[18th Anniversary] Asus Black Pearl - P5k-Premium Testing Report

[holwill]

In reply to haoster, mine is pretty much the same if not worse lol. But do not notice it any other time than running this. My board is 77 serial no




Both drives show Ultra Dma mode 5!

[jonny_ftm]

Ok, here's some more feedback

The 0612 bios gave me some headache. System was still stable at all bench test suites (Small FFT, OCCT in RAM mode, Vantage Suites...) but I expierienced random lockups at startup with the cold boot issue (usually signing an overclock failure). I think it is related to the RAM voltage. In fact, those extra performance compared to 0504 requiered more volts in my RAM modules.

Having to feed it now about 2.20v instead of 2.10v, I preferred for that voltage to use its full bandwidth.

I set a 4:5 ratio now instead of 1:1 to get a 1186MHz RAM at 5-5-5-15-4-45-8-6-6 (I shall tweak tRAS, trfc and other subtimings in a second time to improve memory copy performance) with 2.25v in the RAM now.
The system is rock stable, as usual (OCCT RAM, Prime95 small FFT, Futuremak suites and F@H). I increased vNB to 1.40 and vFSB to 1.30, all other settings are unchanged. In a second time I will try to set vNB and vFSB at lowest as they were in the setup with my signature

Here are the everest scores now, far better than my previous 950MHz CAS4 on the table in my signature



So, be advised, if your memory is running at its limits on bios 0504, 0612 may requiere some more juice in the DDR and NB.

Feed you back soon when the system setup is fully stable and the voltages (vFSB and vNB) tweaked to lower values

[jonny_ftm]

Ah, forgot to say these performances are explained by the strap... 266
Strap 333, CAS5 was crappy performance. Strap 266 at 1186MHz RAM and 475 FSB was not possible on 0504 with low voltages. Now, with bios 0612, this board can sustain those settings. I'll let you know if it handle them with lower vNB and vFSB, even if I'm already happy with stability at the actual voltages. I'll edit my signature soon when 100% finished tweaking

Really, intel should reinvent the P35 instead of flooding us with crappy X38 and X48 chips

[PiLsY]

Jonny_ftm - I just picked up on something you said in the thread that may help me. I got my P5k BP a couple of weeks ago and it ran fine. Ran for a week w/E6750 @ 7 x 533, then another few days @ 8 x 493. Tried it with an E8500 and got a bios update message. Updated to 0602 (latest at the time), but since then ive not been able to get over 390fsb (same as your P5b limit you mentioned?) stable or any memory dividers stable apart from 1:1 no matter what I do. Lockups, no post, blue screens, everything. I can post at up to 450ish, sometimes 475, but no stability at all. At 390fsb im running default volts, TB enabled +2, static read enabled and 3-3-3-3-2-20-2-2-2-2 timings (2.2vdimm), so mega aggressive. Going the opposite direction or full auto settings doesnt help fsb limit. No voltage changes make any difference either. Have tried diff cpus (E6420/E6750/E8400 and E8500, diff gfx cards (8500GT and 7900GTO), retail cooler and scythe infinity, passive cooled or active cooled, no hdds connected, tried everything and i'm getting nowhere!

I've since tried going back to 0204, 0602 and 0612 with exactly the same results. The only thing I have left to try is an EZFlash or afudos /pbnc flash as I updated using asus update, but I cant see this making a difference really.

Did you ever get to the root problem of your 390fsb wall on the P5b? This sounds very very similar to what im experiencing you see .

Im currently running at 8 x 390 @ 1.15vcore, so not cpu related. Mobo voltages dont help raise fsb, vdimm and timings have no effect either. Its driving me mad!

Many Thanks.

[jonny_ftm]

No, I never could pass that limit. I tried 3 Q6600 chips on two P5B deluxe, two memory types (Corsaire XMS2 PC2-6400 and Kingston HyperX PC2-9600). I never could get OCCT stable in RAM mode on my P5B deluxe above 390MHz and that's on 2 different motherboards

The P5K Premium I own rocks with same CPU and memory. The bios now detects 45nm chips and set some default unchangeable settings for 45nm CPUs, maybe that's causing hassle, as ASUS used to mess every thing with bios updates

[sesdave]

I tried increasing the FSB to 400. This lead to increase in core temperature by 6c. It reached a max of 68c under max load and 47c at normal load. I have set the following setting:

CMOS Ratio : 8
FSB Strap: Auto
FSB Freq: 400
CPU voltage: 1.4375
CPU PLL volt: 1.5
FSB term volt: 1.4
NorthBridge volt: 1.4

I don't know how to try reduce these voltage settings .

