From: Aioanei Rares on 15 Apr 2010 20:10 On 04/16/2010 01:59 AM, Stan Hoeppner wrote: > Aioanei Rares put forth on 4/15/2010 3:16 PM: > > >> No, it means that 2 one-gig modules in same-coloured slots will >> theoretically work better. I have one module in a dual-channel mobo and >> it works ok. Most desktop/workstation mobos do. >> > No, that's not correct. If the mobo DIMM slots are colored, you put one > DIMM in each color slot for dual channel operation, not both DIMMs in the > same colored slots. > > This is a very simple concept folks. If you have two sets of train tracks > you can move twice as many trains from point A to point B in a given amount > of time. Dual channel memory works on the same principal, doubling the > number of paths to/from memory. > > Are you guys displaced Ubuntu users who landed on debian-user after a > tornado came through Kansas? ;) > > You made me dig my latest mobo's manual :) [Gigabyte GA-M56S-S3], which says plainly : "The four DDR2 memory sockets are divided into two channels and each channel has two memory sockets as following : Channel 0 : DDRII_1, DDRII_3 Channel 2 : DDRII_1, DDRII_4" Please refrain from making such "wise" comments as , even if you're right, you're damaging Debian's image and yours. Cheers, -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/4BC7AA0C.90505(a)gmail.com
From: Stan Hoeppner on 15 Apr 2010 21:40 Aioanei Rares put forth on 4/15/2010 7:06 PM: >> Are you guys displaced Ubuntu users who landed on debian-user after a >> tornado came through Kansas? ;) I guess you've never seen "The Wizard of Oz". > You made me dig my latest mobo's manual :) [Gigabyte GA-M56S-S3], which > says plainly : "The four DDR2 memory sockets are divided into two > channels and each channel has two memory sockets as following : > Channel 0 : DDRII_1, DDRII_3 > Channel 2 : DDRII_1, DDRII_4" You typo'd that. It should be: Channel 0 : DDRII_1, DDRII_3 Channel 1 : DDRII_2, DDRII_4 > Please refrain from making such "wise" comments as , even if you're > right, you're damaging Debian's image and yours. I'm not sure how you come to that, but, ok.. Gigabyte has screwed up the DIMM slot color coding on the GA-M56S-S3, at least according to the manual. I just looked at Intel, Asus, and other Gigabyte boards and they all follow the correct alternating color pattern. Aioanei, you've based your entire argument on your knowledge of a single main board. This board was incorrectly manufactured, or the manual text is incorrect. Yes, manufacturers do make mistakes. Look at the ATA_HORKAGE blacklist sections in libata-core.c to see lists of hard drives the manufacturers screwed up. All of the Goliaths are in there, Hitachi, Maxtor, Seagate, Western Digital. They're all shipped drives that don't meet one spec or another. If these Goliaths can screw up, I guarantee you a tiny company like Gigabyte can, and probably does so far more often. Now, go look at the DIMM socket colors and layout pattern on the Intel, Asus, and yes, other Gigabyte boards to see what the proper color coding for dual channel memory sockets is, and look at the text descriptions of the slot and channel numbers. This GA board is an anomaly, and is flawed in this regard. Take a look at the RHS of my email address. That's my personal domain. They call me TheHardwareFreak for a reason. ;) The odds against me making a factually incorrect statement here about hardware are so high you should place a standing bet with your local bookie. :P -- Stan -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/4BC7BDDE.40503(a)hardwarefreak.com
From: Aioanei Rares on 16 Apr 2010 09:10 On 04/16/2010 04:31 AM, Stan Hoeppner wrote: > Aioanei Rares put forth on 4/15/2010 7:06 PM: > > >>> Are you guys displaced Ubuntu users who landed on debian-user after a >>> tornado came through Kansas? ;) >>> > I guess you've never seen "The Wizard of Oz". > > >> You made me dig my latest mobo's manual :) [Gigabyte GA-M56S-S3], which >> says plainly : "The four DDR2 memory sockets are divided into two >> channels and each channel has two memory sockets as following : >> Channel 0 : DDRII_1, DDRII_3 >> Channel 2 : DDRII_1, DDRII_4" >> > You typo'd that. It should be: > > Channel 0 : DDRII_1, DDRII_3 > Channel 1 : DDRII_2, DDRII_4 > > Yes, I typoed; my bad. >> Please refrain from making such "wise" comments as , even if you're >> right, you're damaging Debian's image and yours. >> > I'm not sure how you come to that, but, ok.. > > Gigabyte has screwed up the DIMM slot color coding on the GA-M56S-S3, at > least according to the manual. I just looked at Intel, Asus, and other > Gigabyte boards and they all follow the correct alternating color pattern. > > Aioanei, you've