Talk:Intel Core 2

From Wikipedia, the free encyclopedia

This is the talk page for discussing improvements to the Intel Core 2 article.
Article policies

Contents

[edit] Wolfdale Released

Wolfdale is RELEASED... At least in South Korea and is now on Sale.

http://pcdory.co.kr/product/productView.php?nProdCode=615610

Not sure about US. Couldn't find it on Newegg. At anyrate, the article needs to be UPDATED. 121.125.222.222 (talk) 17:11, 10 January 2008 (UTC)

[edit] MT/s is incorrect.

-- None7 -I'd like to disagree, Intel,(and every single website I've almost been to), uses MHz, and if not, none of the motherboards out with an 800, 1066, 1333, or 1600 MHz FSB would be able to support any of these processors, I think it has to definitely be updated, along with Yorkfield and other "future" chips that already came out.

-- None1 - How many bits per second does a core 2 duo process in terms of 10^nth power?

If you're talking about bandwidth, it's 1066 × 106 T/s × 64 bit/T = 68,266 × 106 = 68.266 × 109 bit/s (~68 Gbit/s). 59.92.185.238 02:49, 4 March 2007 (UTC)

Alright, the Core 2 Duo (Conroe) does not have a 1066 MT/s front side bus. If you HAVE to measure the Core 2 Duo's (Conroe) front side bus speed in MT/s, then it's going to be 2133 MT/s. For every 1 MHz, there are 2 MT/s, according to Wikipedia's article Megatransfer. That said, no one uses MT/s -- not even Intel. If this is any kind of "free encyclopedia" for everyone, then the average Joe should (hopefully) be able to come on here and be able to see and understand it. "MHz" is a much more universally known and recognized term for the many types of processor speed and processor compenent (IE, L2 cache) speed

For example, the Core 2 Duo E6600 (Conroe) has a front side bus speed of 266 MHz, quad-pumped for an effective 1066 MHz. The clockspeed of the core is 2400 MHz, 266 MHz by a multiplier of nine. Imagine that with MT/s, now...? Yeah. Let's not. It's MHz, people.

--A Pickle 03:39, 23 December 2006 (UTC)

It isn't wrong, Core 2's FSB is 266MHz quad-pumped, 1066MT/s, but Intel don't use MT/s... Souldn't the article says the same as Intel? EduardoS 18:39, 13 January 2007 (UTC)

Actually Intel uses MT/s in its various official documents. For example, MT/s appears five times in "Intel 965 Express Chipset Family Datasheet", like this (page 27):

Supports Pentium 4 processor and Pentium D processor Front Side Bus (FSB) at the following frequency ranges:
— 533 MT/s (133 MHz)
— 800 MT/s (200 MHz)
— 1066 MT/s (266 MHz)

--renethx 11:35, 12 February 2007 (UTC)

Looks like this and other articles about Intel processors are still using MT/s. BUT BUT BUT, Intel does not use this terminology in their primary web articles nor their packaging. Should Wikipedia terminology match that of the manufacturer whose products are being discussed? See http://processorfinder.intel.com/List.aspx?ProcFam=2558&sSpec=&OrdCode= for an example. It is all in MHz. So whether you like it or not, you are doing the Wikipedia readers a disservice as they don't see the same terms as those used by Intel (or AMD or IBM or ...). 68.115.91.4 16:51, 28 February 2007 (UTC)

I think that we should use MHz, maybe in the format

1066 MT/s(266 MHz)

then mention something about quad-pumped by it.

[4/24/07] —Preceding unsigned comment added by 71.32.189.60 (talk • contribs) 21:03, April 24, 2007

[edit] Updating the Intel List of CPUs

Someone who knows enough about Core 2 needs to update this list [1] which says the Pentium 4 is the current most advanced processor sold by Intel. --71.113.167.60 17:54, 1 October 2006 (UTC)

Done. Guy Harris 17:48, 9 October 2006 (UTC)

[edit] Pricing?

Has there been any talk of pricing on any of the new Intel line of chips? Would be a good addition to the article. --ColinDoody 01:19, 12 May 2006 (UTC)

[edit] Announced yet?

I found a Computer Reseller News article saying that Intel is expected to announce on May 8th that they'll be calling Conroe and Merom "Intel Core 2", but I don't yet see any official announcement from Intel about it. Guy Harris 21:09, 7 May 2006 (UTC)

It's now official - see http://intel.com/products/processor/core2/index.htm. Guy Harris 17:24, 8 May 2006 (UTC)

[edit] Woodcrest?

Why no mention of the server chips? Frankie 20:02, 8 May 2006 (UTC)

Because they'll probably be called "Xeon" rather than "Core 2", so they won't be Intel Core 2 processors, even though they'll use the Intel Core Microarchitecture? Guy Harris 20:17, 8 May 2006 (UTC)

[edit] all-new?

The article says "an all-new CPU design called the Intel Core Microarchitecture". Is this really "all-new" (a marketing phrase if I ever heard one)? I had understood that it was an evolution of the P6/P-M line (Pentium Pro, Pentium II, Pentium III, Pentium M). DHR 20:28, 14 May 2006 (UTC)

Well, every architecture is going to be an evolution of something- nearly every Intel processor is an "evolution" of the original x86 design. However, the Core architecture is significantly different enough from the P6 and Netburst ones to warrant considering Core a new microarchitecture. Go read Intel Core Microarchitecture and if you want even more information, use Google. -- ßίζ·קּ‼ (talk | contribs) 05:52, 16 May 2006 (UTC)
And see also the Ars Technica article on the Core microarchitecture (referred to by the "References" section of the Intel Core Microarchitecture page). Guy Harris 06:10, 16 May 2006 (UTC)

[edit] Two articles!

There are two articles on this topic. I created this one: Intel Core 2 Duo. What to do? Theonlyedge 02:24, 30 May 2006 (UTC)

The best way would be to merge and redirect. Merge in whatever parts of Intel Core 2 Duo that aren't in this article, and then overwrite Intel Core 2 Duo with a redirect to this one (change the article so that the entire content is: #REDIRECT [[Intel Core 2]]). jgp 02:44, 30 May 2006 (UTC)
Yup. There used to be a "Core Duo" page separate from the "Intel Core" page, but "Core Duo" was, appropriately, merged into "Intel Core". The same should be done for Intel Core 2/Intel Core 2 Duo. Guy Harris 06:28, 30 May 2006 (UTC)

[edit] Get Woodcrest outta here!

If Woodcrest is going to be sold as Intel Xeon, then the section on it should be exported to the Xeon article. This article is purely for processors marketed as Intel Core 2, and whether or not differently-marketed products are related in design is inconsequential. -- ßίζ·קּ‼ (talk | contribs) 05:50, 1 June 2006 (UTC)

Yup. Other Xeons aren't on the pages for the corresponding non-Xeons. Guy Harris 07:06, 1 June 2006 (UTC)

[edit] Core 2 Design

I believe the Core 2 design is still a single core/dye approach correct? This isnt't mentioned Nil Einne 01:34, 2 June 2006 (UTC)

Er I guess not since they have a shared cache. In any case, this needs to be clarified, since it's a difference from their previous design. However I believe it still uses the FSB correct? Nil Einne 01:37, 2 June 2006 (UTC)

[edit] Allendale

Why isn't this "Allendale" mentioned? It's in the "List of Intel Core 2 microprocessors" article and should be here. (The E4200 isn't mentioned either) —Preceding unsigned comment added by 159.134.148.149 (talk • contribs)

Is there any talk of pricing and is it released? —The preceding unsigned comment was added by 69.142.156.226 (talk • contribs) 14:33, 29 June 2006 (UTC).

