared to earlier architectures, having already made significant progress, particularly in embedded systems and some specialized applications. That said, it faces an uphill battle in areas like high-performance computing and mobile devices where Arm and x86 dominate.

Some of its limitations include an ecosystem that is still relatively immature with the tools, software, and support still developing compared to established architectures, and possible compatibility issues that could arise in different implementations.

Permalink to story:

 
And transmeta made some mistakes. Biggest mistakes were taking vc money and selling out to Intel (who subsequently destroyed their tech and all emails).
 
If Linus Torvalds is so smart, why doesn't he come out with the next processor then? I'm tired of him, or just the media pointing it out, calling everything garbage.
 
If Linus Torvalds is so smart, why doesn't he come out with the next processor then? I'm tired of him, or just the media pointing it out, calling everything garbage.
Because different professions exist for a reason.
 
Linus has some valid and reasonable concerns. Hopefully awareness can be enough to side-step those issues.

Most people these days hate intelligent people and immediately think they are up to something.
Which is ignorant in the extreme by default. Very sad.

The general populace should revere the smart and brilliant people in life instead we have "Cancel Culture" types making influence.
 
This was the same argument that was made by JLG in his interview in the BeOS Bible. Wintel Microsoft fud against BeOS, OEM bootloader restrictions, etc....



Except instead of a single company, Be Inc, we have an entire nation fighting Wintel and the old ideas, such as "one very inexpensive cpu" versus "many cheaper processors with optimized software".



Certainly RiscV isn't ready to best Intel. It's running in the background like a thousand wildebeests.



RiscV is the only hardware with NO Microsoft operating system. Competition is a good thing.
 
We need to rise above just processor architectures and think about system architecture, and the different styles of system from general multi-processing systems to single-purpose embedded systems.
Here is what I have written in response to an article about what Torvalds said:
medium.com/@ianjoyner/what-we-need-in-computing-is-system-architectures-beyond-processor-architectures-9d3450efb72b
 

Similar threads