Talk:Nehalem (microarchitecture)/Archive 1
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 |
Nehalem to become Corei7 processor
See the expreview.com article. I updated the article with this information, but the title etc needs changing. I hope I did okay. 219.90.249.161 (talk) 12:55, 8 August 2008 (UTC)
- Title doesn't need changing yet, we know the name of the desktop CPU line but the architecture probably won't be called 'i7' or 'Core i7', and i like pie, at least we can't just assume that. (see 'NetBurst', no processors are named anything like that) --Joffeloff (talk) 17:29, 9 August 2008 (UTC)
- [1] Official News Release :) --Vectrox (talk) 13:52, 11 August 2008 (UTC)
- Its a news release saying they are going to again use Core for the deskstop chip names and does not mention a new name for the μarchitecture which is AFAICT being referred to as "codenamed Nehalam"[2]. -- KelleyCook (talk) 17:49, 11 August 2008 (UTC)
- The Core i7 name only applies to the Bloomfield desktop variant.Jasper Deng (talk) 00:54, 13 November 2008 (UTC)
- Its a news release saying they are going to again use Core for the deskstop chip names and does not mention a new name for the μarchitecture which is AFAICT being referred to as "codenamed Nehalam"[2]. -- KelleyCook (talk) 17:49, 11 August 2008 (UTC)
- [1] Official News Release :) --Vectrox (talk) 13:52, 11 August 2008 (UTC)
3DMark Vantage Scores
The article states a 2.93Ghz Nehalem scored 17966 in 3DMark Vantage. It compares it to a Core 2 Duo E6600 (2.4Ghz, 65nm etc) Vantage score. Wouldn't it be better to compare the 2.93Ghz Nehalem to the QX6800 - a quad core with the same clock frequency, and compare the 2.66Ghz Nehalem to the QX6700 (65nm) or Q9450 (45nm)? It seems unfair to compare a quad with a dual core of a lower clock frequency, smaller cache, different manufacturing method (65nm compared to 45nm) etc. —Preceding unsigned comment added by 219.90.155.192 (talk) 00:46, 3 August 2008 (UTC)
It's a comparison to give you a feel for the relative performance. E6600 was a popular CPU. Skilltim (talk) 16:11, 12 August 2008 (UTC)
Rename article?
Shouldn't the article be renamaed to "Nehalem (microarchitecture)" (together with the Intel Core (CPU architecture) and Gesher (CPU architecture) articles) to avoid using the word "architecture" for microarchitecture (hardware) and instruction set (software) possibly confusing a bit?
- I think that maybe they should be renamed to "Nehalem (CPU microarchitecture", "Intel Core (CPU microarchitecture)" and "[[Sandy Bridge (CPU microarchitecture)" instead. Core, Nehalem, and Sandy Bridge/Gesher are microarchitectures, not architectures, but it should be made clear that they are CPUs. Imperator3733 15:36, 24 September 2007 (UTC)
- I mentioned at Talk:Intel Core (CPU architecture) a new, revised idea for renaming these articles. The idea is to rename the articles to Intel Core (microarchitecture), Nehalem (microarchitecture), and Sandy Bridge (microarchitecture). Another possibility would be to have "Intel" in all the names (making the naming more consistent -- i.e. Intel Nehalem (microarchitecture) and Intel Sandy Bridge (microarchitecture). If we do that I would suggest making the change to other uArch pages (i.e. NetBurst to Intel NetBurst (microarchitecture), Intel P6 to Intel P6 (microarchitecture), etc). What do you think? I'll give this a week for comments. If the response is favorable, I (or someone else) will/can rename them. If there is no response, I'll probably wait a bit longer. If anyone sees this, please respond. Thank you. -- Imperator3733 00:22, 16 October 2007 (UTC)
- The pages have now been renamed. See Intel Core (microarchitecture), Nehalem (microarchitecture), and Sandy Bridge (microarchitecture) -- Imperator3733 18:06, 24 October 2007 (UTC)
Processor roadmaps
I noticed that NetBurst and Intel Core (microarchitecture) have the processor roadmap at the top of the page, after the introduction paragraphs, while Nehalem (microarchitecture) and Sandy Bridge (microarchitecture) have it later on in the article. Should the location of the roadmap graphic be standardized, and if so, where? Also, does anyone know why there is another version of the roadmap? This would make it the third version of this roadmap on Wikipedia! The three versions are: my original version, which is now obsolete, WhiteTimberwolf's SVG version, which should be the one that we use, and now the third version. -- Imperator3733 (talk) 15:03, 11 June 2008 (UTC)
