Talk:Clock rate
![]() | Computing Start‑class | |||||||||
|
This is the talk page for discussing improvements to the Clock rate article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
Archives: 1Auto-archiving period: 64 days ![]() |
Moore's Law
I am revising the section on Moore's Law, as it relates to transistor count, not clock speed. — Lomn | Talk / RfC 15:39:21, 31 August 2005 (UTC)
Bleh
Yick... This article is another microelectronic PC-centric mess. Anyone who wants to help improve it would be much appreciated... If not I'll fix this up shortly after I finish working on the Central processing unit article. -- uberpenguin 16:34, 14 October 2005 (UTC)
- I agree. I removed the statement about the clock having a longer "zero" time than a "one" time, but the article seems to be oriented to those who would seek to overclock a chip. Overclocking has nothing to do with binning. However, it results in changes in signal integrity, power consumption, heat dissipation, packaging limitations, reliability, etc. Power relates to I2R power losses in the chip interconnect which complicates signal integrity and reliability of the chip. Its an engineering thing - you might not understand. --71.245.164.83 (talk) 01:38, 21 February 2011 (UTC)
Processing power
Why this redirects here? Processing power has different meaning. kuszi 19:34, 31 March 2006 (UTC).
- Yeah, a high clock rate does not necessarily mean power. How should the page on processing power be if it would no longer redirect here? Geekosaurus 23:18, 27 July 2006 (UTC)
- Agree with the commenters here too. Processing power is controlled by more than just clock rate. Needs work. --71.245.164.83 (talk) 01:44, 21 February 2011 (UTC)
Done. Around here, "processing power" generally refers to the ability of the computer to do useful work. So I redirected "processing power" to "computer performance", which has links to various contributors to that ability, such as clock rate, instructions per second, instructions per cycle, and etc. --DavidCary (talk) 14:49, 18 January 2015 (UTC)
Merger
It does seem to me that this should be part of clock signal, as it is a property of something with clock signals (if this makes any sense). Geekosaurus 23:21, 27 July 2006 (UTC)
speed of light as the upper limit
Someone should mention how the speed of light ensure we'll never have Petahertz chips. —Preceding unsigned comment added by Carlj7 (talk • contribs)
- Why is that so?? Realg187 21:42, 13 December 2006 (UTC)
- Nothing goes faster than c. — RevRagnarok Talk Contrib 22:47, 13 December 2006 (UTC)
- Not necessarily. It's just that it's impossible (with current theory) to send data faster than the speed of light due to special relativity. Anyway, that's pedantic for the purposes of this article. -- mattb
@ 2006-12-14T00:07Z
- Not necessarily. It's just that it's impossible (with current theory) to send data faster than the speed of light due to special relativity. Anyway, that's pedantic for the purposes of this article. -- mattb
- Nothing goes faster than c. — RevRagnarok Talk Contrib 22:47, 13 December 2006 (UTC)
- Assuming a maximum signal path of 1 millimetre, the maximum clock rate assuming c as the propagation rate (which is idealistic) is about 300 GHz. If you wanted to be really pedantic, the limit for any electronic clock rate is limited by the Compton wavelength of the electron to about 123 exahertz. It gets much worse if you try more than one electron at once. :P --AlexWCovington (talk) 06:39, 23 January 2007 (UTC)
Rate
Its dumb how they dont say the clock rates, consumers have the right to know!!! I have a program that tells it very accuatly like to 3 deciamal places ex, instead of 232 or 233 it'll say 232.9!! Realg187 21:42, 13 December 2006 (UTC)
- What are you talking about? Modern consumer microprocessors? If so, the rate is available, just no longer in the name. — RevRagnarok Talk Contrib 22:47, 13 December 2006 (UTC)
Digital logic
The clock rate is the fundamental rate in cycles per second (measured in hertz) at which a computer performs its most basic operations such as adding two numbers or transferring a value from one processor register to another. Different chips on the motherboard may have different clock rates. Usually when referring to a computer, the term "clock rate" is used to refer to the speed of the CPU."
