Talk:Intel Atom
From Wikipedia, the free encyclopedia
"A 1.8 GHz Atom processor's single thread performance is equivalent to its predecessor Intel A100 " - how's the performance of a 1.8GHz cpu possibly equivalent to a sub GHz cpu? I think some clarification might be needed. Anton 24.201.100.166 (talk) 06:05, 26 April 2008 (UTC)
[edit] Merge with MID
I think this is nonsense - you could merge Pentium into Personal computer like this.--Kozuch (talk) 19:05, 8 March 2008 (UTC)
- Agreed. The Intel Atom is just one of many processors used in mobile internet devices... no reason for a merge. ǝɹʎℲxoɯ (contrib) 19:29, 8 March 2008 (UTC)
-
- Yeah. Keep the articles separate. They are different things. -- Imperator3733 (talk) 20:12, 8 March 2008 (UTC)
-
- Part of the problem is that the Mobile Internet Device page speaks not of mobile Internet devices, in the sense of mobile devices that can access the Internet, but about Intel's "Mobile Internet Device" platform. I guess that's why Talk:Mobile Internet Device asserts that "Intel is mentioned because they invented MIDs"; that page also says:
-
-
- A Mobile Internet Device (MID) is an Ultra-Mobile PC (UMPC) initiative raised by Intel for consumers and prosumers. Most other UMPCs are designed for mobile professionals.
-
-
- Unfortunately, "most" is not "all"; the iPhone, for example, weren't first targeted at mobile professionals. I don't know for whom Nokia designed the various Nokia Communicator devices, or whether the Nokia N800 - or, for that mater, the iPod touch - would not be considered sufficiently "mobile" as they support Wi-Fi (and, apparently, WiMAX in the case of the N800), but not any mobile phone Internet access mechanisms such as GPRS, EDGE, HSDPA, EV-DO, etc..
-
- I'd argue that it might be useful to have a "mobile Internet device" page that discusses mobile devices that provide Internet access in general (and would therefore go back at least to the Nokia 9000 - which, amusingly enough, had an x86 processor in it, albeit one from AMD) - and a page about Intel's "Mobile Internet Device" reference design, which should probably be called "Intel Mobile Internet Device" to indicate that it's specifically about Intel's design. One could perhaps then argue that the latter page should be merged into this page, but one could perhaps also argue that the chips and the system are different entities and deserve separate pages. Guy Harris (talk) 20:17, 8 March 2008 (UTC)
-
- I agree. It might be useful to have a generic page but the page about Intel specific MID platforms should not be merged in with pages that describe Intel MID processors (the A100 or the Intel Atom) like this one. If anything needs to be done about this my opinion is the MID page could stand to be renamed to Intel MID and a generic page created talking about MIDs in general beyond Intel MID platforms (and probably linking to the renamed Intel MID platform page). 64.122.14.55 (talk) 16:03, 10 March 2008 (UTC)
I agree, keep them seperate. Colinstu (talk) 18:46, 10 March 2008 (UTC)
Merge templates deleted.--Kozuch (talk) 23:22, 10 March 2008 (UTC)
[edit] Atom NOT a CISC processor!
Whoever wrote this page clearly knows nothing about x86. x86 is the INSTRUCTION SET. Internally since the Pentium Pro all x86 processors have been RISC, with decoders to turn CISC insructions into RISC operations. Thats really basic stuff. Read the excellent anadtech article. I'm on holiday, and do not have time to sort this nonsense out now. Sorry.
81.192.137.26 (talk) 15:03, 6 April 2008 (UTC)
- You are right, internally Atom is a RISC architecture based on micro-operations. On the other hand, you have to agree that IA-32 is an CISC instruction set. Thusly, Atom has to decode x86 instructions into micro-operations; a method that has remained standard to this day for x86 compatible processors. Enter ARM processors! They come with their own instruction set - which is RISC. So they do not need the translation step and save that die-space/power for more important things. Remember, backward compatibility always comes with some penalty. DanTheMan 84.58.151.228 (talk) 19:46, 16 April 2008 (UTC)
That backward compatibiltiy while having some costs helps ensure that DOS 1
still boots on the fastest IA32 processor out there, so it's not all bad. ;-) Regards, Andromeda451