- As long as it's put in the right place in the article I don't think it's that important. (Skilltim (talk) 09:14, 12 June 2008 (UTC))
- The problem is that we already had an image that gave the exact same information as the new on, and that under WP:IUP#Rules of thumb #7, SVG is to be used for drawings and PNG when it is the only choice. The SVG image is also less than a fifth of the size as the PNG version. -- Imperator3733 (talk) 14:52, 16 June 2008 (UTC)
I prefer using the SVG, it means for scalability way better for Wikipedia, PNG quality really bad. The roadmap is meant for presentation not for quality picture like a picture for that shows World Heritage or whatever. --Ramu50 (talk) 22:09, 23 July 2008 (UTC)
Compatibility (LGA 775 sockets)
Will the Nehalem CPUs be compatible with current LGA 775 sockets and current chipsets (P35, G33, G35, etc.) in use today? If not, will there be any special versions that can run on today's computer motherboards? Hellcat fighter (talk) 14:39, 24 November 2007 (UTC)
- I do not believe so, since Nehalem uses an integrated memory controller and QuickPath. This requires a new socket, since there is a new bus and the system is set up differently -- if they used the same socket, a system could end up having two memory controllers. At least the server Nehalems will be using LGA1366 as a socket -- I don't know what the desktop chips will use. Because of the bus and memory controllers, the current chipsets won't work with these either. I think there were rumors of cut down/budget/legacy versions to run on LGA775, but I personally hope that that doesn't happen. -- Imperator3733 (talk) 01:11, 25 November 2007 (UTC)
- If what you mentioned is true, current LGA775 board owners will have to replace their motherboards to upgrade to Nehalem processors. An Intel chipset less than two years old may not be able to support a Nehalem processor. This will be financially costly to the PC upgrader. Intel revises (or changes) their chipsets quite frequently, so upgrading the CPU of an Intel based PC can be expensive thing to do. AMD based PCs do not suffer from as high CPU upgrading costs (since socket AM2). Hellcat fighter (talk) 18:39, 26 November 2007 (UTC)
- This isn't the first time that something like this has happened, and it certainly won't be the last. At least this time it is really necessary because of all the changes that are happening (IMC, QuickPath, etc.) -- Imperator3733 (talk) 18:45, 26 November 2007 (UTC)
- LGA775 has had a disproportionately long lifespan compared to other sockets. If Nehalem was compatible with that socket all the improvements (IMC, QPI) would go out the window. There are other improvements, sure, but without the bandwidth it's almost negligible. --Joffeloff (talk) 16:07, 12 August 2008 (UTC)
- The new socket for all the new Nehalem CPU is LGA 1366
- Uh-UH. The Beckton MP server CPU has socket 1567. The Lynnfield and Havendale CPUS use socket 1156. The Clarksfield CPU uses the mPGA989 socket.These contradict all of your comments assuming the table is correct.Jasper Deng (talk) 00:53, 13 November 2008 (UTC)
- for consumers socket 1156 will be the future! s1366 will be a XEON part, and won't be the defacto after the 1156. Markthemac (talk) 03:41, 29 April 2009 (UTC)
- Uh-UH. The Beckton MP server CPU has socket 1567. The Lynnfield and Havendale CPUS use socket 1156. The Clarksfield CPU uses the mPGA989 socket.These contradict all of your comments assuming the table is correct.Jasper Deng (talk) 00:53, 13 November 2008 (UTC)
- The new socket for all the new Nehalem CPU is LGA 1366
- LGA775 has had a disproportionately long lifespan compared to other sockets. If Nehalem was compatible with that socket all the improvements (IMC, QPI) would go out the window. There are other improvements, sure, but without the bandwidth it's almost negligible. --Joffeloff (talk) 16:07, 12 August 2008 (UTC)
Power
"30% lower power usage for the same performance." - compared to what? Penryn? —Preceding unsigned comment added by 77.183.56.218 (talk) 16:27, 13 April 2008 (UTC)
- I've made the sentence before it make more sense so it is comparing it with Penryn. (Skilltim (talk) 23:01, 17 June 2008 (UTC))
- UH-UH! Beckton, for MP servers (with 8 cores and 16 threads), uses a different socket, LGA1567, the Havendale and
Lynnfield.Jasper Deng (talk) 00:59, 13 November 2008 (UTC)
Question, Corrections...etc.