I'm pretty sure clock rate applies to any digital logic, not just computers. But I'm not bold enough to change it. —Preceding unsigned comment added by 82.130.27.40 (talk) 01:23, 15 November 2007 (UTC)
What limits a chip's clock rate
I came here looking to find out what limits a chips clock rate, but it's not here. I want to know what problems too high a clock rate causes and why it causes those problems.
Done. We now have a section clock rate#Limits to clock rate.
limits to clock rate
This section is very poorly written, and also inaccurate. As I understand it, overclocking does not involve physically replacing anything, as the section suggests. I considered fixing the grammar and wording, but didn't want to risk introducing inaccuracies, since I really don't know all that much about overclocking. Would someone knowledgeable please look at this section? 192.5.109.49 (talk) 21:57, 18 March 2008 (UTC)
- You are right about people overclocking a modern desktop computer. Those people don't actually physically replace clock crystals -- instead, they use BIOS settings to set registers on the clock generator chip.
- However, I've added a reference to WP:VERIFY that, in some cases, overclocking does in fact involve physically replacing clock crystals, and that physically replacing something is the only way to overclock some desktop computers. It may be misleading, but it is not inaccurate. --68.0.124.33 (talk) 02:44, 6 October 2008 (UTC)
- thanks for clarifying it, it sure raises a question while reading the article.
--93.172.97.251 (talk) 23:34, 10 November 2009 (UTC)
cycles per second
The rate need not be measured this way. The means of measurement is independent of the rate. —Preceding unsigned comment added by 80.189.103.145 (talk) 04:41, 4 December 2009 (UTC)
Definition of Cycle
"A single clock cycle (typically lasting only a few nanoseconds in modern microprocessors) toggles between a logical zero and a logical one state."
This is a sloppy defintion; it should be recast in terms of clock edges (as in, edge-to-edge). —Preceding unsigned comment added by 65.80.65.235 (talk) 08:38, 31 December 2008 (UTC)
It is also out of date, 1 ns. is 1 GHz... 2 GhZ = 0.5 ns, modern microprocessors use sub-nanosecond clocks. —Preceding unsigned comment added by 199.212.17.130 (talk) 16:07, 25 March 2009 (UTC)
This Wiki page is extraordinarily poor, considering that CPU clock rates are one of THE most frequent metrics that the modern computer-using person encounters. I'd like to help sourcing the page, as well as delete the 'speed' discussion in the preamble, as it's a misinformed and mostly irrelevant passage. I'll be removing and/or modifying that passage tomorrow unless somebody offers a serious objection...feel free to message me if you think there's a good reason for it to stay. If you know any great sources on clock rates, please let me know and we can begin to verify the information here. JakeKDodd (talk) 08:15, 9 April 2011 (UTC)
CPU
This abbreviation appears without expansion or explanation. —Preceding unsigned comment added by 80.189.103.145 (talk) 04:44, 4 December 2009 (UTC)
- I added a wikilink. Totsugeki (talk) 20:29, 23 January 2010 (UTC)
engineers try to find new ways for both creating
more power efficient processors and more heat controlled materials to build transistors from,
it could be better if this sentence could be referred with a profound article...
regards..--93.173.154.6 (talk) 22:01, 4 February 2010 (UTC)
processor speed
is this what people mean when they say processor speed? If it is, then that should be included in the article, if not why the redirect? —Preceding unsigned comment added by 98.206.155.53 (talk) 16:28, 15 March 2011 (UTC)
The article as presently written makes speed a misnomer. This is incorrect, and represents an unfortunate habit of too many technical people (especially in school), who will borrow words like power and winter—not to mention speed—from the pre-existing language, then back-impute some specialized technical definition onto the pre-existing language, then assert that nontechnical people who use the word in its original sense were wrong. The English word speed is easily broad enough in meaning to include what electrical engineers mean by frequency, especially when the frequency in question is connected with the rate at which data is processed. I wish to remove the sentence in question. If you restore it, would you reword it in a manner that respects the present POV? You just can't call speed a misnomer in this context, because it simply isn't. Tbtkorg (talk) 01:49, 15 June 2012 (UTC)
Historical Milestones biased
The section only covers silicon microprocessors. Earlier discrete component machines ran faster than the first microprocessors. For instance IBM 360's in 1966 at 4 Mhz, or CDC 6600's at 10Mhz, or the CDC 7600 at 34.6 Mhz! 192.139.122.42 (talk) 02:06, 1 November 2012 (UTC)
Poor Definition
"The clock rate typically refers to the frequency at which a CPU is running."
