Jump to content

Corsair TwinX1024PT (Hynix D5) will not work on AMD.

Recommended Posts

Well....i have finished my testing.


This ram will not post on anything higher then 215 on AMD machines. I started off with my Opty 144 and my NF-D...highest i got it to boot windows was 216. Very loose timings and 2.7v. This was not solid for me as i was getting weird BSOD from time to time. I cranked it down to 200FSB and change the timings a bit (5-3-3-2.5) and its running ok.


Next is my neighbors 2 systems. First one is a A64 3000 (winchester) with a Epox board. This machine runs stock as its just some extra parts he had kicking around. I swapped his generic PC3200 and started going to work. I turned down his multi to 6x and up'ed the FSB from 200, 1 FSB each time. Highest i got was 210. I was not able to change the timings at all from 8-4-4-3 , 2t. At 210, 8-4-4-3, 2t, 2.8v the system booted into windows. 211, didnt make it past post. 212 wouldnt even show up post when the bios restarted after the setting changed. Had to clear CMOS and try again.


His second machine is alittle nicer. Its a 3200 (venice) at 2.6 (260) on a Neo2 Platinum board. He has TCCD in there now, cant remember which brand...its running at 260FSB 7-3-3-2.5, 1T, 2.7v. This machine has been running rock solid since the venice came out. I remember going to his house when he got it. So i quickly swapped out his ram for mine and cleared cmos. I pretty much did the same exact thing as before. Started the machine with a multi of 6x and synced up the ram. Starting at 200 i went up by 1 FSB each time. Timings were the recommended 8-4-4-3, 2t, 2.8v. System booted at 210, got just past post at 211 and didnt even get past post at 212.


Third and final machine tested was a newer machine that my friend built a few weeks ago. Its a x2 3800 (2.6 each) on a DFI NF-D ultra SLI version i think. He had the same ram i had before this corsair...his was just PMI mine was made by PQI. However, both used TCCD437. Same thing again, i turned his machine off...did a 10min clear of CMOS and went to town. This is the same board as me, however the CPU is different as i have a opty he has a X2 3800. The funny thing is, his wont even post at 200. Anything i tried, timings, went to 2.9v, swapping slots (yellow and orange).


Ok...now i only used 1 Intel machine but i think thats all i needed...i just needed to know that it wasnt the ram that everyone said...This ram performed awesome and totally within spec on my buds intel system. I cant remember the exact specs, but his ram right now runs 265, 8-3-3-3. We dropped my Corsair right in without a CMOS clear and changed the timing to 8-4-4-3. It booted right up into windows. NOT 1 issue with that. Since it worked there, there was no sense going on more. I just needed to prove that it was NOT damage to the ram. IT WAS NOT cause of removing and replacing Heatspreaders.


One person said that even removing the heatspreader can drop the overclock. This has to be bull crap. Maybe if it was physical damaging to the chips, however this would leave some type of traces of damage. My HS was replaced with a nice copper set and i removed that and looked at the chips with a magnifying glass. NOT 1 bit of Damage to anything...


NOTE TO EVERYONE: THIS ram WILL NOT GET HIGHER THEN 215 on a AMD system. Even that is giving it a bit much. There is maybe 1 person out the the 1000's of owners that have got this to work. The main reason why im saying this is I DONT CARE WHAT ANY CORSAIR REP SAYS...THIS WILL NOT REACH 275 on a AMD system. I pissed cause i even asked on Hardocp.com forums when the deal was going down. The Rep there told me that there shouldnt be any problems. Guess what? Wrong again...i was told that since there was no AMD system at the time it came out, there was nothing to test it. Ok, thats fine and great...but when it did come out...DID they bother to test it? NO...so i bought something thinking that ok...the rep cant be wrong.


No offense to any of the Corsair reps personally. I just wish everyone would have their facts straight before they say "Yes it should work fine" Did you test it personally? If not, then be on your way. I did a week of testing, driving from house to house....going from machine to machine. If you do that and get it to work on say half the AMD systems you use and can verify each, then i might believe it.


If you have any thoughts or comments please feel free to post....im trading this ram for BH5 and never buying another Corsair product agian. Not that i dont like their products...im sure i would if i had a intel system...but the lack of knowledge of their reps has ruined it for me. If they tested it...prove it to me...cause i dont belive it...i think they just saying that it should work without knowing for a fact that it does. You say it can run...PROVE IT!