QPI count
Intel documentetion:
1. does not make difference in processor-to-processor and processor-to-chipset QPI
2. shows that there would be at least 3 QPI links for Xeon core S2 (aka Gainestown)Stasdm (talk) 16:06, 22 January 2009 (UTC)
Cache on Havendale and Auburndale
Does anyone know why the variants table lists Havendale and Auburndale as having 4MB L2 cache with no L3 while all the other chips have and L3 cache? The other Nehalems all have 256KB L2 per core, so changing to a shared L2 in the low-end chips seems to be too much work. Is there any place that specifically said that Havendale and Auburndale have L2 cache? -- Imperator3733 (talk) 00:44, 12 February 2008 (UTC)
- I've updated the cache column, I think you'll find it makes sense now. It looks like (if it is the most recent information) the mainstream LGA1160 form of Nehalem has been pushed back to H2 2009. Is this reliable data? I thought they were going to be released about 3 months after the LGA1366 Nahalems. (Skilltim (talk) 08:30, 7 June 2008 (UTC))
- Thanks for updating the table. The LGA1160 delay seems pretty reliable (X-bit Labs had an article), although I'm still quite mad at Intel about it (I had been thinking of doing an upgrade when those came out). The X-bit Labs article says "second half of 2009". I also noticed that the tables are getting a bit cramped, at least when I'm viewing it. Do you have any ideas on how to make it easier to read? -- Imperator3733 (talk) 04:08, 9 June 2008 (UTC)
- See how you find that, I've cleared the clutter. To read it as it should be you'll need a 19"+ monitor. I can't wait until mid-2009, I may have to get the lowest specification Bloomfield. (Skilltim (talk) 16:41, 10 June 2008 (UTC))
Penryn L2 cache
Above the Nehalem variants table,list a short paragraph which state that the Core 2 Penryn has 12 MB L2 cache.What rubbish is this? Penryn has only max 6MB for the 6MB version and even in the XE too. —Preceding unsigned comment added by 58.182.75.239 (talk) 05:33, 15 June 2008 (UTC)
- I think that person meant Yorkfield as he was comparing like for like in the quad core ranges. I have amended. (Skilltim (talk) 11:06, 15 June 2008 (UTC))
Release date for Mobile Nehalems
The top of the articles says "Mobile Nehalem-based processors will follow in 2010", but the table says that Clarksfield will come "Q3 2009" (which assuming American quarters, ends June 30 2009). Both aren't cited. Any one know of any thing? 81.99.153.69 (talk) 09:03, 3 May 2009 (UTC)
News report
news story - added info on memory controller integration
http://www.betanews.com/article/Intel_Officially_Confirms_Integrated_Memory_Controller_for_45nm_Nehalem/1175123788 —The preceding unsigned comment was added by 24.22.22.101 (talk) 16:10, 29 March 2007 (UTC).