The term "running" is very ambiguous, would "the frequency at which the CPU can execute basic instructions" be more appropriate? I'd like to get some feedback before I edit it.Ducksandwich (talk) 16:56, 5 November 2013 (UTC)
System requirements
In Windows 10]] (system req.): "1 GHz clock rate IA-32 or x64 architecture". I've often seen GHz being mentioned for hardware and maybe as software req. (in WP) and can't really contradict that if sources say that. "Clock rate" is implied, and if we really want to *help* people, by linking to it then maybe we should make more clear how flawed it is. Maybe Hertz alone that is frequently linked gives a better picture.
IF clock frequently were fixed (as on desktop? mostly?) then stated frequency gives some indication - when given some other knowledge. For mobile, I doubt it. Dosn't hardware usually state some top speed given that can't be sustained? comp.arch (talk) 16:52, 16 February 2015 (UTC)
lower boundary
There is much talk about maximum clock rates. But what about minimum? AFAIK usual CPUs fail if they are run too slowly. --Ikar.us (talk) 12:27, 18 July 2017 (UTC)
- Static core is about this. --Ikar.us (talk) 12:34, 18 July 2017 (UTC)
Overclocking & underclocking
Yeah... Nobody really replaces oscillator crystals to overclock or underclock a CPU, it is done by changing clock values in BIOS. I guess that it only changes voltage applied to the oscillator crystal to change its oscillating frequency or switch between resonant modes. This section needs serious updating. — Preceding unsigned comment added by 92.52.23.13 (talk) 20:34, 14 October 2018 (UTC)
Clock Speed Graph is Incorrect
Although it comes from a fairly reputable source, the data in this graph for recent years is incorrect. Clock speed on microprocessors is still not above 6 Ghz even for the fastest production chips. Even with extreme overclocking, clock speed is not above 9Ghz. But this graph shows it reaching 28Ghz in 2016. Looking into your sources, it appears that Kurzweil is incorrect. The data he cites for 2001-2016 is not factual but a projection of the future from 2002: (Data from 2001–2016: ITRS, 2002 Update, On-Chip Local Clock in Table 4c: Performance and Package Chips: Frequency On-Chip Wiring Levels—Near-Term Years, p. 167. ) However, after 2002 the rate of increase in clock speed dropped to very low levels. Instead, chip makers focused on other ways of increasing effective speed, increasing the number of instructions processed per second by clever look-ahead techniques and parallelization.2601:148:200:91F0:D005:CB53:24EE:C25D (talk) 15:16, 6 December 2020 (UTC)Douglas Summers Stay
- Do you have a better / more recent source?
- For CPUs used in PCs the records appear to be here:[1] (8.72278 GHz). As far as I know no supercomputer reaches clock rates that high -- they just have thousands of cores. I wouldn't be surprised if there was some experimental system using a simple CPU that hits 10 or maybe even 20 Ghz, but a quick web search didn't find anything like that. --Guy Macon (talk) 18:17, 6 December 2020 (UTC)
This is okay, if you were to just take the top value for each year:
https://en.wikipedia.org/wiki/Microprocessor_chronology
These are the values I pulled from that chart:
2002 1.35 Ghz
2003 2.4
2005 3.2
2006 4.6
2007 4.7
2012 5.5 Ghz
by 2020 this 5.5 still hasn't been passed.
Of course you are right that with liquid cooling and experimental chips one could do better (I imagine if someone felt like it they could build a 30Ghz 8086 chip) but the values for the first part of the chart are commercial microprocessors so for the second part they should be, too.2601:148:200:91F0:D005:CB53:24EE:C25D (talk) 20:32, 6 December 2020 (UTC)Douglas Summers-Stay