If by "it" you're referring to Allendale (as would be implied by putting the comment here), no, it's not released, and I haven't heard any talk of pricing.
If by "it" you're referring to Conroe and Merom (in which case you should have put the questions in the "Pricing" section at the top of this page), no, they're not released, but the article does mention pricing. Guy Harris 22:05, 29 June 2006 (UTC)

Allendale is a codename of the Conroe-Based processors having only 2MB of L2 Cache, including E4200, E6200 (which will be released in 4th quarter of 2006), E6300 and E6400 (which will be released on july 2006).

why is Allendale in future products? it's for sale RIGHT NOW http://www.alternate.de/html/productDetails.html?artno=HPGI38 --Jmke 17:08, 31 July 2006 (UTC)
Because, at one point, somebody stated that Allendale had 2 MiB of cache and an 800 MT/s FSB, and thus that the E6300 and E6400 weren't Allendales. Then I split the current and future products into separate sections, so that the future products section could have the "future product" template in it, rather than having a common section for both with a "future product" template in it referring to the future products (that combination confused at least two people who removed the "future product" template, presumably thinking it was referring to the current products), and then somebody else stated that Allendales have 2 MiB of cache but don't necessarily have an 800 MT/s FSB, and that the E6300 and E6400 are Allendales, but didn't bother adding an "Allendale" subsection to the "current products" section and putting information about the E6300 and E6400 there. Guy Harris 17:37, 31 July 2006 (UTC)
that last paragraph you wrote here should be somewhere in the article;) because without it, people will continue to refer to E6300/E6400 as Allendale find the future products template to be incorrect --Jmke 18:03, 31 July 2006 (UTC)
I have no idea how Intel uses the "Allendale" code name; i.e., I have no idea whether the "800 MT/s FSB, so the E6300/E6400 aren't Allendale" or "it's only the cache size, so the E6300/E6400 are Allendale" people are correct. The ideal way to resolve this would be something from Intel; I tried a Google search for "allendale site:intel.com", and all it found were some Intel Active Management Technology slides and they only mentioned it as part of a list of code names, they didn't say what made a processor a Conroe or an Allendale. A consensus among third-party sites would probably be the second best answer.
The future products template is correct in either case; it's the future processors section that's incorrect. I.e.:
  • if the E6300/E6400 are Allendales, the correct fix is to change the "Allendale" part of the future processors section to mention only the E4300, with the E6300/E6400 moved to a new "Allendale" section in "Current processor cores";
  • if the E6300/E6400 aren't Allendales, the correct fix is to change the "Allendale" part of the future processors section to mention only the E4300, with the E6300/E6400 moved to the "Conroe" section in "Current processor cores".
Note, BTW, that the E6300/E6400 are already mentioned in the "Conroe" section in "Current processor cores", not in any "Allendale" section.... Guy Harris 18:32, 31 July 2006 (UTC)
Allendales are apparently distinguished by their Core Stepping, which is LX (L2 at the moment) and indicates 2mb total (not just available) cache, whereas Conroes have a BX stepping (B1/B2). Some new e6400/e6300's in fact do have L2 stepping and the true 2mb cache. Hence, I think it would be appropriate to indicate that the latest version e6400/e6300 could be called Allendales, despite the 1066 MT/s FSB. 196.211.15.142 12:21, 7 March 2007 (UTC)
The pricing in the Allendale section seems wonky. It reads: "...$5224 USD each in quantities of 1000." I suspect that 5 is extraneous, since the other processor in the series is rated at only $183, but I don't know where to find the correct information. I suggest that someone in the know correct it.  :)
—ZorkFox (ষTalk) 03:14, 7 August 2006 (UTC)
You find them in the Intel press release for the Core 2 announcement, to which I recently redirected the "Intel's press release announcing Core 2" link on the Intel Core 2 page (the old link went to the announcement of the Intel Core 2 brand, which is rather less interesting now that the Intel Core 2 processors have been announced). That's where I got the correct prices; I just didn't type them correctly. :-) Guy Harris 06:18, 7 August 2006 (UTC)

[edit] Lga 775

Are these going to use the same socket as the current pentium 4's and if so are they compatible with the current mother boards? —Preceding unsigned comment added by 210.246.20.163 (talk • contribs)

Conroe will use LGA 775; Merom will use FCPGA6. Conroe will be compatible with at least some current motherboards, especially those based on the Intel 975X. jgp (T|C) 08:01, 8 July 2006 (UTC)

I don't think the 945G chipset support conroe (as at the time I'm writing this question), please someone, see this link. Intel Product Comparison Chart 203.91.132.17 18:44, 27 October 2006 (UTC)

[edit] Conflicting info

At first, about Merom, this article says "Intel has stated that the first version of Merom will be drop-in compatible with the current Core Duo platform"

And then later on, it says: "Merom requires a newer mobile platform, and it is incompatible with Napa (Core Solo/Duo Yonah) platform."

They can't both be true, and I'm pretty sure the first one is correct, so I'm going to copy/paste the first bit into the second bit.

69.93.96.202 18:02, 11 July 2006 (UTC)

The first is correct. http://www.google.com/search?q=merom+mac+mini However, the 2007 Merom rev2 (800 MT/s FSB) will require a new platform. Frankie

[edit] A blog isn't reliable. And look closely, it still says official availability is the 27th

I reverted the edits that wrongly claim that the official release date is the 23rd, and have restored the more reliable news site. The release date is not the 23rd, it is the 27th, as the blog itself explains "Intel will officially announce availability on July 27, 2006." Dionyseus 01:38, 14 July 2006 (UTC)

You persisting in calling DailyTech a blog shows just how far off the deep end you have gone and that anything you say is suspect. jgp (T|C) 01:40, 14 July 2006 (UTC)
Please assume good faith and stop personally attacking me. DailyTech is a blog. It doesn't even have an article in Wikipedia. Dionyseus 01:41, 14 July 2006 (UTC)
yes it does:) see http://en.wikipedia.org/wiki/Anandtech --84.192.117.172 21:08, 19 July 2006 (UTC)
Intel has officially stated July 27th is the release. If the news.com.com article is not sufficiently unambiguous, this should be. FWIW DailyTech is Anandtech's news component. Aluvus 01:43, 14 July 2006 (UTC)

[edit] Conroe on Newegg

Only been up for a few hours...apparently not July 27. [2]--Cirus206 06:31, 14 July 2006 (UTC)

The official launch is still July 27. The products are being shipped to stores but Intel has suggested that it not be released until the official launch date. This is common practice, and stores can either follow the recommendation or not, and Intel may or may not impose restrictions on the store for future products. Dionyseus 06:45, 14 July 2006 (UTC)
Ah, okay. I wasn't sure why it was up so early. --Cirus206 06:48, 14 July 2006 (UTC)
The fact that Newegg has already received shipments of the X6800 may be a good sign for those who want a Conroe the day it launches. Dionyseus 06:58, 14 July 2006 (UTC)
Only if you are stupid enough to pay an extra premium just to be first kid on the block with a Conroe. Wait until the end of the year, when Core 2 will overlap Pentium D production and prices will no longer be artificially increased because of lack of quantity. The largest part of the Core 2 CPUs go to OEM right now, leaving little for the resale market, those shops who do get their hands on early product (like Newegg) will surely add a premium, the X6800 is available for $1100 at their store...the price of two budget system or one complete high end system. --Jmke 08:39, 19 July 2006 (UTC)

[edit] Merom availability

"Conroe and "Merom," the mobile Core 2 Duo, will be formally launched on July 27 at an event at Intel's headquarters in Santa Clara, Calif." (from [3]) Does this mean that Merom will actually be available soon as well? Not in August like reported previously? --Paul1337 07:20, 14 July 2006 (UTC)

Merom in July is not confirmed, it is speculation. To quote news reports, launch of the Woodcrest consumer server chip in June, Montecito high-end server chip yesterday, Conroe desktop chip July 27, and Merom laptop chip soon afterwords, Otellini said. Frankie
Intel said on page 3 of their July 19 earnings release that "The Intel® Core(TM) 2 Duo processor for desktop PCs began shipping during the quarter ahead of its formal launch July 27 and has already set performance records across dozens of industry-standard PC performance tests. The mobile PC version of the Intel Core 2 Duo processor is also shipping now, one month ahead of schedule." Guy Harris 01:57, 21 July 2006 (UTC)
Intel plans to introduce details about Merom at the July 27th launch event, but the processor won't be available in retail systems until the end of August. [4] --69.81.35.5 03:50, 22 July 2006 (UTC)

[edit] 32 vs 64 bit

The article doesn't as far as I can tell mention whether the Core 2 Duo is a 32 or a 64 bit processor. Which is it?

