Buffalo mT -5BC Results

[johan851]

Alright, some good news guys. I've been testing the sticks independently, and only one of them fits the characteristics described above. The second stick is Memtest86 stable at 230MHz 2.5-3-3-11 CPC Off @ 2.9v. Still not GREAT, but not bad at all either. I'm going to check to see if it runs rated timings of 3-3-3 at stock voltage. I'll be sure to post some more results after I test it more extensively. I feel good about this one.

I'll be returning the bad stick to Buffalo.

 

[Reefa_Madness]

did you try each stick individually? maybe its just one stick thats giving you problems. that would be nice to know cause it could just be one defective stick (more likely than two). if both sticks were defective i would start to wonder...

That was a fine audible.

I've been gone all day, took my 8 year-old to his first summer camp. He is fine, but moma is a wreck.

I am glad to see that you guys are starting to see some improvements from these chips, when tested individually. Is it just dumb luck that each of you got a bad stick of ram or is there more to this? It will take a few more days of people receiving and testing their ram before a better picture develops, but at least some answers to these performance (or lack thereof) riddles are surfacing.

I can't recall who purchased what ram from what vendor. Have the bad sticks come from epowerhouse or tblmemory, or both?

All in all I would say the news is encouraging. I don't have a personal stake, but a buddy of mine (the guy on vacation) did and did so at my suggestion, so I sorta do have an interest in how it performs. I don't want to have to tell him when he comes back that this ram doesn't do as well as I told him it would. He, however, is going to put it in an Intel rig so I'm hoping it won't be as much of an issue for him.

 

[johan851]

My bad stick came from TBL Memory. I think I'm going to shoot for an RMA directly through Buffalo. It'd be great to hear some more results from others though.

 

[Reefa_Madness]

It wont be long before guys like RedScare and far far away start getting delivery. I am quite sure that they will be contributing to the informal performance database that is being built up here. It is not as fancy as the one started on the Mobile XPs, but as long as it gives us the info we want, who cares?

Do you guys think it would be worthwhile to start a thread where the basic system information (mobo, cpu, type of cooling - air, water, etc,), the ram's revision and mfg. date and whether PC3200 or 3700, along with stable results, at various timings and voltages could be posted by each user? It could also include any comments and observations deemed pertinent.

I say this because several people have commented on not knowing where to post about this ram, due to the number of threads. With more and more of this stuff due to be delivered in the next few days, this might make it easier to find a place to post current information, instead of having a bunch of different threads started.

Actually, this one could become the "official thread", since it already has a good, general topic line and a good start on the listing. Any thoughts?

If it is decided to do this, we could just add a post in all of the others to follow a link to this one and consolidate, or if someone wants to keep their own going, for whatever reason, go ahead, especially if you're seeking assistance with specific issues.

 

[johan851]

Actually, this one could become the "official thread", since it already has a good, general topic line and a good start on the listing. Any thoughts?

If it is decided to do this, we could just add a post in all of the others to follow a link to this one and consolidate, or if someone wants to keep their own going, for whatever reason, go ahead, especially if you're seeking assistance with specific issues.

Sounds like a plan to me. I'll make the necessary adjustments to my first post.

 

[2coop]

240 at 2.6v is quite good but the loose timing may negate more than the high speed can compensate for. I don't know if that is true with the 64 or intel platform but I would think tighter timing on anyboard would be better.****edit***no, thats not 2.6v, but they do run error free at 230 with 3.0v in dual channel Cas 2 2-2-2-6 cpc on. ([email protected])

 

[far far away]

i didn't get my sticks today
maybe tomorrow, then i'll give them a try!

 

[johan851]

240 at 2.6v is quite good but the loose timing may negate more than the high speed can compensate for. I don't know if that is true with the 64 or intel platform but I would think tighter timing on anyboard would be better.****edit***no, thats not 2.6v, but they do run error free at 230 with 3.0v in dual channel Cas 2 2-2-2-6 cpc on. ([email protected])

Yes, tighter timings are always better, but sometimes better overall bandwidth can be increased by slightly looser timings (2.5-2-2-11, or 3-2-2-11) and much higher speed. Are those results with Micron chips? Or with BH-5?

 

[Chowdy]

My EB 3200 is pretty nice. Confirmed stable 11-2-23 @ 230FSB, 2.8V. I tried 24 11-2-2-3, but i get instability in windows despite memtest stability running test 5. I'm going to try a CAS of 2.5, as some have said that the NF2's dislike CAS3?

 

[johan851]

I'm going to try a CAS of 2.5, as some have said that the NF2's dislike CAS3?