Your just unlucky with your CPu - it got a high VID and seems to want a lot of volts to overclock. Most should do 3.2 on stock volts and lowest settings or there about.. So if you want to clock that one you have limited option with your ambient temp. Check to see if the IHS and the HSF are flat. Lap whatever isnt flat. That should get some temp improvement which may give some overclocking headroom. Chilled cooling with your ambients would sort you out - but at a price!.. Overall you would be better off with a 45mm as they run cooler and need less volts. As a last go to try to get volts down and a 3.2 overclock try most settings on auto and see what CPUZ says for volts if it boots. Set the vcore to that voltage plus .025 for vdroop and try again. We can start a process of elimination for settings-lol!

AI Overclocking: Manual
CPU Ratio Control: Manual
- Ratio CMOS Setting: 8
FSB Frequency: 400
PCI-E Frequency: 100
DRAM Frequency: 800
DRAM Command Rate: 2T
DRAM Timing Control: Manual
CAS# Latency: 4
RAS# to CAS# Delay: 4
RAS# Precharge:4
RAS# Activate to Precharge:12
TWR: auto
TRFC: auto
TWTR: auto
TRRD: auto
TRTP: auto
DRAM Static Read Control: disabled

Transaction Booster: AUTO
Clock Over-Charging Mode: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

CPU Voltage: auto
CPU Voltage Reference: AUTO
CPU Voltage Damper: Enabled
CPU PLL Voltage: auto
DRAM Voltage: 2.2 or whatever yours need
FSB Termination Voltage: auto
North Bridge Voltage: auto
North Bridge Voltage Reference: AUTO
South Bridge Voltage: AUTO

Advance CPU Settings
CPU Ratio Control: Manual
- Ratio CMOS Setting:8
C1E Suppport: disabled
Max CPUID Value Limit: Disabled
Vanderpool Technology: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled
PECI: Disabled

USB Configuration
USB Functions: Enabled
Legacy USB Support: Enabled

[gmat]

Hey there xtreme systems - been lurking around for long here, and was on procooling years ago. Long winded post warning.

I just finished overclocking my most recent system, which is based around that nice P5K premium. Mine has some pink caps but the serial (on the last PCI slot) is 7Cxx... ? Serial would point at a newer board but i have 0404 stock BIOS, pink caps and older package.

Anyway, the stock heatpipe went off quickly, replaced by 4 EK waterblocks: NB, SB, mosfets (x2). The looks with the black topped EK blocks are rather good !
My system goes like this:
* 2 watercooling loops
* loop 1: CPU (Wolfdale 8400 under a DTEK Fuzion) + GPU (8800GTX under a Dangerden full cover block), D5 pump, Thermochill PA120.2 rad
* loop 2: NB + SB + Mosfets, DDC pump with EK top, no name chinese 120x1 radiator (looks like the BIX but thicker).
So it's 100% watercooled. I used all EK back plates for CPU and Mosfets mounts, so the mounts are flat, and i could use artic ceramique on the mosfets thanks to the reduced flex.

Temps with everything stock were so low it made me feel guilty. So i decided to aim for 4Ghz.
Flashed the BIOS to 612: good choice, the 45nm specifics are accounted for, and the OC setup is a bit easier.
My 8400 despite its lowish VID of 1.115 or something, requires no less than 1.35V at 3.6GHz and 1.40V at 4.0Ghz ! Got a mediocre sample (series 807... dont remember the rest, it's lapped now).

Some weird things that disturb me:
* 4GHz is totally unstable at 533Mhz FSB or 445MHz FSB ! (went up to 1.45V, no success). My memory is Corsair Dominator rated for 1066 so i didnt even push it (and indeed memtest runs flawlessly).
* I have an "island of stability" at 500MHz FSB. Maybe because it's all even multipliers (8.0x, memory on 1:1, etc) ???
* I cant raise the NB voltage over the default 1.25V, result is no post. The limits seem CPU related anyway.
* I really dont see what the GTL Ref voltages are meant for. They do really nothing at all ? What do the values mean ?
* Same for the Clock Over Voltage thingie (it's on auto)
* The memory timings labels in the BIOS dont match the labels on Corsair site or on CPU-Z. I'm at loss to find the tRC (Bank Cycle Time) which is rated at 22 for my Corsair TwinX 8500C2D (according to the SPD).

