Anyone have ASUS A8V Deluxe?

[The Pentium Guy]

Hey,

I'm getting an ASUS A8V Deluxe, and I'm wondering what kind of RAM you guys have (model number..etc) so that I can be sure I'm getting a compatible RAM. I want 1 GB of somewhat fast RAM (around $200), no value RAM.

Also are you guys using it in Dual Channel? If so what slots did you put it in? a1 + b1 or a1 + a2 (or even b1 + b2).

Thanks,
-The Pentium Guy

 

[ts3433]

You can't really go wrong if you buy from a widely-accepted brand like Crucial or Corsair. Check the memory compatibility charts on Asus' website if you want. You'll want to run in dual channel; consult your manual for the slots to put it in.

FYI, on an A64 you'll see no benefit getting "fast RAM" over value RAM. See Zebo's tests in the CPU/OC forum.

 

[The Pentium Guy]

Wow. That just changed my mind. I won't go with the CHEAPEST (In case I want to use the RAM on other machines), but I'll get some "mid-ranged" RAM.

But besides the compatible list, can anyone else add their ram to the compatible list (or a "NOT Compatible!!" list)?

-The Pentium Guy

 

[The Pentium Guy]

*bump*

 

[JRez]

I initially tried using some GeiL PC3200 RAM that worked (and still works) stably in a Pentium-4 rig. Nothing but BSOD's. When I switched over to Kingston (was on the approved list on ASUS' site), it's been stable and performs well. There's a thread somewhere in hot deals for HyperX @ Outpost for $140. Running dual-channel and used slots A1 and B1 as per the manual.

 

[The Pentium Guy]

Thanks guys.
The ASUS site says these are compatible:
CMX256A-3200C2PT
CMX512-3200C2

However, I don't want 256MB RAM so I just changed the model to 512 MB, The 2nd one (CMX512-3200C2) costs a little too much $135 for 512 MB).

So this is the one Im' thinking of getting
http://www.newegg.com/app/View...tion=20-145-420&depa=0
CMX512-3200C2PT

I hope this is OK? The only difference I see is the PT< whicich is the platinum coating (it looks cooler than the non-platinum one )


-The Pentium Guy

 

[lauy]

I would strongly advise you to stay clear of Crucial Ballistix PC4000 modules. I bought an A8V deluxe rev2 with a matched pair of Ballistix PC4000 and have had nothing but trouble when trying to achieve any speeds above 230 fsb.

I think the best memory would probably be the OCZ PC3200 Platinum rev2 or the Corsair modules are also very good in this board I believe.

Just a word of warning anyway about the Ballistix....

 

[adeathangel]

Hmm.. I have an A8V Deluxe, and I can't get my 4 sticks of PC3500 to run at DDR400 no matter what I do. They're Kingmax value ram, I got the value ram because I was told the good stuff doesn't improve performance enough to make it worth-while. I've heard that you can't get DDR400 with four sticks.... So, I'm thinking.. is a viable option to sell off my ram and get two 512MB sticks of some PC4000 in order to achieve better OCing? I'm seriously considering this, and if I should steer clear of the Crucial Ballistix PC4000, any suggestions on what PC4000 would be good? I've heard good things about OCZ, a model # would be good too. Thnx =P

 

[CrucialLabs]

Hello All,

I'm not sure why anyone would advise against using Crucial PC4000 Ballistix modules on the ASUS A8V Rev. 2 motherboard. I can assure you, I have tested this motherboard and I am currently running a pair of 512MB Ballistix modules at spec (500MHz @ 2.5-4-4-8)

Bearing in mind BIOS settings need to manipulated for the FSB, HT Frequency, Memory Bus and AGP/PCI Bus to operate in a stable environment.

Hardware/Software used for memory qualifications are as follows.
ASUS A8V Rev. 2 / 1009 BIOS
AMD64 4000+ (2.4GHz native @ 2.63GHz)
2x BL6464Z505 (500MHz @ 2.5-4-4-8 / 2.8v)
WesternDigital Raptor 74GB SATA
Crucial Radeon X800 Pro 256MB
CoolMax AP-550X (473w) Power Supply
Windows XP Professional SP/2 (Full network install)
CC2004 v1.0.1 w/SP2 Update (35.8 average)
SiSoft Sandra 2005 (RAM Bandwidth Int Buff'd iSSE2 = 5949 MB/s, Float Buff'd iSSE2 = 5960 MB/s)


BIOS Settings as follows:
HT Frequency @ 600MHz - 16bit/16bit
FSB = 250MHz
CPU Multiplyer @ 10.5 (10.5*250MHz=2.63GHz) CPU voltage 1.45v
Memory 1:1 @ 2.5-4-4-8
AGP/PCI manually set to 66.66/33.33MHz

 

[Slug]

CrucialLabs-- Can you run stable at those settings with an HTT multi at 4x?

Also check this thread. What are we doing wrong? I would love to run at 250 mem speed, with a lower cpu speed of course.

Thanks in advance for any input you may have.