I was wondering if anyone thinks that we should add a page on the original Nehalem (the NetBurst version). I also don't know where this page should be located. Imperator3733 19:41, 22 April 2007 (UTC)
"Turbo Mode" for Nehalem
I just came across this article today (Aug. 20, 2008) and believe it would be very relevant to the Nehalem (microarchitecture) page. "Intel's Nehalem Processor Gets Turbo Boost - Technology News by ExtremeTech" It describes a method by which Intel hopes to more efficiently use multicore processors for single-threaded applications. I personally don't have the time to research this information further, nor do I have time to verify through any other sources. If someone would like to do this, please feel free to add it to the main article. It's also entirely possible that I have failed miserably and this is already in the wiki article and I just didn't notice :-). If this is the case, please disregard me.Theweirdguy (talk) 19:06, 20 August 2008 (UTC)
Auburndale/Havendale Canceled?
According to http://theovalich.wordpress.com/2009/01/31/exclusive-intels-cans-45nm-auburndale-and-havendale-fusion-cpus/ and http://www.hardocp.com/news.html?news=Mzc1OTMsLCxoZW50aHVzaWFzdCwsLDE= Intel is canceling Auburndale and Havendale...if it is true should the table be updated? Starrynte (talk) 23:54, 31 January 2009 (UTC)
Rumors and Speculations
nVidia dispute
There has been a dispute between Nvidia and Intel about Nvidia's lisence to make nForce chipsets on nehalam. Basically Intel is telling Nvidia that they will not allow Nvidia to make the chipsets since Nvidia wont lisence its SLI technology to Intel. I think this is pertinant to this article, maybe a section should be added since this is getting ready to turn into a pretty big legal battle? —Preceding unsigned comment added by 155.9.10.2 (talk) 15:15, 13 June 2008 (UTC)
- How is this news relating to the Nehalem (microarchitecture)? If anything it should be added to X58 or PCH with references. (Skilltim (talk) 02:17, 15 June 2008 (UTC))
by the way can somebody fix the table alignment and width. I am not good with any type of markup table related. Was going to suggest Lynnfield and Havendale, but guess my info was one day late - -
Oh I have organized the talk page to order of importance, hope you guys don't mind. The order of importance is as follows
- Decision about article: moving, renaming, merging..etc.
- Roadmap
- Compatibility & Power (some people consider very important, while other think it is not depends on your viewpoint)
- Questions
- News and Rumors (usually I think news that are confirmed or the news reporter don't start guessig then we should start the article, though I don't want to discourage you guys from doing it, because I am Forum fan, but it is a good way for talk page management. --Ramu50 (talk) 00:16, 18 July 2008 (UTC)
- this news is related in the fact that Nehalem has different requirements and these will need to bee licensed as well.Jasper Deng (talk) 22:42, 15 September 2008 (UTC)
- If you haven't read, nVidia may have given up on trying to do this, because it is allowing Intel's X58 chipset to run Tri-SLI. But it also said it would only work with the first CPU in this lineup, the Core i7. I want an explanation.Jasper Deng (talk) 00:59, 13 November 2008 (UTC)
- this news is related in the fact that Nehalem has different requirements and these will need to bee licensed as well.Jasper Deng (talk) 22:42, 15 September 2008 (UTC)
Citation
Images aren't valid citation as anyone can upload it. Even if they are publish by news reporter companies, you are required to cite where it use the image to prevent the misleading of what is the purpose of the image
[http://www.techpowerup.com/img/08-07-09/3.jpg]
QuickPath Interconnect is design for workstation and server class products not for regular consumers. Also the article doesn't support the doubts. Citation removed. Proof is here QuickPath interconnect
The QuickPath Interconnect is Intel's new system architecture interconnect for servers and high-end workstation products replacing its front side bus architecture.
