articles » current » information-technology » software » bios » intel-d815eea-motherboard

BIOS for Intel D815EEA/D815EPEA Motherboard

Download BIOS versions P10 and P11 for the Intel D815EEA/D815EPEA motherboard.

Click here to download BIOS versions P10 and P11.

BIOS ENGINEERING Intel Corporation

5200 NE Elam Young Parkway Hillsboro, OR 97124-6497

DATE: July 18, 2001 FROM: DPSD BIOS Engineering

SUBJECT: EA81510A.86A Standard D815EEA & D815EPEA BIOS Production 11

About This Release

  • EA81510A.86A.0051.P11.0106190714
  • 815 Video BIOS 6.2 Production Version 2313
  • ICH2 PXE Ver. 83 Features/Errata Fixed in This Release


  • Made the "Language" setup item non-editable when user setup access level is set to "View Only".
  • Fixed issue to prevent the BIOS from reporting false information in the Extended System Configuration Data (ESCD).
  • Fixed issue where the Unattended Start password must first be entered to unlock the keyboard so that the Setup password could be entered. The correct behavior is to let the user enter into the Setup Password without the need for the unattended start password upon entrance into setup.
  • Fixed issue of video corruption at the bottom of the screen when booting from a floppy during unattended start.
  • Corrected video corruption and system hang when the setup question "Fixed Disk Boot Sector" is set to "Write Protect" and an OS attempts to write the boot sector.
  • Fixed issue with non-functional PS/2 keyboard and mouse when booting to Windows* 2000 or Windows NT4.0 with unattended start and legacy USB enabled.
  • Corrected issue with USB pointing devices not working correctly in USB legacy mode.
  • Fixed issue of yellow explanation mark appearing in device manager when PS/2 devices were missing.
  • Fixed USB Keyboard POST issue.
  • Fixed "Entering Setup" appearing at the bottom of the screen after pressing F10 while in Setup.
  • Added string translations for PARTIES, USB boot, and BBS 3.0.
  • Added support for PCI IDE Bus Mastering (DMA) for BIOS INT 13h hard disk reads and writes on IDE devices that support IDE Bus Mastering.
  • Added support for doing 32-bit IDE PIO mode data transfers inside BIOS INT 13h.
  • Added support for version 1.11 of Intel Express BIOS Update
  • Updated ICH2 code to account for ICH2 BIOS Specification Update Revision 1.8 documentation changes 8.0 and 9.0 (B4 Stepping Support).
  • Fixed APM so that drives would correctly power down in APM mode.
  • Added new code to do USB polling.
  • Fixed problem with Adaptec 3940 resource conflict with Windows 2000 when using an AGP video.
  • Fixed detection of multi-channel CNR audio upgrade.
  • Fixed issue of Windows S3 loosing some USB devices when connected to USB hubs.
  • Fixed issue of BIOS not detecting PME wakeup correctly.
  • Corrected translations for Wake on LAN, Wake on PME help strings.
  • Added "Mode 5" (UDMA/100) option to the IDE UDMA Mode BIOS SETUP question.
  • Added support for displaying the maximum UDMA mode supported for auto-typed IDE drives in BIOS SETUP.
  • Fixed issue of BIOS incorrectly reporting UDMA modes from IDE devices that do not support UDMA.
  • Passwords are no longer updated when custom defaults are loaded.
  • Fixed auto-display of IDE Multi-Sector Transfers option.
  • Added generic OPSD PARTIES implementation.
  • Corrected functionality of the LBA Mode Control option.
  • Added Intel 815 video BIOS VBIOS 6.2.
  • Added setup option to BIOS to allow users to Enable or Disable USB boot.
  • Fixed issue of shift keys not working in DOS with a USB-only keyboard configuration.
  • Fixed incorrect display of a hard drives capacity for larger hard drives.
  • Updated CNR support to comply with CNR 1.1 Specification.
  • Added new release of PXE code. (Intel Boot Agent 4.0.17 containing PXE build 83)
  • Fixed a problem where the event log would not display correctly if Quiet Boot was disabled.
  • Added AMI BBS 3.0 and USB boot functionality.
  • Added prevention of using Blocked unsupported processors.
  • Added support for latest generation processors.
  • Added update for Pentium III 1.1 GHz processor support.
  • Changed POST screen reporting of ATA/66 and ATA/100 IDE devices to display "Ultra DMA Mode ATA/66" or "Ultra DMA Mode ATA/100" respectively.
  • Fixed issue of BIOS not reporting the resources for, PS/2 keyboard and Game Port, incorrectly.