based your entire argument on your knowledge of a single > main board. This board was incorrectly manufactured, or the manual text is > incorrect. Yes, manufacturers do make mistakes. Look at the ATA_HORKAGE > blacklist sections in libata-core.c to see lists of hard drives the > manufacturers screwed up. All of the Goliaths are in there, Hitachi, > Maxtor, Seagate, Western Digital. They're all shipped drives that don't > meet one spec or another. If these Goliaths can screw up, I guarantee you a > tiny company like Gigabyte can, and probably does so far more often. > > Now, go look at the DIMM socket colors and layout pattern on the Intel, > Asus, and yes, other Gigabyte boards to see what the proper color coding for > dual channel memory sockets is, and look at the text descriptions of the > slot and channel numbers. This GA board is an anomaly, and is flawed in > this regard. > > Take a look at the RHS of my email address. That's my personal domain. > They call me TheHardwareFreak for a reason. ;) > > The odds against me making a factually incorrect statement here about > hardware are so high you should place a standing bet with your local bookie. :P > > Anyway, http://en.wikipedia.org/wiki/Dual-channel_architecture says it better. Obviously, a "tiny company(?)" like Gigabyte is wrong and you are right; then how come all the systems I've put together with dual-channel memory work as expected? -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/4BC860EE.6030204(a)gmail.com
From: Camaleón on 16 Apr 2010 09:40 On Fri, 16 Apr 2010 16:06:54 +0300, Aioanei Rares wrote: > On 04/16/2010 04:31 AM, Stan Hoeppner wrote: (...) >> Now, go look at the DIMM socket colors and layout pattern on the Intel, >> Asus, and yes, other Gigabyte boards to see what the proper color >> coding for dual channel memory sockets is, and look at the text >> descriptions of the slot and channel numbers. This GA board is an >> anomaly, and is flawed in this regard. (...) > Anyway, http://en.wikipedia.org/wiki/Dual-channel_architecture says it > better. Obviously, a "tiny company(?)" like Gigabyte is wrong and you > are right; then how come all the systems I've put together with > dual-channel memory work as expected? O.k. both of you are right. Gigabyte AMD based boards seems to follow "0-0-1-1" scheme and Intel ones goes "0-1-0-1" but that's all. In both cases, same colored slots indicate where to put the matching RAM modules to get a dual-channel setup. It can be worst, though. There are manufacturers that paint all the RAM slots in black or using the same color and so forcing us to read the manual >:-) Greetings, -- Camaleón -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/pan.2010.04.16.13.31.10(a)gmail.com
From: Hugo Vanwoerkom on 16 Apr 2010 15:30 Aioanei Rares wrote: > On 04/16/2010 02:33 AM, Eduardo M KALINOWSKI wrote: >> On 04/15/2010 07:59 PM, Stan Hoeppner wrote: >>> Aioanei Rares put forth on 4/15/2010 3:16 PM: >>> >>>> No, it means that 2 one-gig modules in same-coloured slots will >>>> theoretically work better. I have one module in a dual-channel mobo and >>>> it works ok. Most desktop/workstation mobos do. >>> No, that's not correct. If the mobo DIMM slots are colored, you put one >>> DIMM in each color slot for dual channel operation, not both DIMMs in >>> the >>> same colored slots. >> >> I believe different manufactures follow different conventions. For >> some it's the two in same-color slots, for others in different-color >> slots. When in doubt, check the manual >> >> > One, I work with hardware everyday and I've assembled more computers > than I can count. Two, Stan, I find your remark pretty offensive, > really. Gigabyte, Asus, Intel mobos work by the principle 'same colour > == dual channel'. Please check your facts before trying to sound elitist. > > Checking the wiki page again, it says, "If the motherboard has two pairs of differently colored DIMM sockets (the colors indicate which bank they belong to, bank 0 or bank 1), then one can place a matched pair of memory modules in bank 0, but a different-capacity pair of modules in bank 1, as long as they are of the same speed." Which says to me that if I have a board with 4 memory slots of which 2 are colored light-blue and 2 are colored black and 2 1GB modules of identical speed, then I place both modules in either the light-blue slots of the black slots. Hugo -- To UNSUBSCRIBE, email to debian-user-REQUEST(a)lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster(a)lists.debian.org Archive: http://lists.debian.org/hqadbr$agb$1(a)dough.gmane.org
First
|
Prev
|
Next
|
Last
Pages: 1 2 3 4 5 Prev: Cannot Connect to Debian BTS Next: Configuring exim4 to send emails |