Corsair 8500C5D IC UPDATE

Printable View

Show 100 post(s) from this thread on one page

12-12-2007, 10:31 AM
Yellowbeard

Quote:

Originally Posted by KTE

How far have you pushed these in your labs? :D

2.2V 1200 4-4-4-4 PL5 1M stable?

See post number 2. Bachus has done most of the testing with this memory.
12-12-2007, 04:13 PM
KTE

I've read it all, he didn't really push it at all, it was all stock volts or below. I mean he must have added volts to at least one pair and checked where it goes as part of personal testing?
12-12-2007, 06:05 PM
[XC] 4X4N

I have a set of the new 2.1 version. They don't respond to voltage very well, or really overclock much, but they will do between 1050-1100 with only 1.9v. I only took them to 2.35v, but I couldn't get much more than 1100 @ 5-5-5-15. I have been running them with a Q6600 8x450 5:6 divider and only 1.9v. I have had many D9 sticks die on me with 2.2v or less,(including my ver1.1 of these) so I am pretty happy with what I have now.
12-13-2007, 09:36 AM
Yellowbeard

Quote:

Originally Posted by [XC] 4X4N

They don't respond to voltage very well, or really overclock much, but they will do between 1050-1100 with only 1.9v. I only took them to 2.35v, but I couldn't get much more than 1100 @ 5-5-5-15. I have been running them with a Q6600 8x450 5:6 divider and only 1.9v. I have had many D9 sticks die on me with 2.2v or less,(including my ver1.1 of these) so I am pretty happy with what I have now.

Yeah, they don't have the same scaling with voltage characteristics as Microns but, they do have their own strengths. And, anyone can easily use them with a CPU at 1:1 up to a 550 FSB. Not too many people going way past 550.
12-13-2007, 10:18 AM
kwalker

Quote:

Originally Posted by Yellowbeard

Yeah, they don't have the same scaling with voltage characteristics as Microns but, they do have their own strengths. And, anyone can easily use them with a CPU at 1:1 up to a 550 FSB. Not too many people going way past 550.

And thats the reason I would buy them again :up:
12-14-2007, 06:29 AM
owcraftsman

Quote:

Originally Posted by ZilverPhish

Some questions> I have a problem with booting on my mobo since it tends to revert to default voltage *1.8 or such, on my "old" Micron domis 8500 c5d. As I understand Quimonda runs at lower voltage, so maybe they could boot easier than the Microns`?, Or do I still have to keep a booting crap pair of ram around?

Are there any known problems with "mixing" these two types of ICs in the same system 2*1024 8500 c5d ver 1.x and 2*1024 8500 c5d ver 2.2? Compatibility wise as well as performace wise, without factoring in the generally known overclocking "drawbacks" in regards to 4 gig system memory setups.

Quote:

Originally Posted by ChaosMinionX

The newer revision have ICs similar to what you would call "crap pair of ram" in terms of voltage requirements so even if the BIOS kicked back to 1.8v those sticks would still boot at the timings BIOS defaults back on.

And you should have no problems mixing the 2 different revisions of the 8500C5D the only differences you will see between the 2 revisions now is their "theoretical" overclockability, they still might overclock well, but Micron is tried and true when it comes to overclocking and how well the IC's respond to voltage increases, and how well they handle lower timings and higher speeds and whatnot :)

I'm not sure they are compatible. Not Yet

I have both version mixed for the first time.

TEST BED

EVGA 680i A1 mobo P31 bios|Q6600|8800 GTS 640 sli|OCZ700GXS
CM2X1024-8500C5D 2-2GB kits v1.2 & v2.2

no OS yet booting to iso UBCD v4.1.1 http://www.ultimatebootcd.com/index.html

using Memtest86+ & Mersenne Prime Test to establish stability.

I was able to run both pairs individually dual channel mode 2GB config
@ 400x9 Link/sync 1600 A. FSB, SLI-Ready Mem.=Disabled
all spread spectrum & CPU features Disabled
Auto volts except vdim man set @ 2.15v
Mem. time= man set Expert 4-4-4-15-1T all else auto

Both pairs passed 6 hrs. prime stable & 6 hrs zero errors memtset

Now I did not expect that they would run together @ 1T in a 4GB config
but I did hope for 2T 4GB config at the low latencies.