Temps now are around 35°C idle for both cores (no stuck sensor, woohoo) and 52-53°C full load under Prime95, with the effective 1.400 voltage (read by CPUZ, that means 1.425 in BIOS). The P5K-P board has that very nice feature to avoid the voltage drop under load, and it works wonders (the voltage is rock stable). The motherboard reports around 36°C now. Ambient is around 29-30°C now (hot day !).
Overall i'm very pleased with that Asus board. Excellent OC options, very user friendly, rock stable, and i love the small features they include to make life easier (like the OC profiles stored in BIOS).

[sesdave]

In reply to haoster, mine is pretty much the same if not worse lol. But do not notice it any other time than running this. My board is 77 serial no

Both drives show Ultra Dma mode 5!

I have a 76xx board and now run matrix raid 0/1 on my 2 old Hitachis - this setup spanks your single raptors for performance. With latest sata drives in a RAID 5/10 setup you can achieve upto 450MBs transfer rates and over 3000 - 4000Mbs burst with matrix setup! If you want best performance from drives do yourself a favour and join the matrix. http://www.ocforums.com/showthread.php?t=467848 Reverendmaynard did and he's one happy bunny

[ReverendMaynard]

I have a 76xx board and now run matrix raid 0/1 on my 2 old Hitachis - this setup spanks your single raptors for performance. If you want best performance from drives do yourself a favour and join the matrix. http://www.ocforums.com/showthread.php?t=467848 Reverendmaynard did and he's one happy bunny

Word ses. Incredibly reliable with that added RAID1 partition. Had a mass power failure last night and it did something to RAID1, but now the Intel Storage Manager is repairing it, without even noticing a glitch....in the Matrix lol.

[sesdave]

Word ses. Incredibly reliable with that added RAID1 partition. Had a mass power failure last night and it did something to RAID1, but now the Intel Storage Manager is repairing it, without even noticing a glitch....in the Matrix lol.

Yeah - it fun to watch intel doing something right! You really should get Acronis though if you want RAID0 recovery after a partition or total disk fail cos that is striped not mirrored. It works no problems from a partion backup on RAID1 or using recoveryCD and USB RAID0 partition backup in the event of a total drive fail - I assume you have that for you data backup anyway. I dont know of any other software that offers that for RAID.

[sesdave]

Hey guys, I'm having som serious problems with my P5K Premium Black Pearl Edition (sata HD related). If someone wants to help, please se this thread:

http://www.xtremesystems.org/forums/...d.php?t=186865

Cant help what appears to be a MB fault. RMA and also ask for your $65 dollars back from the shop that tested it. The premium ICH9R problems and drive compatability are well known and Asus will replace it no problem - your seller should be well aware of that. If they are so smart give them the MB and the raptor and tell them to give it back to you fully operating with the raptor connected.!

[ReverendMaynard]

Yeah - it fun to watch intel doing something right! You really should get Acronis though if you want RAID0 recovery after a partition or total disk fail cos that is striped not mirrored. It works no problems from a partion backup on RAID1 or using recoveryCD and USB RAID0 partition backup in the event of a total drive fail - I assume you have that for you data backup anyway. I dont know of any other software that offers that for RAID.

it's only my OS and games on the RAID0...no need to buy Acronis. Anyone who puts precious data on a RAID0 set up deserves to lose it

Oh and we tested some Acronis products for work, nothing even came close to touching ubuntu + fog + PING we image 500+ machine a day.

[sesdave]

it's only my OS and games on the RAID0...no need to buy Acronis. Anyone who puts precious data on a RAID0 set up deserves to lose it

Oh and we tested some Acronis products for work, nothing even came close to touching ubuntu + fog + PING we image 500+ machine a day.

Only suggested it as before you went Matrix you said you were not keen to reinstall as you had your Vista setup the way you liked it I too would not be using Acronis on an industrial scale - only my games PC . Its cheap and does the job . We use Tivoli and Altiris at work for imaging PCs and servers.