Your incorrect info
The mainstream PCH will not have the QuickPath Interconnect so its overclocking potential is called into question.[http://translate.google.co.uk/translate?u=http% 3A%2F%2Fwww.hkepc.com%2F%3Fid%3D568%26page%3D3%26fs%3Didn%23view&sl=zh-CN&tl=en&hl=en&ie=UTF-8] [3]
Personal Comment is consider Original Research, not a citation
[http://www.cpu3d.com/content/view/5087/67/]
Warning to User:Skilltim
You remove somebody's image citation without any edit summary or leave anything in talk pages. Don't do that again, it is consider Vandalism. --Ramu50 (talk) 01:25, 21 July 2008 (UTC)
- Read[4] up some more before you come in here proclaiming you know everything about a CPU that hasn't been released. When it is released we will know the facts but until then this whole topic is speculation. Skilltim (talk) 07:37, 24 July 2008 (UTC)
Wikipedia only allow you to cite blogs if it is written by a professional, which you haven't done so. I know how the proccessors work, its pretty a similar SoC design of UltraSPARC T2, go read Sandy Bridge
* Aimed at Sun's UltraSPARC T1 and its successors
Wikipedia:Verifiability, See #5
Since when I didn't say anything about release date, you got any proof? I only said don't put your own opinion onto Wikipedia article. Read my edit summary
http://wiki.riteme.site/w/index.php?title=Nehalem_%28microarchitecture%29&diff=227434327&oldid=227264171 (invalid references, no mention of overclocking))
If the references does say something like multiplier is locked or some computing technicals term, then overclocking statement can be placed accordingly, but apparently there is no mention of that anywhere.
Also I suggest not to use the undo option, if other users have contrib new edits
The citation you used from Softpedia, I have keep it temporarily, because Softpedia have a very strong tendency to constantly change their viewpoints, I read their news everday so I know e.g. are SSD Energy Efficient
--Ramu50 (talk) 16:59, 24 July 2008 (UTC)
- Screw this, Wikipedia just lost a valuable truth seeker. Skilltim (talk) 18:14, 24 July 2008 (UTC)
- I see you are back to reverting articles to your spelling mistakes, silly layouts and unfree images. (Core i7, etc) Can't you please hold yourself to your promise and be gone? --Joffeloff (talk) 17:28, 9 August 2008 (UTC)
Which we don't need. By the way your citation actually contradict with your statement, it doesn't support it at all. --Ramu50 (talk) 23:41, 24 July 2008 (UTC)
Improvements over Core 2 Duo
So what are the practical improvements over Core 2 Duo or Core 2 Quad processors? Is a 2.6 GHz Core i7 going to be faster than an equally priced 2.6 GHz Core 2 Quad, or is it simply a matter of using less energy, or what? Althepal (talk) 18:03, 17 August 2008 (UTC)
- There are quite a few improvements over Core 2. However, this page is for discussions regarding the article, not the article's subject. I suggest you take a look at various hardware sites, such as AnandTech for more information. Specifically, this article has quite a bit of information. -- Imperator3733 (talk) 18:36, 17 August 2008 (UTC)
- I don't know... readers might be interested in practical information rather than just simply technical stats. But thanks for the links. Althepal (talk) 20:15, 17 August 2008 (UTC)
- There is some basic information on improvements in the "Performance and power improvements" section (which needs to be cleaned up a bit). If readers want more specific information, then they can always look at the external links. -- Imperator3733 (talk) 03:21, 18 August 2008 (UTC)
- Looks good. Althepal (talk) 19:28, 20 August 2008 (UTC)
- Some people say that this CPU will have a clock speed of 4 or even 5 gigahertz later on.Jasper Deng (talk) 22:42, 15 September 2008 (UTC)
- Looks good. Althepal (talk) 19:28, 20 August 2008 (UTC)
- There is some basic information on improvements in the "Performance and power improvements" section (which needs to be cleaned up a bit). If readers want more specific information, then they can always look at the external links. -- Imperator3733 (talk) 03:21, 18 August 2008 (UTC)
- I don't know... readers might be interested in practical information rather than just simply technical stats. But thanks for the links. Althepal (talk) 20:15, 17 August 2008 (UTC)
The Nehalem processor has been released in new zealand . This page refers it to an up comming event . (http://www.computerlounge.co.nz/components/componentview.asp?partid=7487) this it just a sample that i found regarding nehalems release. cheers —Preceding unsigned comment added by Xu-Holland (talk • contribs) 10:05, 12 October 2008 (UTC)
- Interesting. It's already Monday in New Zealand, and Intel usually releases on Monday... -Arch dude (talk) 19:25, 12 October 2008 (UTC)
mmm very however i cannot find any informantion regarding nehalems realese -in New Zealand- on intels website. but its been released . When I checked on thursday the 9th of october it had been released. It may have been released earlier but i was on holiday . only 940 and 920 have come out . http://www.pricespy.co.nz/cat_2.html . —Preceding unsigned comment added by Xu-Holland (talk • contribs) 09:50, 13 October 2008 (UTC)
Removal of Clarkdale entry from the Variants table
While I was re-organizing the table (which I can give the reasons for separately, if required), I researched the processor codenamed Clarkdale. It only turned up in a really old article, and I'm fairly certain that this was a codename for Clarksfield, whether misheard or wrongly edited at the time. If there is any more evidence about it, of course, it will come back.