It supports both. Dionyseus 07:01, 18 July 2006 (UTC)
I.e., the Core 2 (and Xeon 5100 series) processors are EM64T processors, meaning they can run in "legacy mode" as 32-bit processors, with a 32-bit operating system, or in "long mode" as 64-bit processor, with a 64-bit operating system. In "legacy mode", 32-bit applications can be run; in "long mode", both 32-bit and 64-bit applications can be run (if the operating system supports it, which most probably will, for binary compatibility). See the AMD64 article for more information.
I'd call them 64-bit processors; AMD64/EM64T, and most other 64-bit architectures, are designed to support 32-bit application code for the 32-bit version of the architectures on which they're based, for backwards compatibility, and many of them, including AMD64/EM64T, even support 32-bit OS kernel code. Guy Harris 08:51, 18 July 2006 (UTC)

[edit] Neutral point of View

I think that this article is lacking a neutral point of view. It is too much "This is much better than AMD's best" i think.

The problem is that every reference site and every test you can throw at it show that it is "much better than AMD's best" - help find some reference that discusses its flaws relative to AMD and i'd be happy to incorporate them into the article. But cold hard facts like test results are hard to water down. --Trödel 15:39, 24 July 2006 (UTC)
Well, the fairly new anandtech article on the memory performance of AM2 vs. Core 2 might fit the bill here. If somone wants to look through it and add the advantages and disadvantages of AMD's architecture vs. intel's, here's the article: http://www.anandtech.com/memory/showdoc.aspx?i=2800 . 17:23, 27 July 2006
NPOV doesn't mean we have to make out that both processors are equally good, it just means that we have to report them in an accurate, unbiased manner.. Saying "Tests have shown Core 2 to be noticeably faster than the Athlon 64 X2" would be NPOV, for example. On the other hand, saying "Core 2 is the best processor ever made and anyone who says otherwise is an idiot" is definitely a skewed POV. --DaveJB 11:51, 1 August 2006 (UTC)
Well I think we should point out that Athlon 64 is like what, 3 years old now, compared to the relatively new Intel Core 2, and even Athlon X2 is 1 year old, quite a long time in Moore's Law (if it still works). --antilived T | C 04:16, 13 September 2006 (UTC)
In that case, we should also mention that it wasn't fair comparing the Athlon 64 to any NetBurst based processors, seeing as the architecture was 3 years older.
Yeah, we should compare Core2Duo to a non-existing AMD CPU, which will be more powerful (and will be announced in 1 week or 1 year??). You can compare only existing CPUs. 91.127.145.77 19:19, 3 January 2007 (UTC)

The article is... Patetic, it's biased, unsourced, not something i expect to read at Wikipedia, compare Core 2 to Athlon 64 is ok, but on a well written article. EduardoS 18:52, 13 January 2007 (UTC)

I come at this as a long-time AMD shareholder and almost-exclusive purchaser of AMD CPUs for the last seven years... so you should be able to guess my bias. But I think the piece is fair enough, frankly. AMD's current offerings don't match the Core 2 in most respects, in terms of raw performance. I think the article could be enhanced with links to impartial evaluations of Core 2 vs. A64 and Opteron... At most it may be worth noting that server vendors and customers still have some reasons for preferring Opteron in actual deployed servers-- because very few run their servers at 100% usage, Opteron actually holds up pretty well performance-wise and draws less power than comparable Woodcrest servers in the real world. So maybe that needs to work its way into the discussion. See this report from Anandtech... But on the whole I don't think this is unfair, as long as it's kept up to date as new product releases come out.


Quoting sites showing benchmarks of Intel hardware performing better than AMD hardware isn't a bias issue. Core 2 family processors are new architecture, and it should be expected that they are an improvement on most everything that is on the market. <Shrug> just because the occasional fanboy gets out of hand and needs to be edited from time to time doesn't make this article non-neutral. --Taurius 23:06, 30 January 2007 (UTC)

[edit] DDR2 Memory Modules

Do you think a link to this article is worth mentioning when it comes down to stating that Core 2 does not need high speed DDR2 memory to work at its best? --Jmke 15:21, 1 August 2006 (UTC)

I've added it to a section called "DDR2 Modules" with information on running Conroe FSB and Memory in sync having a positive effect on performance. --Jmke 12:03, 2 August 2006 (UTC)

I would like to know where did you get that information about 1:1 ratio between Conroe FSB and memory improves the general performance. This AnandTech article states there isn't any negligible improvement when using different DDR2 memory. This are the links: Part I - http://www.anandtech.com/mb/showdoc.aspx?i=2810&p=1 . Part II - http://www.anandtech.com/mb/showdoc.aspx?i=2813&p=1

the whole point is that no matter what DDR2 speed modules you buy, the performance difference will be very small, negligible as one would say. 57.67.177.26 08:58, 18 August 2006 (UTC)

[edit] DRM on this cpu

AFAIK this processor is the first to integrate the evil DRM. This should be explained on the article. --Licurgo 20:58, 7 August 2006 (UTC)

Do you have any sources on this? What do you mean by DRM? Do you mean TPM or something else? --antilived T | C 04:08, 13 September 2006 (UTC)
it seems among some people drm is synonymous with tpm. where's the evidence? --gatoatigrado 01:47, 26 September 2006 (UTC)
TPM is a separate chip on the motherboard, so it has nothing to do with the main processor. Heck, since the original Pentium 3, we've had 'processor serial numbers' that could be used in relation to DRM to lock content to one computer. 65.100.35.138 09:57, 30 September 2006 (UTC)
For the record, PSN is long gone. IIRC, it was ditched with the Pentium 4 and removed from the P6 line with Tualatin. It was a pretty short-lived misfeature. jgp TC 10:05, 30 September 2006 (UTC)

Here is a link http://www.pcmag.com/article2/0,1895,1983842,00.asp

  1. 8 "Today, only business PCs have the TPM 1.2 protection chip built in. With Core 2 Duo, all PCs will be protected."

One must be very cautious when intel (or mainstream media) talks about tpr/drm. Well sorry for taking long, that was what I read.

And yes, the pentium 3 serial number was discarded on new intel cpu's, not to mention all the bios that disabled it or the software intel released to disable it.

Sorry if I was confusing drm with tpm, I'm not an expert on this issues but neither of them appeared on the article and they should be explained.

Licurgo 04:59, 18 October 2006 (UTC)

[edit] Requested move

Intel Core 2 → Core 2 – All other Intel microprocessor articles except for this article and Intel Core do not include "Intel" in the article title. Intel Core must include the word "Intel" in order to disambiguate it from Core, but Intel Core 2 -> Core 2 has no such restriction (Core 2 is presently a redirect to Intel Core 2). jgp TC 06:48, 12 September 2006 (UTC)

[edit] Survey

Add "* Support" or "* Oppose" followed by an optional one-sentence explanation, then sign your opinion with ~~~~

  • Support as nom. jgp TC 06:49, 12 September 2006 (UTC)
  • Oppose. The registered trademark for this product line is "Intel Core", whereas the Pentium brand did not mandate the word Intel. Also, the reason stated in the request is false: Intel 8086, Intel 8088, Intel 80286, Intel 80386, Intel 80486. -- Frankie 10:40, 12 September 2006 (UTC)
  • Actually, the 80*86 processors must be disambiguated. According to WP:MOSNUM, "A page title that is just a positive whole number is always a year.". Thus, for example, 8086 must refer to the year 8086 CE (even though I doubt Wikipedia will still be around then, the convention holds). jgp TC 20:51, 13 September 2006 (UTC)
Whatever the reason, your original statement ("All other Intel microprocessor articles"...) is incorrect. And IMO the trademark issue is more significant anyways. Frankie
  • Support: Per nom. —Wknight94 (talk) 02:34, 20 September 2006 (UTC)
  • Oppose, as Frankie said. The words "core 2" are too vague to be related to a product, whereas "pentium" is obviously a trademark, not a word. --gatoatigrado 01:42, 26 September 2006 (UTC)
oh, i didn't see peter already mentioned it. --gatoatigrado 01:45, 26 September 2006 (UTC)
  • Oppose Only "Core 2" is too vague, and I support putting manufacturer name in front of its products. --Will74205 06:41, 27 September 2006 (UTC)
  • Oppose. Just Core 2 is too confusing, and "Intel" is more an organic part of the name than Pentium and the earlier processors. (I don't see how it matters all that much given that we have redirects). Herostratus 18:28, 28 September 2006 (UTC)
  • Oppose. In agreement with Frankie and also Core 2 is too vague and not well known enough on it's own (for the time being anyway). -- Borb 11:46, 29 September 2006 (UTC)
  • Oppose. Core 2 is really too vague. First time I went on this article it was buy simply typing "Intel_Core" and following the link to Intel_Core_2. Intel Core is a tradmark while core could be the core of any processor. 70.50.181.110 18:29, 5 January 2007 (UTC)