 

[CrucialLabs]

Slug

I'll give it a shot running the HT Frequency at 4x (800MHz.) I'll report back as soon as I can on this test. I'll also exchange processors and rebuild using a 3200+ and 3500+.

Currently I am running two BL6464Z505 and two BL3264Z505 (still at 250MHz 2.5-4-4-8) to check on any unknown stability issues. **I did have to relax the multiplyer down to 10 from 10.5 for CC2004 to pass a Lightwave error**

Complete current settings are now:
HT Frequency @ 600MHz - 16bit/16bit
FSB = 250MHz
CPU Multiplyer @ 10 (10*250MHz=2.5GHz) CPU voltage 1.45v
Memory 1:1 @ 2.5-4-4-8
2T Command Rate: Disabled/Enabled (both settings run fine)
AGP/PCI manually set to 66.66/33.33MHz

I've just installed Doom3 and will have a run at some saved benchmarks as well as live gameplay (for testing purposes of course) and report back on this as well.

 

[CrucialLabs]

Slug:

I've changed the HT Frequency to 4x (800MHz).
New BIOS settings:
HT Frequency @ 800MHz - 16bit/16bit
FSB = 250MHz
CPU Multiplyer @ 10 (10*250 = 2.5GHz) CPU Voltage now at 1.525v
AGP/PCI manually set to 66.66/33.33MHz
Qty. 2 BL3264Z505 and Qty 2 BL6464Z505 (1536MB total) @ 2.5-4-4-8
CC2004 v1.0.1 w/SP2 update: avg. score 35.7
Doom3: timedemo +5 runs without issue and I "tested" game play @ 40min without issue
SiSoft Sandra 2005: RAM Bandwidth Int Buff'd iSSE2 = 3673MB/s, Float Buff'd iSSE2 = 3686 MB/s on an estimated memory bus of 4000MB/s running at 92% efficiency)

I'm going to rebuild using a 3200+ and 3500+ to see if there are any major changes. I do want everyone to realize that while I will assist in supporting Crucial Memory, I do not want to end up being ASUS motherboard support. I believe I've shown Crucial Ballistix PC4000's run fine on the ASUS A8V Rev. 2 motherboard. If by chance the different processors indicate other issues, I will post these as well.

 

[Slug]

Darn impressive. Thank you very much.

 

[bo3bber]

Why did your Sandra score drop so far? 2T timing perhaps?

Also, if it's easy, please run a looping 3d test overnight, just to add extra confidence in stability. I'm still skeptical that the A8V can run 250 reliably.

Thanks,
-bo3bber

 

[lauy]

Hey CrucialLabs

I have also tried the exact same settings that you are running. Are you running Memtest86 and Prime95 over long periods to show the stability? Because even though they seem stable in games, it doesn't necessarily mean the memory is completely stable. In my opinion, for it to be completely stable, there should be no errors in multiple loops of memtest86 and for Prime95 to pass completely over 12+ hours.

And as I've mentioned in my own thread, the difference between 2T and 1T is quite significant, and I can successfully run my memory at 250mhz @ 2.5-4-4-8 2T with cpu @ 250 x 10 . But who wants 2T when they've just paid the top price for premium overclocking ram? It should be able to do it at 1T and for me, I can assure you it doesn't. Whether it is the motherboard or the memory, I don't know because I haven't had a chance to test the ram in another board, but my guess is leaning towards the board.

Also, I don't know whether you know this, but using a half multiplier actually reduces the speed of the memory without you knowing. So you may think it is running at 250mhz, but if you have a half multiplier, it won't be.

I'm not trying to slag off Crucial or anything like that, but the simple question was asked about the A8V deluxe and memory compatibilities, and it seems that myself and a lot of others have been having issues with this board and the ballistix modules.

-lauy

 

[CrucialLabs]

All,

As it has been shown that the ASUS A8V rev. 2 motherboard and Crucial Ballistix PC4000 modules are able to run at specified speeds I?m bringing this to a close.

Regarding 2T and 1T Command Rates: Unfortunately these are not memory specifications. Many factors can be involved in what 1T/2T Command Rates, on any motherboard, will be including manufacture/process/board timings/noise/etc. to name a small few.

Please see the following for processors used: http://www.crucial.com/calculator/timings.asp

 

[bo3bber]

Actually, you didn't show that Ballistix can run stably at the specified speeds, so bringing it to a close seems awfully premature.

To demonstrate the memory works as expected, I still want to see:
1) 12 hour Prime95 test.
2) 3 hour 3d game testing.
3) multiple loops memtest86 (for lauy)
4) 250Mhz 1:1 memory specification
5) >6000 Sandra score (otherwise, why bother with Ballistix at all?)

I don't care about memory latencies so much, I personally don't require 2-2-2, because it doesn't affect the Sandra bandwidth that much. But, the 1T setting is a must, regardless of whether it's a memory setting or a motherboard setting.

If Ballistix cannot run those settings, then it's a poor choice for this board. Simple as that. (A bad value, in that the performance does not warrant the price.)