Share this post

Link to post
Share on other sites


I did use Corsair TwinX1024-PC4400C25PT 2x 512MB CAS 2.5-4-4-8 @275MHz 1:1 @ 2.8V with my 3800 X2 DFI Ultra-D

I don't know which chips it has, i bought it August 2005


Here some settings i wrote down some time ago

I don't remember which of the settings worked for me



Corsair 4400PT (CMX512)

DRAM Frequency Set............................ - 200=RAM/FSB:01/01

Command Per Clock (CPC)....................... - Enable

CAS Latency Control (Tcl)..................... - 2.5

RAS# to CAS# delay (Trcd)..................... - 04 Bus Clocks

Min RAS# active time (Tras)................... - 08 Bus Clocks

Row precharge time (Trp)...................... - 04 Bus Clocks

Row Cycle time (Trc).......................... - 09 Bus Clocks

Row refresh cyc time (Trfc)................... - 16 Bus Clocks

Row to Row delay (Trrd)....................... - 03 Bus Clocks

Write recovery time (Twr)..................... - 03 Bus Clocks

Write to Read delay (Twtr).................... - 02 Bus Clocks

Read to Write delay (Trwt).................... - 03 Bus Clocks

Refresh Period (Tref)......................... - 3120 Cycles

DRAM Bank Interleave.......................... - Enabled


DRAM Frequency Set - 200 (DRAM/FSB:1/01)

Command Per Clock (CPC) - Auto

CAS Latency Control (Tcl) - 2.5

RAS# to CAS# delay (Trcd) - 04 Bus Clocks

Min RAS# active time (Tras) - 08 Bus Clocks

Row precharge time (Trp) - 04 Bus Clocks

Row Cycle time (Trc) - 09 Bus Clocks

Row refresh cyc time (Trfc) - 19 Bus Clocks

Row to Row delay (Trrd) - 03 Bus Clocks

Write recovery time (Twr) - 02 Bus Clocks

Write to Read delay (Twtr) - 02 Bus Clocks

Read to Write delay (Trwt) - 04 Bus Clocks

Refresh Period (Tref) - 2560 Cycles

Write CAS Latency (Twcl) - 05 Bus Clocks

DRAM Bank Interleave - Enabled

Share this post

Link to post
Share on other sites

If that were the case, i think out of 5 systems that were tested, not 1 of them worked.


For the LAST TIME...the C25 is NOT HYNIX RAM, its TCCD...not speaking about any other corsair ram but the PT. Its Hynix D5 chips. IM NOT SPEAKING ABOUT ANY OTHER RAM BUT THE PT....as this is the only one that uses Hynix. Please NO MORE posting that you TwinX1024C25 will do 275...in fact it should get higher since its TCCD.



I just hate it that corsair reps on the board say that it should work...well, have they personally tested it? Then when i said the HS was replaced...what you think the rep said...Well removing the HS prob damaged the ram...Hmm? Unless you use a screwdriver or something how exactly does removing a sticker damage the ram? I think this was just said as a way to get out of the fact that i was told by a rep that the ram should work, when in fact...its not cause Corsair couldnt test it...its the fact that it DOESNT WORK.


Here is link to PCstats review that i found...can you guess their outcome with the corsair and their AMD system....funny they seem to be experiencing the same exact problem....go figure.....http://www.pcstats.com/articleview.cfm?articleID=1561


They should just be honest with us and say hey, it doesnt perform well on a AMD...rather than have people buying thinking that it will perform as well.


I have tested 5 machine and they all shown the same exact thing. Highest it will reach is 215....FAR FAR from the spec which is DDR550.


In fact, out of the 2 boards that i have posted this on....1 person has had the Hynix D5 chip and got it over 230 on a AMD system.


REMEMBER ....NO C25....C25 = TCCD, PT = Hynix.

Share this post

Link to post
Share on other sites



There is nothing wrong with Corsair products.


The fact is that Hynix chips do not work well on AMD systems regardless of the module manufacturer.


Im not speaking badly about them as a company...i know Corsair has been around for a long time...im speaking bad at the reps and their techs...they say that it should work...in fact, they know it doesnt perform well...but instead of being honest and saying that it doesnt perform well, they let you buy it and find out for yourself it doesnt work. Sorry if i didnt explain that correctly.


Again, im not saying that all their products are bad...in fact this is only the 2nd time ive used corsair ram. Both times were completely below the rated speed on the stick. I thought i would give them another try...found these online for a good price, thought wrong again.


So its not the product..i know . happens....im just mad at them rather then saying it doesnt work, which im almost sure they know it doesnt...they let you waste your money.


Honestly, in my eyes...if they would say that this does not perform well on AMDs on the stats page...or have the reported speed on a amd...which is like 215 (DDR430?), i would of course not bought it, but would think about using their products again since they were honest.

Share this post

Link to post
Share on other sites

So here is the challenge...


I want to see screen shots of your overclocks on AMD systems and Corsair TwinX1024-4400PT ( for the last time...PT only guys...NO C25, NOTHING ELSE BUT PT). Doesnt have to be DFI board...can be whatever...just needs to be a A64 of some sort...no Bartons etc etc...since they dont care what type of ram they used really.


Ok...so MUST BE CORSAIR TWINX1024-4400PT with Hynix chips...nothing else accepted.


Sorry to be soo forward about the ram chips...seriously like 15 people have sent me PMs between 2 forums saying their TwinX1024-4400C25 runs at 8-4-4-2.5 and 270-275 =( ARRGGGGGGG....i replied the same exact thing to all of them...THATS TCCD!!!!!!!!!!!!