[ReverendMaynard]

Only suggested it as before you went Matrix you said you were not keen to reinstall as you had your Vista setup the way you liked it I too would not be using Acronis on an industrial scale - only my games PC . Its cheap and does the job . We use Tivoli and Altiris at work for imaging PCs and servers.

do tell? We are still looking for a solid PXE based mass image/data destruct solution. Are you using this in office, or for retail?

[sesdave]

do tell? We are still looking for a solid PXE based mass image/data destruct solution. Are you using this in office, or for retail?

Talking large office - IBm tivoli is a suite of products with numerous functions, imaging and rollout being one of them.Altiris is another products used for servers - they are used in the large financial companies I work for as an Infrastructure PM. Googkle them for more detail - speak to IBM they are always keen to talk when there money involved

[ReverendMaynard]

Talking large office - IBm tivoli is a suite of products with numerous functions, imaging and rollout being one of them.Altiris is another products used for servers - they are used in the large financial companies I work for as an Infrastructure PM. Googkle them for more detail - speak to IBM they are always keen to talk when there money involved

lol IBM can kiss my ass. I'm looking for a cheap solution.

[jonny_ftm]

Ok, here's my final results with bios 0612, fully tested and tweaked now

Bios 0612 gave me memory performance improvement (cf above posts and my signature link) but I begun getting some random hang-ups on windows startup. During daily use or benching it was still stable Prime95 small FFT and OCCT in RAM mode. I couldn't sort it out increasing vNB and vFSB. Finally, I realized it was memory issue, now needing looser timings (it became unstable) or more voltage (2.20 instead of 2.10v with bios 0504). Since I opted for 950Mhz CAS4 to have 2.10v on my RAM, and since I now had to give it 2.20v, I opted for a larger bandwidth setup (if it has to burn, at least it gets for respected performances ). Here are my expierience facts
- 8x475 at strap 266 with 1186 DDR2 CAS5 and 2.25v were perfectly stable on this new bios. It also gives a good performance increase compared to strap 333. I didn't need to boost my NB or FSB voltages

- 8x480 at strap 266 with 1200 DDR2 CAS5 and 2.25v needed 1.30v FSB instead of 1.20v to stabilize the CPU at 3.84GHu instead of 3.80GHz without adding vcore. So, now, this is my new setup, still rock stable and no more hang ups on boot.

- My Q6600 is fully stable P95 small FFT at 3.90GHz (9x434MHz) at PLL 1.5v and vFSB 1.30v but a vcore of 1.47v. I couldn't stabilize the 3.9GHz at 8x488 or even 8x485 despite any increase in vFSB, vPLL, vNB... It could be the CPU, the memory or the motherboard limiting me. I'd have to try strap 333 to exclude memory from equation and to loosen NB stress to partially exclude motherboard. Maybe I'll test that later, but no real benefit as I won't gain much CPU speed and the memory will loose performance

Here are screens of this new setup. Note that I opted for shorter periods of benching, especially with OCCT in RAM mode. OCCT will burn the RAM, and I've never seen errors after 2h of run in RAM mode. Also, at the end, you'll find all bios screens for this config

I could tweak the trfc and other subtimings (it seems stable with a trfc of 25 and settings of 5-5-4-10-4-25-3-3-3), but it only gave a poor boost in memory copy/read performance, so not worth going the hassle of OCCT RAM stability sessions to tweak it













I also edited my signature with my actual results, if it could help some one to boost his board

[R@ptor =]]]

Guys..what is this options?
FSB strap to north bridge
CPU and nb gtl voltage reference
DRAM Static Read Control
Clock Over-Charging Mode
CPU Voltage Reference
CPU PLL Voltage
Load line calibration
with a 450fsb on my E6750 in this mobo, whats values i have to set in this options?
thanks!

[jonny_ftm]

Guys..what is this options?
FSB strap to north bridge
CPU and nb gtl voltage reference
DRAM Static Read Control
Clock Over-Charging Mode
CPU Voltage Reference
CPU PLL Voltage
Load line calibration
with a 450fsb on my E6750 in this mobo, whats values i have to set in this options?
thanks!