--Masudr (talk) 05:23, 22 November 2008 (UTC)
marketing hype
I commented out "Nehalem is the largest change in Intel's system architecture since the introduction of the Pentium Pro" which was cited to "http: //www.realworldtech.com/page.cfm?ArticleID=RWT040208182719&p=2 Real World Technologies - Inside Nehalem: Intel's Future Processor and System]". The page linked does not support the claim that Nehalem is the largest change since the ppro, and that claim itself is very dubious. The ppro was an enormous leap over earlier x86's, introducing a micro-op architecture with 3 pipes and out-of-order execution that could keep dozens of pending operations in the air, vs the Pentium which was basically two traditional 486 pipes stapled together. That's not to disparage the Nehalem, which is another in a good steady stream of incremental improvements in the series. But Nehalem is a few percent faster than Penryn while the PPro was probably close to double the speed of the pentium. Architecturally the Nehalem seems like a smaller change from Core 2 than Core 2 was from the P4 (among other things, that change went from 32 to 64 bits). I also removed another sentence that seemed like uninformative marketing. 208.120.235.110 (talk) 10:57, 31 December 2008 (UTC)
- David Kanter is a credible tech journalist who runs a credible site. Here's what it says;
- 'Nehalem represents a complete and total overhaul of Intel’s system infrastructure, one of the most dramatic changes since the introduction of the P6 in the latter part of the last decade.'
- SYSTEM ARCHITECTURE. Not the CPU architecture. The pipeline has received very little changes, the whole point is that the FSB, memory controller hub, all that, it's gone. Nehalem isn't a 'few percent' faster than Penryn, it's twice as fast in situations where the old system architecture (lack of bandwidth) was the bottleneck. It's a huge change.
- Oh well.. Play it your way. Sure, it's not a big change if all you do with it is to play games. For serious work the revolution can clearly be seen in the results. --Joffeloff (talk) 17:53, 1 January 2009 (UTC)
Core i5
Please stop adding the name 'Core i5' to Lynnfields - CPU-Z is not an official program, i5 has not been officially confirmed. This is the same as adding 'Core 3' to the article before the i7 name was known (as people did). --Joffeloff (talk) 20:06, 6 January 2009 (UTC)
New table is flawed
![](http://upload.wikimedia.org/wikipedia/en/thumb/1/15/Ambox_warning_pn.svg/48px-Ambox_warning_pn.svg.png)
/!\Who messed up the table like that? Here is a list of the flaws:
- No mobile section in Westmere catagory
- DP server and Desktop sections not seperated into rows
- # of cores not mentioned (not even a column!)
- Other columns missing.
This list sadly is incomplete. Jasper Deng (talk) 00:38, 7 January 2009 (UTC)
- Can you give me a link to the diff where these errors have appeared? I can't seem to find them myself. Masudr (talk) 03:09, 7 January 2009 (UTC)
- Mostly in the Westmere catagorie at the bottom. —Preceding unsigned comment added by 209.155.146.2 (talk) 00:37, 21 January 2009 (UTC)
- I'll try to address Jasper Deng's issues point by point:
- There is no mobile section in the Westmere category because we do not have any data about Westmere-based processors apart from the entry that is written.
