Talk:Transfers per second

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
WikiProject iconComputing Start‑class Low‑importance
WikiProject iconThis article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
StartThis article has been rated as Start-class on Wikipedia's content assessment scale.
 Low This article has been rated as Low-importance on the project's importance scale.
Taskforce icon
This article is supported by Computer hardware task force (assessed as Low-importance).

Discussion of multi-pumping needed?[edit]

The double data rate page mentions what double-pumping means (transfers on the rising and falling edge of the clock); should there be a page on multiple-pumping to which the megatransfer page refers, or should that stuff be moved to this page, or the reason why the MT/s rate is 2 or 4 times the clock speed otherwise given on this page?


What's a transfer in this context? Is it a byte? A packet? E.g. how much bytes would 10MT/s transfer (the verb) in 1 second? 10 million bytes, or maybe 10 million packets of e.g. 4KB making it 4 billion bytes? (talk) 15:35, 17 January 2008 (UTC)Reply[reply]

One bit per data line. Multiply by the bus width (e.g. 1 bit for PCIe x1, 16 bits for PCIe x16, 64 bits for SDRAM) to get bits per second. (In theory, multi-level signalling could encode more than 1 bit per bud, but that's not currently used in such contexts.) (talk) 14:12, 7 July 2009 (UTC)Reply[reply]


This sounds like a somewhat more specific and useless version of "baud", doesn't it? Something like x bauds = x symbols per second = x/(data pumping rate) transfers per second? (talk) 15:12, 17 October 2008 (UTC)Reply[reply]


Trying to reinvent the Hertz; no, not all frequencies are clock frequencies. Destroy with prejudice. (talk) 00:18, 10 February 2009 (UTC)Reply[reply]

There's a difference. Hertz is cycles per second, i.e. the signal returns to its initial state. Double-pumping works by observing that one data bit is at most half a cycle, so you can transmit two bits per cycle. (talk) 14:14, 7 July 2009 (UTC)Reply[reply]
I have to agree with this. Wikipedia has gotten away with retroactively renaming things with stupid substitutions for too long. The original definition was in Megahertz, it doesn't need to be changed. The same goes with the Megabyte; We don't need a consortium panel of idiots telling us what we need to call things. Ggigabitem (talk) 09:45, 9 August 2009 (UTC)Reply[reply]
Ditto. Superfluous idiocracy. What a waste of time. Let's invent a non-technical, unofficial (and officially unrecognised) term to buff some editor's ego. I clicked on a link named MT in the Ivy Bridge article with some apprehension, sure I was about to see yet another abhorrence similar to the delicious mibibyte et al. Yep; sure enough, here we are. Blitterbug 11:33, 14 April 2013 (UTC) — Preceding unsigned comment added by Blitterbug (talkcontribs)

Rewrite needed[edit]

The article mixes transfer and transfer rates. A rewrite is needed to clarify this and the comments above. I've tagged the lack of references and problems. Widefox (talk) 10:46, 17 July 2010 (UTC)Reply[reply]

Yes, I noticed the foo versus foo/s confusion too. I'll try to clear that up now. However I'm not sure what to do about the other problems mentioned above. —Preceding unsigned comment added by (talk) 09:58, 1 May 2011 (UTC)Reply[reply]
Ok, I believe I improved this, but it needs a lot more work. The part about the physical data rate being limited by the speed at which one can change the field down a wire seemed to me to be a garbled/confused version of Shannon's Law; I just deleted it, but I think a reference to Shannon's Law might be appropriate. —Preceding unsigned comment added by (talk) 10:29, 1 May 2011 (UTC)Reply[reply]