[edit] Discussion

  • Relisted. --Dijxtra 10:28, 20 September 2006 (UTC)
  • If this matters, maybe the title should be "Core 2 (Intel)"; unlike Pentium 2, Core 2 can be somewhat ambiguous. Peter O. (Talk) 20:36, 21 September 2006 (UTC)
  • Move has been declined. Teke (talk) 05:21, 29 September 2006 (UTC)

[edit] Core 2 and Athlon 64 in same Generation?

Core 2 has the lowest power consumption of any desktop chip this generation, including both Prescott's TDP of 130 W and San Diego's TDP of 89 W. Isn't Core 2 the successor to the the position previously held by Prescott? Then how can Core 2 be in the same generation to Prescott or Athlon 64? --antilived T | C 04:02, 13 September 2006 (UTC)

Prescott is being assembled as well, so it is the same generation, or define "generation" :) 91.127.145.77 19:24, 3 January 2007 (UTC)

I think it should be changed to "at the date of its release", and changed to "consumer desktop ix86 architecture". The Cell microprocessor takes up 40 watts of power and kills the core 2. --gatoatigrado 01:44, 26 September 2006 (UTC)

[edit] No SSSE3

The introduction to the article mentions SSSE3, which of course is a typo. Made the necessary changes. —The preceding unsigned comment was added by Vliktor (talk • contribs) 08:21, 29 September 2006 (UTC).

No, not a typo. See the top of page 5 of this Intel white paper, which says "A further advancement, Supplemental SSE3, is now available in Intel Core microarchitecture. Included in Intel(R) Xeon(R) 5100 processors (server and workstation) and the 350 Intel Core 2 Duo processors (notebook and desktop) processors.". Guy Harris 16:18, 29 September 2006 (UTC)

[edit] Sorry, just creepy timing.

I was reading this article, and came upon the 'Kentsfield' section. I noticed that it still called them 'Core 2 Quadro', so I hit the section 'edit' to correct it. The text in the edit box had already been edited! So in the time it took me to read the article, someone else had already fixed it. I love Wikipedia. (Sorry for the off-topic rant, but I just found it cool.) 65.100.35.138 10:00, 30 September 2006 (UTC)

It means so many people are reading this wikipedia page. 210.246.73.12 18:38, 13 April 2007 (UTC)

[edit] Power consumption

(AMD codenamed processor known as) San Diego's TDP of 89 W. - (with the possible exception of the AMD "energy efficient" line of processors announced around may of 2006 which include 65 W versions and even 35 W versions of some AMD processors)

I've removed the misleading bit on power consumption cf AMD. AMD AM2 Orleans which was released a while before the Core 2 has a TDP of 62W. The 939 Venice which was released way way before the Core 2 has a TDP of 67W so while the TDP of the Core 2 is still slightly lower it's not the 89W that was used for comparison. While it might be fairer to compare X2s to the Core 2, the article simply talked about desktop chips not desktop dual cores. Even the bit about EE AMDs was poorly written. The EE AMDs have been available for a while (although they did take a while to become available from launch) yet the article just talked about them like they were irrelevant. Finally as has been discussed in many places, TDP cannot be used to compare power consumption for many reasons including the fact that AMD and Intel define TDP differently. Nil Einne 09:53, 6 October 2006 (UTC)

As an additional note, TDP is merely the estimated maximum using whatever methodology, not necessarily the actual usage. Furthermore, TDP is listed for usually the most power-hungry processor in the whole family, so the lower speed related processors will generally have lower TDP. To summarize, TDP is calculated totally differently between AMD and Intel and is generally an overestimate. However, when comparing chips of a single company, lower TDP can probably be interpreted as lower power consumption, though how much is anybody's guess. --Taurius 22:57, 30 January 2007 (UTC)

[edit] Removing reference to iMac/Macbook

I removed the fact that iMac use the merom CPU (since it's not the only desktop doing so) as well as rumors (rumors on wikipedia, WTF!) about a coming Macbook Pro/Macbook with a C2D.

Mac-users/zealots, please think twice before adding any informations related to Apple/Mac as, to put it bluntly, most of the world don't give a shit about what Apple do or do not. —Preceding unsigned comment added by 81.233.73.177 (talk • contribs)


[edit] Fan speed Error

I don't know if it deserves inclusion but I have one and did a bit of research and for some motherboards the core 2 duo causes a fan speed error since the idle fanspeed doesn't have to be as high as a P4 chip. The BIOS on the motherboard is expectign 20%-30% higher fanspeed and gives a warning. I hear this is common for Asus boards.

Sounds like the old error that affects any slow CPU fan, although if this is happening with the heatsink and fan included with the CPU then it's probably worth mentioning.

[edit] CPU-Z version 1.37 & E6400

Processors Information

Processor 1 (ID = 0) Number of cores 2 Number of threads 2 (max 2) Name Intel Core 2 Duo E6400 Codename Allendale Specification Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz Package Socket 775 LGA (platform ID = 0h) CPUID 6.F.6 Extended CPUID 6.F Core Stepping B2 Technology 65 nm Core Speed 2133.2 MHz (8.0 x 266.7 MHz) Rated Bus speed 1066.6 MHz Stock frequency 2133 MHz Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, EM64T L1 Data cache 2 x 32 KBytes, 8-way set associative, 64-byte line size L1 Instruction cache 2 x 32 KBytes, 8-way set associative, 64-byte line size L2 cache 2048 KBytes, 8-way set associative, 64-byte line size FID/VID Control yes FID range 6.0x - 8.0x max VID 1.213V

?

CPU-Z 1.38 has been fixed and now correctly reports the E6300 and E6400 as Conroes. -- Borb 17:53, 25 November 2006 (UTC)

[edit] kentsfield Quad core proccesor

I have a question, i found info on this proccesor in the Dell´s website. This technology is included in the XPS 710 (Intel® Core™2 Extreme QX6700 (8MB L2 Cache,2.66GHz,1066 FSB)) But i want to know one thing... it says that the proccesor has 4 cores in one proccesor(all of them clocked at 2.66 GHZ), Are we talking about a total proccesing power of 10.64 Ghz, or what? I really don´t understand this and I want to know, because i´m planning to buy a new PC for gaming. Can someone explain me the way this proccesor works and it´s power??? is this proccesor is as powerful as they say it is??????? Somebody please, explain me this.


This goes completely off topic but... If you haven't built it already, you should know that there are currently only a handful of games which can take advantage of a multi-core setup, it is mostly useful if you are running multiple programs which have heavy CPU usage: CD/DVD Ripping and compressing large files for example. If you are building a gaming system you want to focus more on your RAM and video card (personally I would say that if you have enough money for the kentsfield, you should go with the AMD socket F FX series, but then again I've always favored AMD).
As far as I know, neither Intel or AMD has come up with a way to combine the processing power of multiple cores together efficiently. The Kentsfield CPUs run at 4x2.66 Ghz, and the total load on the system is divided between the four cores. Most games will only be able to use one core at a time so you might as well have a single 2.66 Ghz processor. I don't know whether that is clear or not, perhaps someone else could phrase it differently? -RK 66.72.64.57 22:28, 7 January 2007 (UTC)

Further note: computer programs run in what are called threads, which are more or less a list of tasks for the processor to do. The speed of the processor is the rate at which it can perform those tasks. Most programs, as RK mentioned, are single-threaded, so only one processor core can work on that list at a time. Very few games (if any) are multi-threaded, so the increase in performance seen with the new processors is mostly due to faster speeds and better instruction sets (which enable tasks to be grouped to be performed more efficiently), as well as more complicated things that I won't get into here (maybe I'll make a new page :P). Does that answer your question? --Taurius 22:50, 30 January 2007 (UTC)

There is a multi-core article. It could stand some expansion in this department. -RK 66.244.123.101 17:18, 12 February 2007 (UTC)

[edit] Core Duo Vs. Core 2 Duo

Is there any difference between the above two except for the obvious difference of "2" added to the latter?