I'm not trying to slag Ballistix either, and if you demonstrate they work as specified, then I'm sure other readers and I will be _very_ interested. So far, you have failed at that task, which still leads me to believe that there is a problem.


From your prior post, If you can set this up:

HTb Frequency @ 600MHz or @800 - 16bit/16bit (at 3x or 4x, either is fine)
HTT/FSB = 250MHz
CPU Multiplier @ 10 (10*250MHz=2.5GHz) CPU voltage 1.45v (up to 1.6v OK)
Memory 1:1 @ 2.5-4-4-8 (3-4-4-10 OK too)
2T Command Rate: Disabled
AGP/PCI manually set to 66.66/33.33MHz

And then succeed at my tests above, then you've got a winner.

Please do those tests, and let us know. I for one would be interested in switching from my Corsair if you can demonstrate success.

Thanks,
bo3bber

 

[adeathangel]

I too am interested. I would most assuredly switch memory if those tests were to prove successful.

 

[lauy]

I would too. And tbh, it seems like Mr CrucialLabs won't even do all the tests and will just give some simple screenshots of cpu-z. Anyone could do that? hardly needs a stable PC.

I've been ripping my hair out with these Ballistix modules trying to get it to run 250 @ 1T and seems like I've been completely wasting my time. And on top of that, someone from Crucial won't even give minimal support and proof that they will run at those speeds stably with tests. >:|

Overall, not exactly impressive for high priced top premium memory from a leading memory manufacturer.

 

[bo3bber]

I sent my comment directly to Mr Crucial, and haven't received any response, nor any comment in this thread for three days.

No one is questioning whether Ballistix are good or not, just whether they work in this board. See:http://forums.anandtech.com/messageview.aspx?catid=29&threadid=1469337


I take his silence as a tacit admission that Crucial Ballistix PC4000 modules are a poor choice for this board. Beware.

-bo3bber

 

[pizzamampf]

Hello,

i´ve got a similar problem with my adata vitesta ddr500 ram.
Anything above 220 MHz occurs errors.

`CrucialLabs´ writes:

CPU Multiplyer @ 10 (10*250MHz=2.5GHz) CPU voltage 1.45v
Memory 1:1 @ 2.5-4-4-8

My question is: What does Memory 1:1 mean for him ?

If 1:1 is switched in BIOS, this means RAM is at 1/2 of FSB.
To run RAM really at same speed as FSB he must switch to 2:1 in BIOS.

This point perhaps explains his low results with sandra.

- pizzamampf

 

[CrucialLabs]

I apologize for the delay in responding to the most recent posts about Ballistix?s performance. Rest assured that there is no subterfuge in the delay; it is simply a matter of workload. Our department has been very busy in recent days, and you must realize that responding to forum issues is only one of many aspects of my job in the Crucial Performance Lab.

While we have our own stringent requirements for in-house memory validation, I did decide to go ahead and run bo3bber?s tests as requested because I knew that Ballistix does meet the parameters he outlined. (But please be advised that we cannot continue to run individual testing requests, nor can we serve as ASUS motherboard support, as I mentioned previously.)

It does seem puzzling that even with Lauy?s POST of ?I can successfully run my memory at 250mhz @ 2.5-4-4-8 2T with cpu @ 250 x 10,? Crucial is still being questioned about Ballistix?s validity in this motherboard

Here are the results to the tests that bo3bber requested:
1. 12-hour Prime 95: Finished ? 19 hours. 1 error that cannot be attributed to memory.
2. 3-hour 3-D game testing: I ran 3DMark03 build 360 looped for 3 hours and ?tested? Doom3 for 45 minutes straight ? NO ISSUES to report.
3. Multiple loops Memtest86 (for Lauy): Not sure what constitutes ?multiple loops.? Memtest86 would run only when the CL was changed from 2.5 to 3. (All other tests were done at CL2.5).
4. The BIOS settings were: 250MHz 1:1 memory/fsb.
5. >6000 Sandra Score (otherwise, why bother with Ballistix at all?)

Screenshots can be found here: http://www.crucial.com/calculator/timings2.asp

I hope that this provides bo3bber?s with the information he?s searching for and quells fears surrounding Ballistix compatibility in this board.

 

[Slug]

Could be a combo of things, ie- the winchester cpus and the board, or even all three together. I have tried every combo of settings to try to get 250 stable with 1:1 and I simply cannot do it, even at cas 3. Again, I seriously doubt it's just the ram. I have toyed with the idea of getting a Neo2 platinum to test with.

CrucialLabs, Am I correct in assuming stock cooling on mobo and ram? No extra fans blowing on components?

 

[dfinder]

bump

 

[CrucialLabs]

Slug,

Motherboard was setup on a test bench / open air with stock CPU heatsink.

Memory modules had active cooling similar to this one shown at: http://reviews.pimprig.com/cooling/super_talent_memory_cooler.php since I did not have it in a case under any type of circulation. I did notice a huge decrease in surface temps when using the active cooling.