Intel D875PBZLK

[Sandog]

Originally posted by Paragon
Guys,

I think most if not all the memory problems we have experienced with the D875PBZ are totally related to not having adequate vdimm...2.5 volts just seems inadequate when most manufacturers recommend 2.75-2.95 volts for their current lot of memory...even Mushkin in their forum recommends 2.75 volts...That's why I have such high hopes for the Desktop Control Center which is hopefully not some apparition...Why would Intel target the D875PBZ and the D865PERL only for this product if they didn't realize there were some problems...I don't think Intel just had a dream one night and decided that they want to start catering to the enthusiast market...hopefully this new product if and when it materializes will help significantly with memory issues. Just my $.02

I think your right because why else would Intel come out with a program that lets you change memory voltages. At least I hope you are right. I'm at 5-2-3-2 right now and 5-2-2-2 wouldn't hurt.

[CrowdAds]

[Sandog]

I tell you what I think. Intel is trying to satisfy the people who are looking for PAT or who have bought low latency RAM and expect it to work at SPD but it doesn't. They are providing a solution by giving us a way to up the voltage to the memory. I mean who is gonna buy the mediocre DDR400 memory that is listed on Intel's memory certified sheet. Anyway I have no problems with my board. If Intel throws in some extra FSB speed besides 4% I will be the first to OC this 3.0C. It's looking up from here.

By the way friends, I hope you realize as I do that this is totally new ground for Intel. Their practice has always been you get exactly what you pay for. Nothing less, nothing more. For them to offer voltage mods (we'll see) or FSB adjustments is new. The D875PBZLK board is going down in Intel history one way or another lol.

[Sandog]

Originally posted by Ralf Hutter
Didn't anybody notice that the release notes also talk about BIOS P10 and P11, released on 8-15-03 and 8-19-03 respectively? Where are they? They're not in the archives or on the ftp site. Wierd.....

Maybe they are way ahead of themselves.

[Sandog]

Originally posted by AHLnut
The Santa Cruz is one awesome sounding card for audio purposes. For games, in side by side comparisions, the Audigy 2 is the hands down winner in my opinion. The driver for the Santa Cruz is not nearly as intrusive as Creatives.
For the Audigy, I just install the driver and I can use my mic. What I do is go into control panel,sounds and audio devices, then voice then test hardware. It does it all for you then.

All I want to do with my mic is record what you hear. I play the file I want to record on my speakers and thats it. If it does that then I will do a complete uninstall/restall like you said.

[Sandog]

Originally posted by Riptide
I had problems getting OCZ memory working properly with this board and that was my suspicion as well: that it was a voltage issue. OCZ spec'd the memory @ 2.6v and as soon as I ran into the spec for the board (2.5) I figured that was probably the problem..

The Mushkin runs fine, including the earlier pair of PC3200 I used. It's spec'd to run from 2.5-2.75 which is why it probably played better with this motherboard.

I prefer a BIOS fix for these adjustments vs. a software fix, but whatever it takes to get this problem solved is a step forward I suppose.

I prefer bios adjustments to. Maybe later Intel will incorporate it. I've come to the conclusion we are only halfway thru the bios updates lol.

[opdirect]

Originally posted by Sandog
I tell you what I think. Intel is trying to satisfy the people who are looking for PAT or who have bought low latency RAM and expect it to work at SPD but it doesn't. They are providing a solution by giving us a way to up the voltage to the memory. I mean who is gonna buy the mediocre DDR400 memory that is listed on Intel's memory certified sheet. Anyway I have no problems with my board. If Intel throws in some extra FSB speed besides 4% I will be the first to OC this 3.0C. It's looking up from here.

By the way friends, I hope you realize as I do that this is totally new ground for Intel. Their practice has always been you get exactly what you pay for. Nothing less, nothing more. For them to offer voltage mods (we'll see) or FSB adjustments is new. The D875PBZLK board is going down in Intel history one way or another lol.

Competition forces companies to do things that they rather not do. Maybe Intel figured out that IT departments (bread and butter) are not going to go around overclocking their machines to make a 2.4 a 3.0 and risk stability and possible voided warranties. This application must be a response to Abit, Asus, and everybody else selling adjustable boards. This is too big of a market to ignore anymore. $$$ We're the guys buying the expensive 3.2 processors. (wll maybe not me. 2.4 cheapie)

I can't imaging Intel coming out with an application just to bump up DIMM voltages to manufacturer recommendations. Wouldn't they just make the DIMM voltage adjustable and leave the FSB at 400 and below? That would protect their processor market but allow some adjustment for the 875/865 memory issues. If something fried, they could always blame it on the RAM manufactures.

Maybe with this application they are testing the water. If it turns out to be a bad idea for them, they just won't provide the application for their next mobo. Maybe it's just a corporate policy thing. It seems easier to nix an application than to justify an overclockable D875PBZ BIOS and then have later mobo releases non-overclockable. Now that the put "Burn-in" in the BIOS, I think everyone is going to expect that to be in the BIOS of future releases.