Is there any difference between the Pentium and the Pentium II other than the obvious difference of "II" added to the latter?
The answer, in both cases, is "yes". For one thing, the Core 2 supports the 64-bit EM64T instruction set, as well as SSSE3, while the Core doesn't. In addition, the internal design of the Pentium and Pentium II are very different; the internal design of the Core (which had the Pentium M derivative of the Intel P6 microarchitecture as its internal design) and the Core 2 (which has the Intel Core microarchitecture as its internal design) are also different, although not as different as the Pentium and Pentium II. Guy Harris 11:25, 30 November 2006 (UTC)

[edit] Re: Dualcores

Ok, I'm slightly confused. Let's take the E6400 (Core 2 Duo, 2.16 ghz).

Does each core run at 2.16 GHZ or in unison is their standard clock speed 2.16 ghz, hence their individual speed in 1.08 ghz? Logical2u 22:37, 3 December 2006 (UTC)

An E6400 is a single chip with two processors on it; both processors on the chip run at 2.16 GHz. Guy Harris 23:22, 3 December 2006 (UTC)

[edit] dont use pwned.

I think the phrase "pwned the athlon64" should be taken out. its just unprofessional. Discussion of why it is better than athlon64 is fine, but pwned is just vague and sounds opinionated. —The preceding unsigned comment was added by 76.185.114.198 (talk) 19:25, 8 December 2006 (UTC).

Already reverted, it was vandalism. — Aluvus t/c 22:29, 8 December 2006 (UTC)

[edit] Overclocking and comparisons against Athlon 64

Why are there so many references to overclocking and Core 2's performance vs. A64? I think they are completely unnessessary - it even sounds a little fanboyish. I don't think the average joe would care how well his CPU overclocks. Not only are some of the statements uncited but I would imagine that most Core 2 owners didn't buy their CPUs to see how much they could overclock, or how long it takes to calculate pi to 1M decimal points. --MaXim 05:24, 13 December 2006 (UTC)

These chips are well known for their amazing overclocking ability, which may not mean much to an average user purchasing a pre made computer based on the Core 2 platform, but for the large enthusiast market it is a very good attribute to have. A lot of online reviews will include overclocking results when reviewing CPU's. -- 58.107.223.215 00:46, 31 December 2006 (UTC)
The time it takes to calculate 1M PI shows the CPU power, but I agree, it tells nothing to an average person, not interested in computers. The same with overclocking. The question is whether we should provide these information... I think it is better to have an unknown statement (for an average user) than to have no statement (which some people may be interested in).91.127.145.77 19:31, 3 January 2007 (UTC)

This may be true, but we need to explain the significanse of overclocking, either here or in a separate article. Significance:

  • The "overclocker culture" is a set of people who like to explore the limits of a system.
  • Historically, a system that performs well when overclocked is well-designed and relatively better-behaved when NOT overclocked. The design is not pushing its limits. By contrast, as system thts behaves poorly when overclocked probably has much less margin in at least one dimension (e.g., ambient temperature or voltage fluctuation) than a system that can be overclocked.

-Arch dude 01:39, 3 January 2007 (UTC)

There appears to already be a perfectly good article on Overclocking availble. -- Rcrowley7 04:12, 9 February 2007 (UTC)

[edit] MT/s unnecessary?

Intel FSB speeds are generally still rated by MHz even though the the conroe and kentsfield core 2 processors currently use a 266mhz quad-pumped bus and the meron chips use a 166mhz quad-pumped bus. - just a thought, RK —The preceding unsigned comment was added by 66.244.123.101 (talk) 17:02, 14 December 2006 (UTC).

I like the idea of specifying MT/s as opposed to MHz -- it's more accurate and less ambiguous. Personally I've been using things like "266 MHz QDR", which is fairly specific. 59.92.138.0 12:55, 20 February 2007 (UTC)

[edit] A note on fab locations

I'm sure I'm just being picky, but I thought wikipedia was created to further the global knowledge pool and otherwise benefit the greater good. It's no big deal, but it seems a tad ignorant to give the locations of US fabs only by their state - when the locations of fabs located in other countries are given by their country, and not the province or locality. This has been the case for so long that people don't seem to pay mind to it, but by the same token as 'pwnage' seems inappropriate, it seems inappropriate to assume that US-located facilities need no mention of the fact that they are in the US, as if everyone automatically knows, or 'should' know. Where I live, at least, the states of the U.S. are not part of the school curriculum.

What do others think? Discuss :) --Growly 15:34, 23 December 2006 (UTC)

  • "Where I live, at least, the states of the U.S. are not part of the school curriculum.". Oh, man! Where do you live? In the USA? ;^) Giving the locations by state helps North Americans learn their geography, in case they didn't have time to at school, between "Why we should rule the world" and "Firearms use for kids" classes. — Isilanes 17:03, 23 December 2006 (UTC)
  • Be bold. You do not need anyone's permission to make this change: just do it. Get on Google and find the city locations for the fabs in the US and in foreign countries, and then edit this article. Alternatively, create a new article the lists fabs, put the locations in that article, and reference the article from here, Or create a whole new set of articles, one per fab: a fab is at least as significant as (say) a naval ship or a CD, as Wikipedia has articles for individual ships and CDs. -Arch dude 02:00, 3 January 2007 (UTC)
  • I added the County in Ireland that Intel's Fab 24-2 is located in, in the introduction section. Flash fact: Intel's Fab 24-4 in Leixlip in Co. Kildare in Ireland is the largest fabrication plant outside of the US.

[edit] chip core

is the chip core built off the Pentium 3 core? —The preceding unsigned comment was added by Falcon866 (talk • contribs) 02:01, 15 January 2007 (UTC).

  • Not as such, no. The Core 2 microarchitecture is built off lessons learned from the Pentium M but is mostly a new design.
Well, the Pentium M was based off the P3... —The preceding unsigned comment was added by Cooldude7273 (talk • contribs) 03:17, 21 January 2007 (UTC).

[edit] Well respected??

From the Section titled "Conroe":

"According to well respected reviews, the E6300 and E6400 only suffers on average 3.5% because of the smaller L2 cache size."

Surely if this is to be an unbiased article, this needs some kind of citation?? CheShA

Looks like Ploxhoi has changed the wording and added references. 59.92.138.0 12:52, 20 February 2007 (UTC)

[edit] Quad-core drop in upgrade

This statement is incorrect, but I don't know how to reword it:

  • "What this also means is that it will be easy for Core 2 Duo early adopters to upgrade their LGA775 platform by just dropping in a quad-core Kentsfield, as has been demonstrated by AnandTech when they dropped in two Core 2 Quad processors in place of two Woodcrests in a Mac Pro."
  • The MacPro actually has two LGA-771 sockets, not 2 LGA-775 sockets. He put in early samples of Intel's quadcore Xeons, not Core 2 Quad. I beleive the information is still correct regarding LGA-775, but this citation is incorrect.

http://anandtech.com/mac/showdoc.aspx?i=2832&p=6 —The preceding unsigned comment was added by 17.232.47.219 (talk • contribs) 13:28, January 26, 2007 (UTC)

A better citation might be the Anandtech article specifically about Kentsfield motherboard compatibility. — Aluvus t/c 23:07, 26 January 2007 (UTC)

[edit] Neutrality? and layout overhaul

Has this page been recently reevaluated for neutrality? Just wondering, because it seems to be okay now that most of the AMD bashing, slang, etc. have been removed. Also, the layout needs to be overhauled, because there are some future products that are no longer future. Some information probably needs to be reverified as well, e.g. Allendale, Conroe-L and some other items.