NO JOY

Now I'm not saying this is a flaw I'm hoping I've made a minnor miss calulation
and someone can help me find the love.

Qusetions:
1) Am I wrong to go the low latency route or would you recommend higher FSB for best performance?
2) Is this an unrealistic expectation to pair these two revisions in the same bed @ low latencies or otherwise?
3) What volt would best suit this pairing?
4) What man sets to use? or all auto see where it goes....
mmmmmm I'm gonna try that but I digress
5) I've noticed the PCB HS seems to be somewhat different in that
it appear to be a different color or possibly grade. Could this account for the better thermal performance.

Please Advise!
12-14-2007, 11:44 AM
Yellowbeard

Quote:

Originally Posted by owcraftsman

Qusetions:
1) Am I wrong to go the low latency route or would you recommend higher FSB for best performance?
2) Is this an unrealistic expectation to pair these two revisions in the same bed @ low latencies or otherwise?
3) What volt would best suit this pairing?
4) What man sets to use? or all auto see where it goes....
mmmmmm I'm gonna try that but I digress
5) I've noticed the PCB HS seems to be somewhat different in that
it appear to be a different color or possibly grade. Could this account for the better thermal performance.

Please Advise!

1. I would go for the highest FSB you can get. Don't let latencies hold your CPU back. CPU cycles are good.
2-3. No, they should run fine together. 2.2v or 2.25v is fine when mixing them. See #1 for latency.
4. Manually set 5,5,5,15 and 2T, all else auto or defaults.
5. No, there has been no change in the PCBs that would affect the thermal performance.
12-27-2007, 02:03 AM
MavXS

Qimonda versus Promos chips

C4DHX XMS640X v2.x...PROMOS........64MX8...90nm
C4DHX XMS640X v4.x...QIMONDA......64Mx8...Rev.C

Which are preferable, the Promos or the Qimondas IC's?

Purpose: for 2 x overclocked Q6600 systems with P5K-E and GA-P35-DS3P M/B's.

Thanks
12-27-2007, 06:57 AM
Yellowbeard

Both of those are going to perform about the same and they'll easily run spec. They are not going to respond to voltage increases and OC quite like a Micron D9 set will.
12-27-2007, 01:21 PM
MavXS

Quote:

Originally Posted by Yellowbeard

Both of those are going to perform about the same and they'll easily run spec. They are not going to respond to voltage increases and OC quite like a Micron D9 set will.

That's understood and there was never a question about that issue, ie. a comparison with the D9's liking for voltage.

The question was if you had to pick between DHX's with Qimonda or Promos IC's, which kit would you choose based on whatever anecdotal feedback has come in on these different revisions of the same product line.
12-28-2007, 04:10 AM
Yellowbeard

I honestly have not compared the 2 so I can't make an educated guess based on any anecdotal feedback as I really have none. We don't get much hardcore enthusiast feedback on our non-Micron memory these days.

If I was forced to chose, I would take the Promos. IIRC, we have been using the Promos ICs longer than we have the Quimonda. The failure rates are VERY low so I know it is a very reliable IC. The OCing is limited somewhat due to the voltage regulation on the IC. But, I have no other evidence or experience to offer.
12-28-2007, 10:34 AM
KTE

I've had the C4 PC2-6400 v2.1 XMS2 (rated 2.1V) which are ProMos IC's. They dislike and do not scale with voltage above 2.35V but they run very cool, far cooler than D9's so dying at stock (overheating) is very unlikley unlike with Crucial RAMs.

Once you hit 2.6V on them or a little regular 2.4V the IC's will be dead and start throwing thousands of errors. This is regardless of temperature, doesn't matter if you have them at 1℃ all the way. However, this is a 2.35V run on them: http://valid.x86-secret.com/show_oc.php?id=268251
01-25-2008, 05:40 PM
hagatha

Because of the difference in voltage requirements between the two versions... can they be mixed?

I currently have a 1.* and I'm adding a pair... can they be mixed with the voltage differences?