- The sections aren't separated since we don't know the specifics of what the DP desktop and DP server sections will contain.
- There is a No. of cores column.
- There are no other missing columns.
- --Masudr (talk) 04:38, 21 January 2009 (UTC)
- I'll try to address Jasper Deng's issues point by point:
That's only after you changed it. I still would prefer a single table vs. two tables though.Jasper Deng (talk) 23:36, 21 January 2009 (UTC)
- Of course. I performed several edits on the table from [5] dated 16:11, 21 November 2008, to [6] dated 05:46, 20 December 2008. The reason I was confused about your comment is that it came on the 7th January, but all the things you complained about were there since the 20th December.
- I personally think the table as it is now is immensely more useful than as it was prior to the 21st November. As for splitting the two tables apart, I have little issue with it, and don't mind seeing it either way (i.e. split apart or together). Note that I wasn't the one who performed the split.
- I hope that clears things up. --Masudr (talk) 00:49, 22 January 2009 (UTC)
- That does not since much of the previous data over there was deleted (for example the Clarksfield processor's socket number). —Preceding unsigned comment added by Jasper Deng (talk • contribs) 23:50, 22 January 2009 (UTC)
- Both links I have provided in grandparent comment (and indeed the current version [7] of the article) all list the Clarksfield socket number as mPGA-989. What, exactly, was deleted? --Masudr (talk) 01:26, 23 January 2009 (UTC)
Core i-
I am coming to believe that the number after i will be the number of threads minus one.Jasper Deng (talk) 23:36, 21 January 2009 (UTC)
- Doesn't make sense at all. What are they supposed to call Lynnfield then? It's a quad-core with HT too. Joffeloff (talk) 21:25, 27 January 2009 (UTC)
- They might use a different letter in that case.Jasper Deng (talk) 23:47, 27 January 2009 (UTC)
Xeon 3500 series is missing from table
The table is missing information about the Xeon 3500 serie to be relased at the end of March with the 5500 serie.
The Xeon 3500 serie is the uni-processor (UP) server model for Nehalem.
The 3 model of the 3500 serie are the W3570, W3540, W3520 and they match Bloomfield Desktop 964, 940 and 920 model specifications (but have ECC enabled). Pricing is pretty similar to the desktop version (both articles don't have similar pricing). —Preceding unsigned comment added by 138.11.0.153 (talk) 00:05, 4 February 2009 (UTC)
Jasper Forest
Revealed at this IDF Beijing, I originally placed this under 'Future Xeons' but it might not be a Xeon. I can't edit the table without screwing it up, anyone else?
Jasper Forest
Jasper Forest is a Nehalem-based embedded processor with PCI Express connections on-die, core counts from 1 to 4 cores and power envelopes from 23 to 85 watts.[1]
—Preceding unsigned comment added by Joffeloff (talk • contribs) 20:50, April 10, 2009
- The (extremely brief) blurb from Intel describes it as a Sub 25-Watt Xeon (although higher-TDP parts are also described) intended, in part, for blade servers. Unfortunately, that link will die once the item scrolls off Intel's "Chip Shots" page, so it's no help if you want a reference. — Aluvus t/c 04:11, 11 April 2009 (UTC)
- Oh, well, that solves it then - it states clearly it's a Xeon. I added it back to the future Xeons section now. --Joffeloff (talk) 00:59, 12 April 2009 (UTC)
Article title
Intel Nehalem (microarchitecture) is redundant and clumsy. Unless I am missing something, it suffices to be either Intel Nehalem or Nehalem (microarchitecture), or if you really want to spell it all, Intel Nehalem microarchitecture. Please comment. - Altenmann >t 18:38, 26 October 2009 (UTC)
- Nehalem is both the general name used for the micro-architecture, and the first of the cores to use it. By analogy, the USS Nimitz (CVN-68) was the first Nimitz class aircraft carrier. My guess would be that the current name was an attempt to disambiguate the core and the micro-architecture. P6 and Core use the same name format (in the case of Core it is necessary to disambiguate from the Core series of processors; no need with P6). NetBurst does not. On the other side of the fence, AMD K9 and AMD K10 are currently the only AMD micro-architectures with their own articles and they are named as written. It would probably be sensible to get all of the micro-architecture articles named in some consistent fashion (one might argue that simply moving P6 to Intel P6 would make everything consistent). — Aluvus t/c 02:21, 27 October 2009 (UTC)
Core i9 clock speed citation needed, and other conflicting info on Gulfttown . Also is it really going to be a single variant of Core i9 with a single number??