Taurius 22:28, 30 January 2007 (UTC)

Let's put it to a vote then: should the neutrality notice on this article be removed? My vote is yes. Taurius 15:14, 13 February 2007 (UTC)

I've gone through and updated information on Conroe, Allendale, Kentsfield and Conroe-L and added references for the updates (including pricing and availability.) The text looks fairly neutral to me. 59.92.138.0 12:35, 20 February 2007 (UTC)

[edit] Inaccuracy in article

The article says, "Core 2 was to mark the retirement of Intel's Pentium brand name that had been used since 1993". How can this be, when the Core 2 is the successor to the Core? It was the original Intel Core which marked the retirement of Intel's Pentium brand name. In fact, the original Core was just a rebranded Pentium M. TwinTurboZ 04:25, 16 March 2007 (UTC)

Intel's desktop products continued to carry the Pentium brand when Core Solo and Core Duo products had taken over Intel's laptop lineup. Core Solo/Duo never crossed over to the desktop; Core 2 Duo did, and in so doing it displaced the Pentium products. Additionally, Core Solo/Duo is not just a rebranded Pentium M. Performance of Core products is better, they are on a smaller process, and of course some are dual core. — Aluvus t/c 04:54, 16 March 2007 (UTC)

[edit] processor list of Core 2

Hi can we remove the tables of future Core 2 Processor and their info? Its getting more and more techno geek language and less user-friendly(read = noob).

We already have another page where the tables of all the present Core 2 processors are at,so maybe we can put the future ones there as well.

There is some dispute as to where the future Intel processors section would go, as the charts of Intel Core 2 on those pages only list available processors not future ones.Coldpower27 16:06, 24 March 2007 (UTC)

i think the list page would be a more apropriet place for the tables. maybe we should take a vote. also the layout of the tables on this page is slightly inconsistant but i cant be botherd to change it. Nicoli nicolivich 17:16, 24 March 2007 (UTC)

[edit] Doubt about this procesor

Hi there, is it true that the Core 2 duo Conroe E6600 emulates to have 2 cores? Somebody told me that this proccesor doesn´t have 2 physical cores and that they r emulated. I ask this because im planning to buy a new PC for gaming and multimedia (specially HD content), and i dunno if someone can reccomend me some cool components for my future system. By the way, is the nVidia GeForce 7300GT at 512Mb a good graphic card? or i should search for another card?Big butt gangsta 19:07, 5 May 2007 (UTC)

Well that somebody doesn't know what they are talking about, all Dual Cores on the market are 2 physical cores, none of them are emulated. The 7300 GT is a budget graphics card. Get something from the 8600 or 7900 line minimum. Coldpower27 19:21, 5 May 2007 (UTC)
Which video cards and motherboards would you reccomend me? I want to use the core 2 duo E6600Big butt gangsta 21:23, 5 May 2007 (UTC)
Your question belongs on a discussion forum like this one. — Aluvus t/c 00:13, 6 May 2007 (UTC)

[edit] Matched processor and RAM ratings

Hi! If I get E4300 (800MHz FSB) together with two dual-channel DDR2-800 modules in a G965-based board (ASUS P5B-VM), will they work in dual channel mode to achieve a nice throughput of 12.8 GBps, or will they only work in single channel mode to achieve merely 6.4 GBps? The table on the page doesn't address this. Essentially, what is the maximum memory speed this processor and chipset can handle? Thanks, Constantine. MureninC 15:25, 12 May 2007 (UTC)

You will get 12.8Gb/s but the 800FSB will only use 6.4Gb/s there is nothing wrong with over saturating the memory bus with more then enough bandwidth. But with an IGP motherboard such as G965 some of the memory bandwidth is used for graphics as long as you don't have a discrete PCI-E graphic card added in so more then enough is a good thing. Coldpower27 14:48, 13 May 2007 (UTC)
Thanks. I've got my system working, and I'm running memtest86+ v1.70, and it shows that the memory speed is only 3070MB/s. Why is it more than 2 times slower that it is expected to be? When the system booted, it did say that it was running "PC2-6400 Dual Channel Interleaved". Thanks, MureninC 21:23, 16 May 2007 (UTC)
Speed showing in memtst86 is inaccurate, use Everest/AIDA32 memory benchmark. Also, you are limited by the FSB of the CPU.
Are there any UNIX (OpenBSD/FreeBSD) utilities to test the raw read/write memory speed? I don't run win32, so I cannot verify the actual speed with Everest. Regards, MureninC 16:24, 19 May 2007 (UTC)
This disscussion is not about you, it's about the Wiki article, please stick with the topic.--68.147.37.49 00:58, 7 July 2007 (UTC)

[edit] Price drop dates?

According to the article, Intel is dropping prices on June 22. I see July 22 from other sources, however, e.g. [5]. —DemonThing talk 22:05, 18 May 2007 (UTC)

[edit] Overclocking to 4 GHz

The way this article currently reads is that 4 GHz speeds are easily achievable on most E6600/E6700 chips, this is absolute nonsense. Many chips (including those in 2006 early 2007) top out around 3.2-3.6 GHz and require high vcore to get there. My own chip from 2006 needs vcore of 1.3825 to make the "tiny" overclock to 3 GHz. Just because some cherry picked engineering samples can overclock very high doesn't translate to the typical Conroe. In fact the entire article reads this badly, such as "At full load the X6800 does not exceed 45 °C (113 °F)" - whose X6800 exactly? Does the acronym YMMV mean anything to anyone? It makes it sound like every chip is identical and everyone's thermal solution is identical. Every chip will have its own overclocking and thermal peculiarities, but this is not reflected in the article with sweeping statements about what a particular chip model will or won't do.Darkfalz 12:19, 19 May 2007 (UTC)

True enough, I think overclocking numbers should not be posted, as different batches have different capabilites. It'll only serve to annoy people. 85.19.140.9 16:10, 11 July 2007 (UTC)

[edit] Penryn?

I thought Penryn was the codename for the 45 nm shrink so why is it the codename for the mobile version of the 45 nm shrink as well? in this article?--Sat84 04:05, 20 May 2007 (UTC)

Penryn is the codename for both the shrink as a whole and the mobile dual-core, actually. The desktop dual-core version is codenamed Wolfdale, the desktop quad-core is codenamed Yorkfield, while the quad-core Xeon version is codenamed Harpertown. See here. Suigi 05:21, 20 May 2007 (UTC)
Ok thanks for the info but it's a little confusing that they had the same codename. I was about to change it because on the intel page it said it was for the 45 nm shrink--Sat84 05:42, 20 May 2007 (UTC)
It's no different to "Barcelona" which refers to the quad core sever chip, but most people are using to refer to the entire K10 range Kuma/Phenom etc.

[edit] Use of code names

I understand the use of "code names" prior to processor launched, but after a product is launched people know a product by its marketed name, not a code name. I suggest reducing the use of "code names" after a product has been launched. The masses looking for processor information don't care about "code names" and makes the article more confusing. Daniel.Cardenas 14:33, 23 May 2007 (UTC)

Unfortunately, due to Intel's marketing people assigning their products offical names which are often either vague, confusing or unwieldy, usually it is necessary to use the appropriate codename in order to identify concisely a specific variant of an Intel processor. Of course, it's not just Intel products that suffer from this - Apple are another culprit who spring to mind. Letdorf 11:01, 16 July 2007 (UTC).

[edit] Remove list of future processor tables

[edit] Survey

Add "* Support" or "* Oppose" followed by an optional one-sentence explanation, then sign your opinion with ~~~~

  • Support The paragraphs can be resident on this page, but the tables should reside in the list of future processor page. It seems redundant and out of place here. Kyberesh 15:40, 11 June 2007 (UTC)

[edit] Core 2 is NOT the same as 2 cores!

The article starts off with saying:

"The Core 2 brand refers to 64-bit dual-core CPUs"


That is false, Core 2 is next type after Core 1, or "Core" as it was called. Like Pentium II came after Pentium. They should have named it better I guess. After Core 2 comes the indicator of how many cores there is - Duo (2) or Quad (4). —The preceding unsigned comment was added by 84.189.114.33 (talk • contribs) 19:53, 10 July 2007 (UTC).