  • Addresses the Pentium III processor E76 Erratum


  • Added new Processor update for Pentium III processor.
  • Updated the Intel Copyright String.


  • Banner change to P09-0040.
  • Added support for programming the back-panel diagnostic LEDs to all green on resume from ACPI Suspend-to-RAM. We were not programming the diagnostic LED on S3 resume and it would come up with default values according to hardware strapping.
  • Fixed a SMBIOS problem where the Type 9 System slot - "current usage" field always showed up as "unknown" instead of "Available"/"In Use" for the AGP slot.
  • Resolved issue with waking up from ACPI S3 where some video cards would hang the human interface to the system. B-0039
  • Made password entries case-insensitive.
  • Enabled 2nd USB host controller to respond to wake events.
  • According to the ADC Self-Test (v2.4) and the ICH & ICH2 BIOS specs, Delayed Transaction Enable (DTE) should be set = 1.
  • Changed Secondary Master Latency Timer setting. According to the ADC Self-Test (v2.4), the Secondary Master Latency Count should be set to 0x4.
  • On resume from S1 (Standby), OS software, as part of its normal power management path, enables the USB host controllers to generate an interrupt (not SMI) as part of transactions occurring on its ports. This would not usually be a problem, but because the USB Legacy Keyboard/Mouse Control register is shared between USB HC 1 and 2, the act of enabling HC interrupts (USBPIRQEN) affects both host controllers.
  • Fixed a problem where POST did not display the correct amount of memory if more than 1GB of memory was used in the system.
  • Fixed PCI subsystem device/vendor ID programming for onboard video.
  • Fixed initialization of OPROM area in RAM. The code wasn’t making D8000-DFFFF R/W before writing FFh’s to it. Now we write FFh’s from C0000 through DFFFF.
  • Fixed string corruption interaction with the floppy interface with ABSCMOS.
  • In Multi_Sector_Transfers_Proc and Ultra_Dma_proc added check for ABSCMOS running.
  • ADC Self-Test Version 2.4.12 found several 815 and ICH2 registers that were not at recommended values. Thus changed the 815 and ICH2 registers to recommended values.
  • Added support for Standard 3 SKU with no VGA connector.
  • Removed unnecessary AMR-specific code. One path in the code was only necessary for AMR support. It isn’t ever executed for AC97ONLY or CNR. I removed that path and optimized the routine. This change does not apply to ICH. We can make this change for ICH2 because we have no ICH2 AMR products.
  • Fixed a problem where the SMBIOS information for the onboard Audio (On Board Devices Information TYPE 10) always returned "Device status" set to Enabled.
  • Insure that the PCI Configuration register is really restored when returning from SMI.