Yes. I spoke with Buffalo Tech this morning, and CAS 3 is the reason that nForce2 platforms overclock poorly in comparison to A64 and Intel rigs. The i865 and i875 chipsets are spec'd to run 3-4-4 timings, while the nForce2 chipset is designed to run CAS timings of 2 and 2.5.

Those are some great results Chowdy.

 

[Reefa_Madness]

Catchfrase wrote in one of the other theads that he spent all weekend messing with his PC3700 with no luck reaching specs.

It just doesn't make sense that everyone would be struggling with this ram after it was supposed to have been "tested" to run at 233 and yet everyone seems to be having trouble, not just an isolated case or two. What gives? You would have thought that this would have been the cream of the crop.

We need the people from Buffalo that did the testing to do some explaining about their testing procedures, otherwise, there is going to be a truckload of RMA'd ram at their front door in the near future.

 

[Ramlaen]

so not what i wanna hear lol, mine arrives tommorrow hoping to not hit all these problems

 

[Reefa_Madness]

I agree, this is not what anyone wanted to hear, much less, experience.

I am still perplexed about the 3700, considering what some of the earlier 3200 was doing, with the same chips.

Buffalo Tech...what's going on here?

 

[johan851]

Not sure what's going down at Buffalo tech...for my RMA, Bruce is speed tesing the RAM to match what my other stick does. BTW, my good stick is running rated speeds - 230MHz, 3-3-3-11 @ 2.6v.

Most of the results have been with AMD platforms, haven't they? Remember that the nForce2 chipset isn't built to do CAS 3 well, and that AMD systems usually don't run high FSBs well either. Other than that...I'm out of ideas.

 

[Reefa_Madness]

Not sure what's going down at Buffalo tech...for my RMA, Bruce is speed tesing the RAM to match what my other stick does. BTW, my good stick is running rated speeds - 230MHz, 3-3-3-11 @ 2.6v.

Most of the results have been with AMD platforms, haven't they? Remember that the nForce2 chipset isn't built to do CAS 3 well, and that AMD systems usually don't run high FSBs well either. Other than that...I'm out of ideas.

I'm glad that you were able to contact Bruce and that he is trying to get you right.

Your comments about AMD systems not usually running high fsb is somewhat / mostly correct, except your DFI's claim to fame is just that, being able to hit in the 240 plus range, isn't it? I understand an NF7 struggling to hit 230 with this ram, they are somewhat inconsistent boards, they either seem to break 230 or struggle with 220. If I remember correctly, your board has gone 240 before, so it should not be unrealistic for you to be able to hit the 233.

I am not really critical of the ram, mostly perplexed and unsure as to what to expect from it. It would be nice to have a few people hitting at least 240-250 at 1:1 on Intel rigs, then you could at least say, OK, hit specs plus some, not bad for $225/gig and be done with it, you know?

 

[johan851]

Yeah, I hear you. From what I've seen though, if the sticks are good the results are good as well.

I understand an NF7 struggling to hit 230 with this ram, they are somewhat inconsistent boards

I'm just saying that some of the inconsistency might be inconsistent boards. It's a bit of a reach.

 

[Reefa_Madness]

I hope so. Everyone (me included) had high expectations from this EB value ram. Logic says it should do well, and in alot of cases has. I think that a "wait and see" what results others get, in order to have more test data, is the only thing we can do. After the dust settles we can then make a valid judgement.

By the way, I almost forgot to ask, did Bruce tell you when to expect your stick and did he indicate whether or not he could match it with a "C" revision?

 

[johan851]

He did say that he had 200 sticks in stock with which to RMA mine. He also let me know that he would be hand picking another stick to match the good one I have. I can only assume that he'll be replacing my "C" stick with another "C" stick. I can expect it shipped via Fedex 2 Day 1-2 days after he recieves mine. Mine will ship out tomorrow.

 

[wannaoc]

Oh you guys keep talking like this and now I'm expecting to rma this as soon as it gets here Thurs. Oh please, oh please let mine run AT LEAST spec! I may have to get ahold of Bruce as well if not.

 

[Rabbi]

Mine should be here tomorrow.

I have 3.2v at my disposal and this board has done done 1x512mb BH5 255FSB 2-2-2-11 and 2x256mb 43B 256 2.5-3-3-11 so I know the board won't be a limiting factor...... unless compatibility is an issue, in which case i will try one of the MANY modded BIOSes I have saved.

I'm making up some custom alluminium heatsinks to epoxy onto these sticks and i have 60mm and 40mm fans pointing onto my dimms... no excuses