Talk:Berkeley RISC

When were the projects producing their results?

What were the years in which the seminal research projects were actively producing their published academic results? optikos

Excellent point, adding... Maury 13:13, 10 January 2006 (UTC)

RISC article

I don't propose merging the two, but I believe that some parts of Berkeley RISC should move to RISC, and some parts of RISC (particularly in RISC#Early RISC should move to Berkeley RISC. Thoughts? StuartBrady 13:37, 12 January 2006 (UTC)

Yes, the section "The RISC concept" is most probably redundant with the article RISC. Somebody would have to put in the work and move such content to the RISC-article, making it better. I do not volunteer. User:ScotXW^t@lk 16:24, 12 July 2014 (UTC)

faster and slower

I'm confused.

The article currently claims

"... the RISC I was twice as fast as the VAX, ... Even though the RISC design had run slower than the VAX, it made no difference to the importance of the design."

Which at first glance appears to be a contradiction.

And what exactly was "the importance of the design"? --68.0.124.33 (talk) 20:03, 22 February 2008 (UTC)

"the RISC I was twice as fast as the VAX" means "the simulations of RISC I indicate that, had they gotten the RISC I to take .4 μs, rather than 2 μs, to complete an instruction, it would have been twice as fast as the VAX".

"Even though the RISC design had run slower than the VAX" A design doesn't run, a hardware implementation of the design runs; I changed it to "Even though the RISC I hardware had run slower than the VAX".

The key here is

In testing, the chips proved to have lesser performance than expected. In general, an instruction would take 2 μs to complete, while the original design allotted for about .4 μs (five times as fast). The precise reasons for this problem were never fully explained. However, throughout testing it was clear that certain instructions did run at the expected speed, suggesting the problem was physical, not logical.

so, for some instructions, the chip didn't run at full speed. Had it done so, the simulations indicated that it would have been twice the speed of the VAX, but that didn't happen.

And what exactly was "the importance of the design"? Presumably it was that, had they been able to fix the "physical" problems with RISC I, it would've outperformed the VAX, at least on some tasks, with a much simpler design. Guy Harris (talk) 22:19, 24 April 2025 (UTC)

"circuitry that eliminated one line per bit"

The following sentence appears in the article: "The key difference was simpler cache circuitry that eliminated one line per bit (from three to two)"

The meaning of this is not completely obvious to me and would probably not be clear to many others. My guess is that the writer was referring to a single line-width in layout for the the memory cell used for cache. If that impression of mine is correct, the text needs to be altered to specify what is meant by a line or a line-width, and the concept of a memory cell may have to be introduced. I can do the clarification, but I am still not sure I get what's meant by "one line per bit". Can anyone help?Dratman (talk) 19:55, 20 October 2020 (UTC)

General RISC history in the "Follow-ons" section

I redid most of the stuff in Berkeley RISC § Follow-ons to give credit to Stanford MIPS and the IBM 801, as well as Berkeley RISC, for RISC taking off in the 1980s and 1990s, but that's all generic RISC history, and probably belongs in reduced instruction set computing rather than here. I'm inclined to leave just a general "Amid the takeoff of RISC in the 1980s, Sun developed SPARC based on Berkeley RISC." note there. Guy Harris (talk) 21:55, 24 April 2025 (UTC)