For verification i will need CPU-Z uploaded to the internet...it should show the exact same stats in SPD tab. (CMX51-4400). If you dont have it, download, its tiny.


It must be within the spec on the chip....275+ 8-4-4-3 or better.

Share this post

Link to post
Share on other sites

I have some of the memory in question here and I can test it later. However, you keep ranting about some Corsair rep and their statements about what will and won't work. I'm the only Corsair rep posting here and I have NEVER replied to you and NEVER stated that the Hynix based memory will work with AMD platforms.


Here is the only thread I have ever even participated in discussing these ICs/DIMMs: Hynix Based Thread


If there is any way I can help, I'll be glad to but, let's keep it factual.

Share this post

Link to post
Share on other sites

Also, you stated that you have bought 2nd hand memory that have had the heat spreaders removed. There is no way to know if the DIMMs have been damaged. Instead of blaming it on us, you should take responsibility for your actions. Granted, the Hynix DOES suck on AMD platforms but, you are only telling 1/2 of the story here and trying to make us look bad in the process.

Share this post

Link to post
Share on other sites

I wasnt talking about you, well...wait a sec...are you Redbeard on Hardocp?


I'm the only Corsair rep posting here and I have NEVER replied to you and NEVER stated that the Hynix based memory will work with AMD platforms.




If so then i am..sorry man. If you remember the post before on hardocp? You said the only reason why it said it was recommended for intel systems only, was cause there wasnt a AMD system on the market. Fine and dandy...when there was one, did Corsair or you bother to test it?


Basically, you said it would work...when in fact you just guessing and not knowing. Why couldnt you just said it then when i posted?


I dont care what you say about damaged ram. Ive been around a while...have used ram and never ever seen in my life ram being damaged from removing the HS. Its a giant sticker basically...its not like its soldered on the ram...In fact, its not even very tacky/sticky. People remove the heatspreader on ram on a normal basis. You ever try to fit ram with a Big Typhoon and the bottem ram slot? You almost need to remove it, or you can always exchange it for ram without a HS. How many people do you see saying...well ok, i cant use TCCD cause of the HS...i might as well sell it and buy something smaller. HELL NO, they remove it.


That was the FIRST post you said once i said the HS was removed. You stopped all contact and that was your last post.


So there is the whole story...if you like i can post the hardocp thread so you can see that i was the same person. You seem to know that Hynix was a problem..i asked the simple question...instead of saying there were problems (not 1 mention of it) you said it should of worked. Without knowing yourself that it does. You know your ram...you know what system i had (posted in my sig on every forum)...you know that Hynix D5 doesnt work on DFi, but instead of saying its known not to work...you said it SHOULD work. So my question ....Once AMD came out with a 275 FSB chip, did Corsair bother to test the Hynix?


Im very sorry if this wasnt you and you have to read through all that....but if you know who redbeard is, i would find out why his facts are different? If at that point he said that it wasnt a good match with my A64 system, i wouldnt of bought the ram, however would of bought another Corsair Product in the future.


Also, reguardless of you being redbeard or not...thank you for offering to test this for me. If you can post your highest clock via CPU-Z uploaded online that would be great...also show me a pic of your settings in A64 tweaker. Thats if you get it to work. I also HIGHLY doubt you will though. Also, if you can post all the stats of the testing machine, that would be great also.

Share this post

Link to post
Share on other sites

REMEMBER ....NO C25....C25 = TCCD, PT = Hynix.


the PT dont mean $hit. Its just means what kind of heat shields are on the ram


I got stuck with some Hynis D43s and they worked up to 290 but were a pain in the -ss


here find out what you got


But without the stock heat shields I cant see were you know what Ver they are so without that you really dont know

And i dont see no 4400s that ant pros so what is it you got Do you even know?

Best advice I can give at this point is look at the chips themselfs. and if your lucky there not relabled and go from there

Share this post

Link to post
Share on other sites

First, I am not Redbeard so it is not me you dealt with at HARDOCP. Second, you asked the question in such a way that it was not possible for him to give you solid advice. You basically asked "will it work" and he said yes. If he said it will work, that means we've tested it. He knows what he is doing.


So, when you stated you had issues, without hesitation, he offered to replace your modules. Only then did you tell him that you bought memory with the heat spreaders removed and thus a voided warranty. If you had told him about the spreaders before you made your purchase he would have told the same thing I would, buy at your own risk as they modules may be damaged and the warranty is voided.


I understand you are frustrated but, you made the decision to buy modules that were not recommended here at DFI AND that could not be RMAd. By omitting this little nugget of info you basically put DFI and Corsair in a position where they could not help you. I really wish I could but I can't. So, please don't try to blame all of this on Redbeard or Corsair.

Share this post

Link to post
Share on other sites

This topic is now closed to further replies.

  • Create New...