  • Added audio codec signature.
  • Fixed CMOS corruption error.
  • Fixed a problem where hard drives in DMA mode would not return from S3 with the default Microsoft IDE drivers. The reason for this is that the commands that are passed back to the OS to initialize the drive are executed before the drive is ready. To solve the problem we are waiting for the drive to be ready before passing the values back to the OS. This fix also includes a fix to the timings for ATA-100 drives. This may increase S3 resume time, as we now need to wait for the drives to become ready before we give any IDE data to the OS.
  • Added SMBus to the Slot Characteristics 2 of PCI slot 2 SMBIOS data structure.
  • Fixed a security issue, that started with BBS 3.0, so that in system config (Maintenance) mode when Unattended Start is enabled the system still enters Setup.
  • Fixed a security issue, that started with BBS 3.0, so that when Unattended Start is enabled the system will not boot from floppy or CD-ROM.
  • Resolved issue with some ATAPI CD-ROM not being seen when an ATAPI Zip drive is attached to the same channel.
  • Fixed a problem where an FDC error was displayed during POST if a SCSI/ATA100 card was connected to the system, and the user had never entered setup.
  • VBIOS 6.0A Production Version 2269
  • Initialization of ARMD limited setting of ARMD device emulation type to either floppy or hard disk. B-0036
  • Fixed issue where bootable Windows 98SE CD-ROM would hang during boot if CD-ROM was before hard disk in the BIOS boot order, and the user selects boot hard disk from the Windows 98SE CD-ROM boot menu.
  • Previously the CTRL, ALT and SHIFT keys were being reported as stuck keys if held down. Also the shift state bits at 40:17 were being cleared as part of USB init. Now, CTRL, ALT and SHIFT keys (6 keys total) will not be recognized as stuck keys, the keyboard error will not be reported if they are held down during POST. The 4 shift state bits at 40:17 are preserved at USB init. The OS can use these keys held down during POST to enter Safe Mode, bypass startup files, etc.
  • Removed obsolete AC97 code. This code was only needed for AMR support. B-0035
  • Added ICH2 workaround: Assertion of the ICH2’s RTCRST# signal does not reliably reset all RTC-well register bits to their default states as documented. Some RTC-Well configuration bits may not be restored to their expected default states when booting after RTCRST# assertion. BIOS’s already should be checking RTC_PWR_STS bit during POST (this bit is not affected by the issue). If this bit is set, BIOS should explicitly reset all RTC-well register bits to the desired default states. D-0034
  • Fixing up our drive table late in POST so Windows can find everything.
  • Fixed Memory Size Decrease error reports to the correct CMOS Battery Low.
  • Fixed the issue of Bios not displaying date and time error when CMOS battery is corrupted.
  • Fixed a problem where the CNR LAN was disabled coming out of POST.
  • Checked in the translation for some untranslated strings. And updated string translations.
  • After all ATA/ATAPI devices are initialized, Security Freeze Lock command is issued to all ATA hard disks if the respective device supports Security Mode Feature Set. After this command is executed, the device rejects all other security mode commands until next power cycle. This may help eliminate the possibility of installing password into the devices by any un-authorized entity. This is the default implementation of the Security Mode Feature Set. However this default implementation can be changed according to OEM requirements through the hook provided.
  • Changed the Timer 0 from mode 2 (rate generator) to mode 3 (square wave generator). This is the default AT setting defined by IBM. Awhile back this timer was changed to mode 2, because of possible performance improvements.
  • Under the IDE screen menu for each drive the informative text was always showing up. Now it disappears when a drive is not present, or is selected as type None. This was accomplished by converting the static text items to dynamic text items.
  • Rearranged Power Management Setup Options to clarify under which context each option functions. The Setup Options having to do with power management were often confused so this change is to clarify which options are used under ACPI, which are used under APM, and which are used regardless of power management mode.
  • This fix obsoletes the MKF_VIDEO_REPOST_OPTION SUPPORT. If you have video repost in your current BIOS, this will not effect you. If you did not have video repost in your BIOS, this may change the behavior. We will always check Q_VIDEO_REPOST. Microsoft says that we should NOT repost video, so Q_VIDEO_REPOST should default to disabled.
  • Removed an old workaround that breaks current products OS’s return from S3.
  • Memory count during POST is displayed in the same row where "Checking NVRAM "message is displayed. This is ok for English language but in some other language (e.g. German) the "Checking NVRAM" message is longer than the length of memory count string, which is causing garbage on the screen. Removed the checking NVRAM message from POST screen. This message is not required because the OPSD BIOS always require NVRAM.
  • Added AMI BBS 3.0 and USB boot functionality. For more information, read AMI BBS 3.0 specification.
  • To facilitate development for all chipset North Bridge and South Bridge combinations, we are separating all South Bridge code and tables from the North Bridge component.
  • Fixed the problem where Win98/2K/ME device manager in ACPI mode cannot display COM port IRQ correctly when COM port IRQ is shared in BIOS setup. COM port IRQ sharing in SMSC super I/O needs a special register setting. First, configure UART1(or 2) to use the desired IRQ. Second, configure UART2(or 1) to use No IRQ selected. Third, IRQ share bit in UART1 Option register should be set. The original ASL code for super I/O didn’t consider this SMSC specification. Added necessary checking to meet this specification.
  • Fixed issue where the PS/2 mouse device was disappearing from Device Manager if Windows 2000 was installed with no mouse attached and then the PS/2 mouse was attached after OS installation.
  • Fixed an issue where unpopulated DIMM slot is shown as ’Unknown’ in type 17 of the SMBIOS structure.
  • Re-initialization code for the SIO
  • Corrected LAN string and DIMM strings to match silk screen.
  • Optimized OPROM and PAM registers so ASL can better report to OS (Win2K/WinME) usage regions.
  • The NO_POP signal mutes audio in an attempt to prevent the user from hearing popping noises as the audio device is initialized. It was being asserted when waking up from S3, but never deasserted.
  • To prevent D3h hangs of SMBus locking we clock the SMBus clock so the SPD state machine can finish if it was previously interrupted.
  • Option ROMs which are shadowed should not be declared as a part of the _CRS for the root PCI bus since these addresses are decoded by the memory controller, not the PCI bus.