First of all, the 3.06 GHz clock speed listed in the Westmere variants table has a Citation Needed tag on it. Can someone fix it?
Second, the table's 3.06 GHz speed conflicts with the Gulftown article.
And Third, I do not believe that Intel Core i9 will come in just a single variant with processor number 1000 (Intel never uses a hundred, only a hundred-and-something (like 1020). Jasper Deng (talk) 01:06, 8 November 2009 (UTC)
- The entry you are describing says "1000 series", implying several products whose model numbers fit the form 1xxx or possibly 10xx, not a single model number. In any event, model numbers ending with two zeros (like the Core 2 Duo E6300 and friends) are not unheard of for Intel. The clock speed could certainly use a citation, but then so could most of the future products that are listed. As a minor aside, I would encourage you to use shorter section titles. This section could just as easily have been titled "Core i9/Gulftown". — Aluvus t/c 06:34, 8 November 2009 (UTC)
Duplicate information in Table
The large table in this article is IMHO both unreadable because of its size and duplicates the information already available in the "List of Intel XXX processor" articles. When 218.186.12.226 removed the table recently, I added back a simpler one with Links to the articles explaining the cores and to the lists, but without duplicating all the information for them. User:Jasper Deng then reverted both with the comment "Unfortunately when the series number is known, you cannot just simply say something like 1xx-that is for CPUs whose model # is unknown". This seems inconsistent with what we do at other articles like Intel Core or Lynnfield (microprocessor), which all use this form to link to the real lists.
I really think we need to cut down the duplication between detailed tables in the CPU articles. Arndbergmann (talk) 10:23, 14 November 2009 (UTC)
- Actually you're not correct. Wikipedia always lists the different variants and not just things like 9xx, which is only for speculation. These types of simpler tables are only used for summarizing CPU series and not for a full listing of variants as this article does. The information is NOT duplicated, sorry for what you think. And anyways, most of these lists of different CPUs do not exist or use 1xx. In essence, this article is the list and not the summary.Jasper Deng (talk) 03:54, 19 November 2009 (UTC)
Intel Core i7 980x vs. Intel Core i9 10xx as names for Gulftown
Gulftown will be called Core i7-980x by Intel apparently according to a Russian site. However this site didn't cite good sources and its the only reference.
Add a better reference or do something else about this. Jasper Deng (talk) 03:25, 11 December 2009 (UTC)
Multiple QPI Link Confusion
Currently on the page there are 4xQPI, 2xQPI, and 1xQPI listed under Interfaces->Chipset table columns. What may be confusing to some people is that Bloomfield and Gainestown are actually made from the same die, so Bloomfield has two QPI links with one disabled. That's why a lot of Bloomfield die diagrams, such as [10] (posted by Hans de Vries on a blocked website), show two QPI links. The additional QPI link is for additional sockets, and is disabled for desktop processors and the Xeon 3000 series. The pending Gulftown Desktop CPU, still being socket 1366, will probably have two QPI links for a probable Xeon Multi-CPU version, with one disabled for Single-CPU/Single-Socket Desktop versions. But the link is still there. 71.179.243.191 (talk) 23:11, 12 December 2009 (UTC)
Archiving
Does anyone object to me setting up automatic archiving for this page using MizaBot? Unless otherwise agreed, I would set it to archive threads that have been inactive for 60 days.--Oneiros (talk) 23:17, 29 December 2009 (UTC)
Done--Oneiros (talk) 01:04, 2 January 2010 (UTC)