Template talk:Infobox Machine

Template error

What's with this? busarch is broken. FireflyST (talk) 00:24, 16 May 2007‎

Clock speed / cycle time

So, as we discovered with PDP-8/PDP-8/S, 'clock speed' can be confusing, since some machines take a basic clock and immediately divide it down, thereby giving an incorrect sense of their speed. 'cycle time', although similarly varying in meaning from machine to machine, is at least one step up the architectural ladder, and thus more likely to be a compare-able measure of the machine's speed.

I have left the old 'clock speed' parameter, so we won't have to run around and change every place that uses this, but either will show the caption 'cycle time'. Jnc (talk) 14:22, 22 August 2019 (CEST)

There has been a query about dropping 'clock speed'. Maybe we should have both? What do people think? Jnc (talk) 14:11, 24 August 2019 (CEST)

OK, no feedback, so I'll go ahead and add it, on the principle that more info is never bad. Jnc (talk) 13:50, 25 August 2019 (CEST)

This is a really tough question, because it encompasses decades of computer architecture

In the beginning, everything was measured in clock cycle time,

Then they went to clock speed...

Then to MIPS. ( Vaxen mips.

Then to iComps...

Then smushed among it all LinPack Flops, MegaFlops, Giga Flops, TeraFlops, and now Petaflops.

PDP-8s were measured in cycle times. VAX 11/780s began the MIPS.

"For a while the VAX-11/780 was used as a standard in CPU benchmarks. It was initially described as a one-MIPS machine, because its performance was equivalent to an IBM System/360 that ran at one MIPS, and the System/360 implementations had previously been de facto performance standards. The actual number of instructions executed in 1 second was about 500,000, which led to complaints of marketing exaggeration. The result was the definition of a "VAX MIPS," the speed of a VAX-11/780; a computer performing at 27 VAX MIPS would run the same program roughly 27 times faster than the VAX-11/780. Within the Digital community the term VUP (VAX Unit of Performance) was the more common term, because MIPS do not compare well across different architectures. The related term cluster VUPs was informally used to describe the aggregate performance of a VAXcluster. (The performance of the VAX-11/780 still serves as the baseline metric in the BRL-CAD Benchmark, a performance analysis suite included in the BRL-CAD solid modeling software distribution.)"

[1]