I've updated the page not to say that Core 2 refers to dual-core CPUs; it later talks about dual-core and quad-core CPUs, so that takes care of the types of CPU. Guy Harris 08:34, 11 July 2007 (UTC)

[edit] Way to overclocker biased

no regular person wants to hear the ravings of a newbie who just bought a DS3 and an E6600. we need more generic audience friendly info.

also i deleted the following part from the 2nd conroe paragraph as it is wrong: These impressive speeds are possible due in part to the fact Conroe series chips have an adjustable multiplier (previous Intel Processors lacked this, with the exception of very high end Extreme Edition™ models). The Conroe multiplier can be adjusted from x6 to x9 — allowing higher clock rates with the ability to strap the MCH (northbridge) down for stability depending on other hardware.

Any conroe can have its multiplier adjusted downward, to a minimum of six, but cannot go higher then its stock multiplier. The X6800 has only the bios's mulitplier imput as a limiting factor.

also MCH is the name of the northbridge only on intel chipsets, its SPP on Nvidia chipsets, and RD (I think, that bridge spicifically is called RD600) on AMD's chipset(s?).

Do we even need to go into how well Core 2 overclocks, is this really encylopedic information?? I think the exact values should be removed, though mentioning it in the general sense is tolerable. Coldpower27 22:48, 11 July 2007 (UTC)

[edit] Quad Core double performance of Dual Core

http://techreport.com/reviews/2007q2/core2duo-e6750/index.x?pg=12 It is rare, but it can happen if the FSB is not the bottleneck. That is why I used the word rarely user 151.202.72.24. As well faster low latency RAM benefits all processors and it is not limited to just Intel's Quad Cores but Intel's Dual Cores as well, hence why the lack of need to mention it.Coldpower27 20:18, 12 July 2007 (UTC)

Reference link number 33 gives a 404. Maybe the register has made a site redesign and changed permalinks.

[edit] Number of Transistors?

In the olden days, it was common to report the number of transistors on the die. Can we get figures like that anymore? I'd like to know how fast that number is ramping up as the generations change. No need for exact numbers, I'd like like ballpark figures. 207.154.79.131 07:40, 19 September 2007 (UTC)

BalusC says that the Conroe/Merom 4 MB die is 144 mm^2 and has 298.0 million transistors, while the Allendale/Merom-2M die is 111 mm^2 and has 167.0 million transistors. (http://web.archive.org/web/20070629171523/http://balusc.xs4all.nl/srv/har-cpu-int-c2.php) Pgk1 03:30, 25 September 2007 (UTC)

Holly Lord! How do they get to put 300 million transistors in 144 square mmm. Trully impresive 190.189.10.235 (talk) 20:34, 26 April 2008 (UTC)
Nothing special as most of the transies are Cache SRAM. Intel is known to have very high packable (=small die-size) Cache cells. --Denniss (talk) 04:05, 27 April 2008 (UTC)

[edit] Updated summary for Merom-2M/Allendale die and the Core 2 Duo ULV CPUs

I added the information about the L2 and M0 stepping Merom-2M die and the Core 2 Duo ULV CPUs to the "Merom" section. I also updated the main summary to include the lowest speed of the ULV CPU (1.06 GHz, U7500) as well as adding the micro-FCBGA interface for the C2D LV, ULV, and some T5000 series CPUs and updated the core names to include the 2 MB dies' core names. Pgk1 03:29, 25 September 2007 (UTC)

[edit] SSE4

According to this (http://www.tomshardware.com/2007/10/29/intel_penryn_4ghz_with_air_cooling/page4.html) SSE4 is supported in VirtualDub already. Update time for "SSE4 is not yet supported in any app"! Just a little reminder! 76.227.11.127 16:00, 12 November 2007 (UTC)

[edit] How can 2 (or 4) processors share the same memory?

I was hoping to find the answer in this article, but no luck. Seems to me that the processors would be constantly "butting heads" trying to access memory at the same time (thus forcing one of the processors to wait & stop execution). How does Intel get around this problem? - Theaveng (talk) 20:17, 21 November 2007 (UTC)

See SMP "Access to RAM is serialized" or "Non-Uniform_Memory_Access". The "butting heads" problem has to be handled in software (typically the OS) but it is only really a problem for writing. Accessing the same memory does not really mean, they are accessing the same memory ranges at the same time but rather they share the same memory in contrast to having dedicated memory each. The advantage is that you do not have to transfer between the processors which would require more bandwidth and be slower than synchronisation via semaphores or similar mechanisms. --217.87.87.117 (talk) 19:50, 22 November 2007 (UTC)

[edit] Clovertown missing

Where are the core 2 xeon clovertown cpu specs? —Preceding unsigned comment added by Spydercanopus (talk • contribs) 02:31, 23 November 2007 (UTC)

That is because Clovertown is a Xeon and is therefore discussed in the Intel Xeon article instead.Coldpower27 (talk) 20:14, 24 November 2007 (UTC)

[edit] Core 2 Octo

With the advance of the 45nm fabrication process, and the resultant decrease in heat emission by the CPU's transistors, can 8-core 45nm Core 2 CPUs become a reality? If Intel decides to release some Core 2 Octo CPUs, will they be compatible with the current chipsets (Intel Bearlake chipsets)? Will they require special voltage regulators that may not be present on current LGA775 motherboards?

Or will 8-core CPUs be realized only on the post Core 2 architecture (Nehalem, if I'm correct here)? —Preceding unsigned comment added by Hellcat fighter (talk • contribs) 16:14, 8 December 2007 (UTC)

[edit] Possible inaccuracy about Conroe and Allendale cores

The article states that "The currently available Core 2 Duo E4300 only uses an Allendale core, released on January 21, 2007."

Yet, I appear to have a piece of hardware on my hands that defies this statement. —Preceding unsigned comment added by 67.55.8.197 (talk) 22:21, 8 January 2008 (UTC)

All L2 Stepping cores are indeed Allendale SKU's, CPU-Z sometimes doesn't read accurately. L2 or M0 Steppings are Allendale while B2, G0 steppings are Conroe. That is more accurate and CPU-Z won't screw that up.Coldpower27 (talk) 16:18, 19 January 2008 (UTC)

[edit] MacWorld

What is the processor announced by Intel at MacWorld today that was 4MB, 1.6 or 1.8GHz Core 2 and shrunk 60%? Nja247 (talk • contribs) 21:43, 15 January 2008 (UTC)

A rather special Merom. — Aluvus t/c 00:46, 16 January 2008 (UTC)
It is a Merom core with a TDP of 20W so between Low voltage and Standard voltage, which uses the pacakaging technology designed for the SFF Montevina platform due to come out in the middle of this year. Originally it was to be used only with Penryn's, but for Apple Intel designed something special. —Preceding unsigned comment added by Coldpower27 (talk • contribs) 16:20, 19 January 2008 (UTC)

[edit] compatibility

penryn/wolfdale will go on my snata rosa chipset? yes or no? Need new bios i suppose? 78.86.18.55 (talk) 15:52, 28 January 2008 (UTC)

[edit] Real World Comparisons

There is a lot of great tech jargon to talk shop with in this article, but what is the improvement in speed for each chip in real terms? A chart and a link at the end of it to one or more test comparison sites would be appreciated by many readers. (And, if not for this article, can anyone direct me to a site that gives some kind of statistical comparison?)

I am going to buy a new high-end laptop for graphic design. The current processor options are: T9500 (2.6 GHz), T9300 (2.5 GHz) T8300 (2.4 GHz), T8100 (2.1 GHz), X7900 (2.8 GHz), T7800 (2.6 GHz), T7700 (2.4 GHz), or a T7500 (2.2 GHz). - - (Today is 10 fevr. 2008) - - These chip options come on the Sony AR line, LG W1 Line, and Dell Precision line, supposedly their best laptops. - - Is an X7900 faster than a T9500?