[AHLnut]

All I want to do with my mic is record what you hear. I play the file I want to record on my speakers and thats it. If it does that then I will do a complete uninstall/restall like you said.

I will gladly check that for you when I get back to the computer with the Audigy 2 after this weekend.

[fire1]

Hi !

i have flash now to Bios P12 and i see a new entry in the Boot Bios with the name :Scan User Flash Area can you me say what this is make it enable ore not ?? Is this for the new controll panell other for the the new cpu prescott . Where has information for me


Thx fire1

The PC is not faster other slower i find with this BIOS version.

[Miles]

Well maybe we are making progress towards the Desktop Control Center, but the fact remains that PAT is still disabled in CTIAW and CPUz and benches are back to where they were with P09

[fire1]

Originally posted by Paragon
Well maybe we are making progress towards the Desktop Control Center, but the fact remains that PAT is still disabled in CTIAW and CPUz and benches are back to where they were with P09

Yes i think is not the performence bringer this bios ;(
have you an answer for this option in the boot: Scan User Flash Area ?? What is this
Who has INTEL Bios P10 and Bios P11 in the AREA 51 other in a dungeon .

Will flash back soon to P09 other P05

greetings from VIENNA fire1

[Sandog]

Originally posted by fire1
Yes i think is not the performence bringer this bios ;(
have you an answer for this option in the boot: Scan User Flash Area ?? What is this
Who has INTEL Bios P10 and Bios P11 in the AREA 51 other in a dungeon .

Will flash back soon to P09 other P05

greetings from VIENNA fire1

I'm not sure but scan user flash area might deal with the logo storage area. I assume it would check there for a custom logo you could put up instead of Intel's logo.

[Miles]

I flashed back to P05, P12 was very stable but still no PAT...P05 just seems to perform better in most software, nothing more than a subjective thought here.

[Ralf Hutter]

Okey-dokey, got BIOS P12 last night and wanted to see what's up with PAt being enabled or not. Had P05 on my test rig so I ran an entire series of benchmarks prior to flashing to P12.

So I flashed (from DOS, uneventfully) to P12 and booted back into windows. checked PAT with CPU-z 119, no big suprise, it says PAT is disabled. Ran all my benchmarks, will post numbers below:

=====================================================

Here's the hardware that I used for this test:

P4 3.0C running at default speed (no burn-in)
2 x 256MB Mushkin PC3200 Level II RAM @ 2-3-3-6 in the "A" slots

on this board:

D875PBZ "Revision #2"
Serial #BZBT33006812

This is a brand-new board. It came with BIOS P05 and it's been running Prime95 for the last 14hrs until I did this little experiment.

=====================================================


I ran Sandra Max3 and PCMark2002 benches with P05 and again after I flashed to P12. Here's the lowdown:

Results:

Sandra Max3 v2003.7.9.73


Math: P05-9138/5712, P12-9154/5382

Memory: P05-4756/4789, P12-4721/4724 <----------------------------------------LOOK :eek: RAM benches with P05, "PAT ON" are almost identical to P12 "PAT off" benches!!

MultiMedia: P05-13814/21723, P12-13815/21692

======================================================

PCMark2002 build 100:


CPU: P05-7378, P12-7354

Memory: P05-8132, P09-8022<----------------------------------------LOOK :eek: RAM benches with P05, "PAT ON" are just slightly better than with P12 "PAT off"!!


HDD: P05-878, P09-875

======================================================

Cachemem benches:

P05: 304/3.0 = 101.3333

P12: 319/3.0 = 106.3333

Damn near the same as each other and much better than the last set I ran on P05 vs. P09. That was 106 vs. 120-ish.

======================================================

ctiaw says PAT is "Fully Enabled" for P05 and "Reserved" for P12

======================================================

CPU-z 1.19 says PAT is "Enabled" for P05, "Disabled" for P12

======================================================


Conclusion:

PAT is still disabled in P12 but Intel's done something to increase the RAM performance up to where it's real close to the PAT enabled version. Pretty neat. I'll stick with P12 on this board and play with it for a while because there's effectively no difference in the RAM performance between the new and the old BIOSes.

[TQ]

Good work, RH. Very much appreciated.

[Miles]

I tried flashing in DOS as you did Ralf and reran the benches, here are mine:

Sandra Mem: P5 - 4817/4758 ... P12 - 4755/4722
========================================
Sandra Math: P5 - 8576/5201 ... P12 - 8548/5028
========================================
Sandra MM: P5 - 12898/20154 ... P12 - 12900/20254
========================================
AIDA 32: P5 - 4732/1493 ... P12 - 4579/1549
========================================
Cachmem: P5 - 302/2.8 = 107.9 ... P12 - 332/2.8 = 118.6
========================================

P4 2.8C running at default speed (no burn-in)
2 x 512 MB Mushkin PC3500 Level I RAM @ 2-3-3-6 in the "B" slots, I guess in the grand scheme of things the result differences are less neglible than I first thought