FYI Possible AMD A64 Memory Controler issue discovered

[imported_Aelius]

I got my new system up and running. Will try to OC soon and see how well it works out.

I'll post an update this week.

 

[Markfw]

edit: forgot I already replied to this...... sorry.

 

[DAPUNISHER]

Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Keep in mind that PC3200 RAM might not overclock well (depends on a lot) and it may have squat to do with the memory controler. It could very well be the RAM.

Real nice, how about you put the disclaimer in the first post or edit the sensationalistic thread topic? :roll:

 

[AWhackWhiteBoy]

Sounds a lot like the Barton problem and how everyone blamed it, and all it took was a GOOD motherboard design to get around this. How about the new DFI mobos that are almost on the market that have been getting 550mhz?

 

[DoobieOnline]

I think this issue mostly affected those with the early CO revision Clawhammers, and/or mobo's without AGP/PCI locks. The later CO rev chips seem to OC just as well as the CG rev. The newer CG revision Clawhammers and Newcastles seem to work fine at higher than 220 FSB with core/mem at 1:1. My 3000+ Newcastle hits 10x250, 1:1 with 2x512MB OCZ PC3200 EL Platinum Rev2 which has the new Samsung TCCD chips. My DTR 3200+ Clawhammer CG rev is able to do 10x240, 1:1 with the same RAM. This is all done with the Soltek NF3 250GB board.

doobie

 

[Zap]

Originally posted by: DoobieOnline
I think this issue mostly affected those... mobo's without AGP/PCI locks.

I was just thinking this as I read this thread. The real test is to put one of those CPUs that seem limited to 220FSB into a motherboard/RAM combo that has already been tested to a higher FSB, such as the combo that DoobieOnline has with the NF3-250GB chipset. I think very few people got high FSB clocks until the boards with AGP/PCI locks became available (K8T800 Pro, NF3-250, NF3-250GB, NF3-150 on certain boards).

 

[LocutusX]

Originally posted by: DAPUNISHER
Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Agreed. This thread smells like a load of expired baloney to me.

 

[Neptune3000]

wow, this is interesting. I recently overlcoked my athlon 2800+ to 3375+ pr. It runs perfectly at 250x9 with stock everything. Temps dont even pass 50C

my HT is default so is my voltages. I purchased some crucial DDR 3200 from bets buy for 159.00 thought i got screwed but i did'nt for it.

 

[imported_Aelius]

Originally posted by: DAPUNISHER
Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Keep in mind that PC3200 RAM might not overclock well (depends on a lot) and it may have squat to do with the memory controler. It could very well be the RAM.

Real nice, how about you put the disclaimer in the first post or edit the sensationalistic thread topic? :roll:

How about you take your testy little attitude and shove it. Come back when you have something constructive to add.

 

[imported_Aelius]

Originally posted by: LocutusX

Originally posted by: DAPUNISHER
Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Agreed. This thread smells like a load of expired baloney to me.

I smell trolls.

I believe your cave is back the other way towards P&N.

 

[imported_Aelius]

Originally posted by: Zap

Originally posted by: DoobieOnline
I think this issue mostly affected those... mobo's without AGP/PCI locks.

I was just thinking this as I read this thread. The real test is to put one of those CPUs that seem limited to 220FSB into a motherboard/RAM combo that has already been tested to a higher FSB, such as the combo that DoobieOnline has with the NF3-250GB chipset. I think very few people got high FSB clocks until the boards with AGP/PCI locks became available (K8T800 Pro, NF3-250, NF3-250GB, NF3-150 on certain boards).

The problem is that it's happend on boards with with known working AGP locks. Assuming the board isn't defective. That remains to be seen.

OCZ is not one to jump on a bandwagon and even they have recognized this as a legit issue, hence they are trying to figure it out with AMD.

Maybe there's a workaround or maybe if it's a fault in the process then AMD will be able to correct it.

I don't expect an RMA but it's nice to know that finally someone is looking at this issue that is constantly overlooked as "lucK" when it may not have anything to do with luck at all.

EDIT: Forgot to add. So far the only update I seen on this issue is that some kinds of RAM will work on some CPUs that did not work well with some kinds of RAM where the CPU was thought to be at fault but that has to do with a lack of troubleshooting and not the absence of the problem. Some have tested their CPUs on multiple known working motherboards with different kinds of RAM and still it wouldn't go beyond around 220. No real udpate yet from AMD or OCZ except that they are looking at it and are gathering info from various sites. Maybe we will know in 6 months.

 

[LocutusX]

Originally posted by: Aelius
How about you take your testy little attitude and shove it. Come back when you have something constructive to add.

His question was valid... where's the proof? Or is it because you have none that you must resort to flaming?

Until you provide some, this thread is really nothing more than a bunch of flagrant whiners. Whiners led by... you!

And trying to get an RMA because your A64 won't go above 220HT? Ha, that's just pathetic.

 

[DAPUNISHER]

Originally posted by: Aelius

Originally posted by: DAPUNISHER
Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Keep in mind that PC3200 RAM might not overclock well (depends on a lot) and it may have squat to do with the memory controler. It could very well be the RAM.

Real nice, how about you put the disclaimer in the first post or edit the sensationalistic thread topic? :roll:

How about you take your testy little attitude and shove it. Come back when you have something constructive to add.

Still waiting for you to add something constructive, besides bullsh!t speculation. :clock: Where's the proof? BTW, OCZ is well known as one of the worst trolling companies ever! so I'd believe the problem is with their ram before I'd believe the problem is with a reputable well respected company like AMD. Don't like it? To fvckin' bad, 'cause there ain't a damned thing you can do about it 🙂

 

[imported_Aelius]

Originally posted by: LocutusX

Originally posted by: Aelius
How about you take your testy little attitude and shove it. Come back when you have something constructive to add.