This article needs someway for me and other readers to evaluate the bang-for-buck factor in these chips. What is the real incremental speed improvement in each one anyway? Charvex (talk) 08:28, 10 February 2008 (UTC)

You will need a way to define faster. Some tasks dont benifit from multiple cores, some do. Some tasks can be optimised with new CPU instructions, some can't, some arn't. Some tasks are limited in speed by hard disk throughput, harddisk seek time, memory size, memory speed, cache size, network speed, network latency etc. --87.127.117.246 (talk) 01:16, 7 March 2008 (UTC)

[edit] KiB, MiB, etc. and reverts

One or two users continue to revert my edits where I change these terms to the traditional variants. Exactly zero percent of the referenced sources use these terms. Indeed, only two minor pieces of documentation on intel.com make use of 'kib' and 'mib,' and neither consistently, let alone the x86 architecture documents! When I can find these terms in the Oxford English Dictionary, I'll consider them valid. Objections? Give your arguments here. Regards, 130.49.2.244 (talk) 23:14, 23 February 2008 (UTC)

Common use of MB/KB does not mean they are correct, they are even plain wrong. 1 KB means 1000 Bytes but actually it's 1024 Bytes thus 1 KiB. Did you ever try o read the articles like Kibibyte ? --Denniss (talk) 23:20, 23 February 2008 (UTC)
This has been discussed to death at WP:MOSNUM#Binary_prefixes, where the current compromise phrasing includes, "There is consensus that editors should not change prefixes from one style to the other". The general sense there is that articles should continue using whichever set of units have become "established" in that particular article. If you disagree, take it up there and not here. Everyone will be happier if we don't have yet another edit war over this. — Aluvus t/c 04:27, 24 February 2008 (UTC)
Well, I've read nearly the entire discussion on the use of the IEC prefixes, and it seems to me that the argument 'against' is certainly the stronger case. I understand the merits of compromise, but why would I take up further discussion there when this matter concerns this "particular article" (in your own words)? I'm also not sure what you mean by "established," because from reviewing the edit history, a number of editors have tried to change this article to the traditional prefixes to have their changes reverted. If you don't want an edit war, then we can discuss compromise here, until there is a consensus that doesn't include the phrasing "particular article." If you refuse to debate the matter on this "particular article," then I almost certainly have the stronger case.
Remember, Wikipedia is not for things that the IEC made up one day, ten years ago (Wikipedia:Wikipedia_is_not_for_things_made_up_one_day). Perhaps in the interest of 'disambiguation' an enterprising editor should take it upon himself to replace all homonyms in the English language with less-ambiguous equivalents. Sounds ridiculous, right? Aside from adoption, which is very limited, that seems to be precisely what you're trying to do here. 130.49.2.244 (talk) 19:01, 24 February 2008 (UTC)
If you disagree with what is in MOSNUM, take it up at MOSNUM. In the past year or so of edits (around 1,500 individual edits) since the last big project-wide fight over the issue, outside of some hit-and-run edit warring (which usually involved changing just a handful of instances of MiB to MB), this article has consistently used KiB and MiB. Even discounting that, the literal phrasing of the text at MOSNUM comes down directly in opposition to edits that swap one of the styles for the other. If you dislike what MOSNUM says (and there are plenty of reasons to do so), then bring it up at WT:MOSNUM. — Aluvus t/c 03:26, 25 February 2008 (UTC)
I'll give you this one, because the other articles on Intel microprocessors also use the IEC prefixes and I simply don't have the time to support an edit war on all articles. In any case, if you're going to quote this (flawed) policy, you better make sure that "consistent" or "established" are very well defined. 130.49.2.244 (talk) 03:05, 26 February 2008 (UTC)
A very good definition of "consistent" is "the same style is used throughout the article". I'm not going to take a position on xB vs. xiB, but I will take a position on an article using {K,M,G,T}B in some places, and {K,M,G,T}iB in other places, to mean {2^10,2^20,2^30,2^40}B, which is that it's a Very Bad Idea. I.e., if somebody wants to change the notation used, they should change all instances where xiB is used, or all instances where xB is used to refer to a power of 2, rather than leaving the usage mixed. (The Manual Of Style speaks of a consensus that the existing style of an article shouldn't be changed:
There is no consensus to use the newer IEC-recommended prefixes in Wikipedia articles to represent binary units. There is consensus that editors should not change prefixes from one style to the other, especially if there is uncertainty as to which term is appropriate within the context—one must be certain whether "100 GB" means binary not decimal units in the material at hand before disambiguation. When this is certain the use of parentheses for binary prefixes, for example "256 KB (256×210 bytes)", is acceptable, as is the use of footnotes to disambiguate prefixes. Use of IEC prefixes is also acceptable for disambigation (256 KiB). When in doubt, stay with established usage in the article, and follow the lead of the first major contributor. Prefixes in directly quoted passages are never changed; if explanation is necessary, use a more exact measurement in square brackets.
which is a consensus that should, if nothing else, fend off edit wars.) Guy Harris (talk) 03:33, 26 February 2008 (UTC)
That is not a very good definition because it only refers to the particular article. If an article is in the context of an encyclopedia, it should share the same general style as comparable articles. Now there is clearly an issue with policy here, compare the articles for iBook and MacBook for instance. How can you possibly defend a policy that effectively says that you should look no further than the particular article and it's "established" usage to decide which usage is appropriate? Here's an exercise in thought: how many featured articles in computing Wikipedia:Featured_articles#Computing use the IEC prefixes? 130.49.2.244 (talk) 22:43, 26 February 2008 (UTC)
Wikipedia is not for stuff you, i.e. the contributor, or your friends, i.e. the contributor's friends, made up one day, as the page you cited says. It does include stuff somebody made up at one point, as that page notes ("Of course, everything in Wikipedia was made up by someone at some point in time"), such as calculus, general relativity, quantum mechanics, the C programming language, etc., so that is, by itself, not sufficient reason to avoid the IEC prefixes. Now, that page also does indicate that some degree of notability is required - but that page is primarily intended to keep people from putting up pages about the club they and their friends started last week at school. The IEC is, I suspect, not known to the majority of the educated populace, and probably not even to the majority of the technically-educated populace, but they're not exactly an obscure body of whom nobody has heard and that has no influence, either. Guy Harris (talk) 03:42, 26 February 2008 (UTC)
I wouldn't deem my reasoning up to this point to avoid the IEC prefixes sufficient either, since this has mostly been a discussion on policy. There are many other reasons, but that one is very well suited to analogy. I'm not going to argue with you over what is 'implied' by that page, or what you think is "primarily intended." Though, if you read the second sentence, it does mention "not yet become well known to the rest of the world." This goes hand in hand with 'notable'. 130.49.2.244 (talk) 22:43, 26 February 2008 (UTC)

[edit] Merom etymology

The article claims that: Merom (מרום) is the Hebrew word for a higher plane of existence or a level of heaven, BaMerom (במרום) means "in the heavens". The name was chosen by the Intel team in Haifa, Israel, who designed this processor. Are you positive? Any reference on that? Because there is also a biblical lake Merom which was situtated in what is today Hula Valley, and my gumption tells me that it's thence the processor code name is derived from. Thanks, Maikel (talk) 17:08, 12 March 2008 (UTC) PS: I wish they'd just number the f*ckers.

[edit] Yorkfield released

Yorkfield has been released, At least in the US, and in OEM form.

http://www.tankguys.com/product_info.php?products_id=1817

Not sure about elsewhere.

Thus, this article should be updated soon to reflect this.

67.224.19.160 (talk) 20:29, 26 March 2008 (UTC)

I've updated it with the most recent information so far. --GuitarFreak (talk) 13:59, 9 April 2008 (UTC)

[edit] MiB and KiB v.s. MB and KB

Note. For interested authors, debate and a vote is ongoing on Talk:MOSNUM regarding a proposal that would deprecate the use of computer terms like “kibibyte” (symbol “KiB”), “mebibyte” (symbol “MiB”), and kibibit (symbol “Kib”). It would no longer be permissible to use terminology like a “a SODIMM card with a capacity of two gibibytes (2 GiB) first became available…” and instead, the terminology currently used by manufacturers of computer equipment and general-circulation computer magazines (“two gigabytes, or 2 GB”) would be used. Voting on the proposal is ongoing here. Greg L (my talk) 21:16, 27 March 2008 (UTC)