I won't go into technical description. There are many threads here covering them, and also anandatech reviews. I'll go rather on important notes:

Strap: is NB latencies, the lower will give better memory performance, but can limit memory overclock. So, you need to find the sweet spot using Everest benches to reach a good compromise between memory bandwidth and latencies

GTL ref voltages: are safe to play with, use the maximum setting, than when stable, try to lower them. On a C2D they won't impact much, so keep them on smallest setting

Static Read Control and Booster: are related to memory performance level and latencies: you have to find, like with strap, the best spot for performance. Usually, activating both on very high memory overclocks is impossible

Clock Over-Charging Mode: Keep lowest setting on the P5K, won't help you in OC

CPU PLL Voltage: don't exceed 1.6v for your C2D, usually useless, the most harmul voltage for the CPU with vFSB. I'm sure you won't need more than 1.5v

Load line calibration: to bypass vdroop. Can be helpful if you use high vcore on a quadcore. But for under 1.40v of vcore and on a C2D, I won't activate it to limit the overshooting problems

[R@ptor =]]]

I understood everything...thanks a lot!
But tell me any reference to these options with 450fsb in my cpu and my memos at 1050mhz 5-5-5-15 ....please???
thanks

[jonny_ftm]

I understood everything...thanks a lot!
But tell me any reference to these options with 450fsb in my cpu and my memos at 1050mhz 5-5-5-15 ....please???
thanks

I won't make the effort to search for your exact setup as it'll cost me much time, and would be useless, as it will depend on the motherboard/ram/cpu chips quality you have, like with CPU batches (people forget that the NB is just like a CPU, and its performance can varry depending on batches/samples

To start with your setup, you can go with these settings:
FSB strap to north bridge 333
CPU and NB GTL Voltage Reference 63% and 67%
DRAM Static Read Control: Off (enable later)
Clock Over-Charging Mode: 0.70
CPU PLL Voltage: 1.50v
Load line calibration: on
vNB 1.25
vFSB 1.30 to start than try to lower to 1.20 to see if still stable (this is to limit needed vcore)
vDDR: depends on your RAM modules, but 2.20-2.35v depending on the sticks you own

Later, you can try strap 266 if still stable and enable SRC/Booster and see in everest which combination suites you

[dingmel]

all right, i'm gonna ask a really noob question here - just a word of caution, as i ran a quick yahoo search and came up with vague answers ; here goes :

I'm considering a 2nd 8800 to try my hand at SLI on my P5k-P. i've read from various forums that intel chipsets don't support SLI. what does that actually mean?

my initial idea was due to the limitations of the P35 (having 20x lanes) that the 2nd 8800 will be capped at 4x, but would be no biggie, meaning that SLI will still run, but performance will be capped on that 2nd card.
then i came across another thought from something i read iirc "all intel chipsets with the exception of ST do not support SLI" - meaning they just won't run 2 cards together or linking them with a SLI bridge won't enable SLI.

could someone shed some light on this? i'm actually loving my P5k-P a lot, and its pointless to go X48 / P45, seeing as there are negligible gains really, overclocking wise.

Thoughts?

[impar]

Greetings!

For SLI you need a nVidia motherboard.
You can have Crossfire, though.

[Leeghoofd]

all right, i'm gonna ask a really noob question here - just a word of caution, as i ran a quick yahoo search and came up with vague answers ; here goes :

I'm considering a 2nd 8800 to try my hand at SLI on my P5k-P. i've read from various forums that intel chipsets don't support SLI. what does that actually mean?

SLI is NOT officially supported on Intel chipset based motherboards, getting a 2nd card will only help you to steer more monitors but will not help gaming performance at all... some hacked drivers enable SLI on Intel mobo's but these drivers are way dated and will prolly not work in modern games or with modern cards...

No need to buy a 2nd Nvidia card mate, will not help at all... save ya cash...

[dingmel]

SLI is NOT officially supported on Intel chipset based motherboards, getting a 2nd card will only help you to steer more monitors but will not help gaming performance at all... some hacked drivers enable SLI on Intel mobo's but these drivers are way dated and will prolly not work in modern games or with modern cards...

No need to buy a 2nd Nvidia card mate, will not help at all... save ya cash...

ahah! thanks for that. thought so myself too lol. anyhow, i updated my thread on the 9650. i'm getting 400x11 at 1.485V actual stable. is that considered high?