His question was valid... where's the proof? Or is it because you have none that you must resort to flaming?

Until you provide some, this thread is really nothing more than a bunch of flagrant whiners. Whiners led by... you!

And trying to get an RMA because your A64 won't go above 220HT? Ha, that's just pathetic.

Don't associate me with your kind. If you got a valid point then make it without restorting to childish personal attacks. What do you expect when someone addresses you with a confrontational comment? I don't take it litely and don't appreciate it.

So until you provide something more than troll bait I suggest you take your attitude elsewhere. If you wish to comment on this issue in a more civilized manner I will be more than happy to respond in kind.

 

[imported_Aelius]

Originally posted by: DAPUNISHER

Originally posted by: Aelius

Originally posted by: DAPUNISHER
Where's the empirical data to back this up? I must have missed it, and I don't see anything in the OCZ forum@bleedingedge I'd call proof positive.

Keep in mind that PC3200 RAM might not overclock well (depends on a lot) and it may have squat to do with the memory controler. It could very well be the RAM.

Real nice, how about you put the disclaimer in the first post or edit the sensationalistic thread topic? :roll:

How about you take your testy little attitude and shove it. Come back when you have something constructive to add.

Still waiting for you to add something constructive, besides bullsh!t speculation. :clock: Where's the proof? BTW, OCZ is well known as one of the worst trolling companies ever! so I'd believe the problem is with their ram before I'd believe the problem is with a reputable well respected company like AMD. Don't like it? To fvckin' bad, 'cause there ain't a damned thing you can do about it 🙂

Do us all a favor and stop thread craping. Do I need to dumb it down for you or what? Take a hint.

Either respond respectfully, I'm happy to discuss issues in a civilized manner, or piss off.

 

[DAPUNISHER]

You make demands you can't enforce so just stop now 😉 The chronology of this thread is as follows

AMD A64 Memory Controler issue discovered

but then your own posts go on to issue disclaimers that it's not known if it is in fact a memory controller issue or not. You even go so far as to say that maybe in 6 months we will know. Now, how is that not at odds with the thread title? 😕 It states that there is a issue, yet the thread contains no evidence beyond some peoples best guess. I requested you edit the title to reflect this, you became surly. I gave as good as I got and you returned fire=fair enough.

Now, might I suggest we dispense with the unpleasantries and defensiveness on both our parts, and you explain why the thread title says there is a issue with the memory controller and the substance of the thread does not? Or why there is no solid evidence to support the title? Those were my only objections, why the thread didn't contain hard evidence, and why the speculatory nature of the "issue" isn't reflected by the title.

If you can't answer those 2 questions with a valid response I will take my leave of this discussion as you dersire, certain that you are just spreading rumors like the Inq does and not informing we your fellow forum members of anything useful. Answer them with a valid response and I'll happily appologize for my offensive behavoir towards you 🙂

 

[acivick]

Just for the sake of adding some info, as opposed to bickering, I just got the new DFI board and some OCZ PC4000 Gold Rev2. Paired it with my 3200+ NC and I can't seem to get it past 225-230. I've tried a bunch of different settings, but Memtest (typically during test #5) reports errors at or below 230.

So far here's what I've gathered:
Not the RAM (tried Corsair PC4000 as well)
Probably not MB (DFI does have working locks and should be able to handle it)
Most likely not heat (have lowered CPU multiplier AND I'm watercooling)

What it COULD be:
Raidmax PSU (only provides 15A on the 12V rail, will be putting in an Antec PSU in the next day or two)
CPU/memory controller (admittedly less likely than the PSU, but I'll provide an update once I've changed PSUs)

 

[imported_Aelius]

Originally posted by: DAPUNISHER
You make demands you can't enforce so just stop now 😉 The chronology of this thread is as follows

AMD A64 Memory Controler issue discovered

but then your own posts go on to issue disclaimers that it's not known if it is in fact a memory controller issue or not. You even go so far as to say that maybe in 6 months we will know. Now, how is that not at odds with the thread title? 😕 It states that there is a issue, yet the thread contains no evidence beyond some peoples best guess. I requested you edit the title to reflect this, you became surly. I gave as good as I got and you returned fire=fair enough.

Now, might I suggest we dispense with the unpleasantries and defensiveness on both our parts, and you explain why the thread title says there is a issue with the memory controller and the substance of the thread does not? Or why there is no solid evidence to support the title? Those were my only objections, why the thread didn't contain hard evidence, and why the speculatory nature of the "issue" isn't reflected by the title.

If you can't answer those 2 questions with a valid response I will take my leave of this discussion as you dersire, certain that you are just spreading rumors like the Inq does and not informing we your fellow forum members of anything useful. Answer them with a valid response and I'll happily appologize for my offensive behavoir towards you 🙂

Misery loves company.

Thread title updated to be more appropriate for the discussion on the subject.

 

[acivick]

As it turns out, my PSU was not to blame. I still get similar results, despite going to this Antex PSU. Has to be the CPU and/or memory controller.

 

[imported_Aelius]

Originally posted by: acivick
As it turns out, my PSU was not to blame. I still get similar results, despite going to this Antex PSU. Has to be the CPU and/or memory controller.

It's not easy to figure out what's wrong exactly. I got my system fully up and running but gota wait for the thermal paste to go through the burn-in process. After that I'll OC.

 

[MOHO456]

I have C0 and mine is fine at 230+

 

[imported_IKnowNothing]

Here is some info that might help.

Link: Hard|OCP Overclocking Athlon 64 3000+