If so... what to set the voltage to? 2.2 or 2.1?

thanks
hagatha
01-28-2008, 08:58 AM
Yellowbeard

Set them to 2.2v as the revision 1.* is tested at 2.2v. The extra .1v will not damage the newer RAM. If you want, you can always test at 2.1v to see if it is stable there.
01-30-2008, 03:54 AM
mad_skills

Quote:

Originally Posted by KTE

I've had the C4 PC2-6400 v2.1 XMS2 (rated 2.1V) which are ProMos IC's. They dislike and do not scale with voltage above 2.35V but they run very cool, far cooler than D9's so dying at stock (overheating) is very unlikley unlike with Crucial RAMs.

Once you hit 2.6V on them or a little regular 2.4V the IC's will be dead and start throwing thousands of errors. This is regardless of temperature, doesn't matter if you have them at 1℃ all the way. However, this is a 2.35V run on them: http://valid.x86-secret.com/show_oc.php?id=268251

I had a TwinX kit of those rev. 2.1 with ProMOS chips and i couldn' get them higher than 1000MHz or perhaps 1066MHz at some point, but they did rune really nice on speeds below that with very tight latencies. In my signature there are images of what i achieved with them, most impressive was SPi stability at 960MHz 4-4-3-5-2T.:cool:
Not so bad chips if you haven't got D9's.
03-07-2008, 02:05 AM
GunterFalstaff

Hey, i just bought 4 gigs of this dominator 8500 ram at the same time from newegg... and i got 2 different versions... version 3.1 and version 4.1 rated at 2.2v on thier label... but in cpu-z it says the 3.1 runs at 2.1v and the 4.1 runs at 2.2 is it wise to run these 2 different RAM together? will i have any problems from mixing them?
03-07-2008, 07:10 AM
Yellowbeard

Quote:

Originally Posted by GunterFalstaff

Hey, i just bought 4 gigs of this dominator 8500 ram at the same time from newegg... and i got 2 different versions... version 3.1 and version 4.1 rated at 2.2v on thier label... but in cpu-z it says the 3.1 runs at 2.1v and the 4.1 runs at 2.2 is it wise to run these 2 different RAM together? will i have any problems from mixing them?

You shouldn't but you may need to set the RAM timings and voltage manually. If they are not stable at 2.1v, 2.2v is safe for both versions.
03-07-2008, 07:13 AM
zfactor

wait you bought it as a kit? or do you mean you bough 2 4gb kits? you imo should not have recieved 2 different version in a matched kit..
03-07-2008, 07:18 AM
GunterFalstaff

no no... i bought 2 2gig kits (1gigx2)
03-07-2008, 07:22 AM
Yellowbeard

Quote:

Originally Posted by GunterFalstaff

no no... i bought 2 2gig kits (1gigx2)

If you have the option to exchange them, a 2 x 2gb kit is a much better solution. But it is up to you.
03-21-2008, 03:40 AM
jakku

Ver 4.2 realy :banana::banana::banana::banana: ?
04-06-2008, 08:18 AM
JimF

Ver 4.2

I just received Ver 4.2 also from Newegg.
(1) What chips do they use?
(2) What is the technology (e.g., 65 nm)?
(3) Will they boot at 1.8 volts?

I don't have a motherboard yet, so what timings can I expect at 1.8 volts?
04-06-2008, 11:50 AM
zanzabar

Quote:

XMS850X v4.X...PSC..............128Mx8 (2.1 Volts)

they are single sided powerchips, and if they are like the double sided kits they should do 1000 5-5-5-15 with 1.8
04-06-2008, 02:33 PM
JimF

Thanks. I missed that. That works for me.
04-30-2008, 06:52 PM
tbone8ty

what should tRC be set at?

in CpuZ is says EPP #1 setting should be at 22 but when its on auto in bios is shows 31

http://farm3.static.flickr.com/2169/...a422a0e5_o.jpg

i have them set at 2.1V in bios, will they do tighter timings? say 4-4-4-15-2T?

also any tips for OC these with the phenom 9850???

Show 100 post(s) from this thread on one page