  • Fixed game port for boards with hard audio, was not an issue for soft audio boards.
  • Enhanced CD-ROM boot emulation detection.
  • Reset SIO keyboard controller early in POST to prevent hangs of D1, 0C, and 15 POST Codes.
  • Fixed error where ABSCMOS would return corrupted strings.
  • Added second COM port to SMB structure, resolves LDCM issue.
  • Under the IDE screen menu for each drive the informative text was always showing up. Now it disappears when a drive is not present, or is selected as type None. This was accomplished by converting the static text items to dynamic text items.


  • Removed AUTO as an option for the "Use ARMD drive as" question.
  • Previously we only supported 1 onboard audio device. If onboard PCI audio was detected, AC’97 audio was automatically disabled. This prevented CNR audio from ever working on a board with PCI audio stuffed. Now we allow the user to enter Setup and select which audio device they want to use.
  • Updated the Italian setup strings for the setup keys. Modified the help text for the "Enter key" to add the "L " symbol. Also modified the abbreviation for the word selez. to use Seleziona.
  • Fix up the drive table late in POST so Windows can find everything.
  • Fixed code to check the Validate command when calling a SMBIOS function 54 subfunction 4006. This fixes the trackers: BEL:05B: DMITEST (Post-Haste I): Mux value is read instead of parameters valid.
  • Fixed bug where the AC'97 PCI SSID and SSVID were incorrectly programmed (their locations were swapped).
  • Added support for the SST FWHx(SST49LF004) 2nd stepping silicon.
  • Changed AML code to consistent Byte wide access for PS2E.
  • Remove code that checks for PnP OS in COMA SIO setup question code.
  • The previous method of determining DCT (Tras and Trc) was based on DIMM SPD Data. However, there is ambiguity in the spec in interpreting the SPD Data such that some PC 133 DIMM that are supposed to run Tras=5, Trc=7 will fail randomly. This fix will always programmed PC 133 Memory to Tras=7, Trc=9.
  • Fixed the issue of bios not being able to recover from CD-ROM with floppy drive attached to the system. The reason was that during recovery, the OS issued an INT10 function 0Fh to check for the monitor mode and used the return value in AL to index into some other procedures. Since our BIOS’s do not support video during recovery, INT10 function 0Fh would return whatever value happened to be in AL register at that time. And it happened that prior to the INT10 call, the OS read our bios data area at location 40:10 and stored it in AL register. Bit 7 and 6 of this location define number of floppy disk drives and bit 0 determines the presence of floppy disk drive A. In the failure case, 40:10 contained a value of 41h that was accidentally returned by INT10 function 0Fh to the caller to be used as an index (an invalid index). The system hung because the code was being indexed to no where. However when floppy drive was removed, 40:10 contained a value of 1. This was a valid index. So there was no problem.
  • Windows makes an INT 13h AH=48h call to the BIOS to get floppy media size. It uses this in the format size option that it offers to the user. Previously, we would return no information about the drive size if no media was present. Now we ask the drive about media size even if no media is present. ATAPI Zips will return 100MB when no media is present, so this becomes a format option once Windows 98 has booted.
  • Fixed the issue of BIOS not meeting SMBIOS 2.3.1 spec
  • Removed DMA 0 from the list of possible DMA's for the ECP LPT port (ACPI).
  • Fixed the issue of 766Mhz Celeron processor speed being displayed as 800Mhz.
  • Improved AGP card speed display in setup.
  • Fixed Memory Size Decrease error report to the correct CMOS Battery Low.
  • Fixed the issue of Bios displaying "Memory size decrease", not displaying "Cmos battery low" and date & time error when Cmos battery is corrupted.
  • In suspend handler added code to set PWRBTN_EN so power button will function as expected.
  • Added Ghz support.
  • Changed the default of Q_On_Lan to the stay_off option. This resolves the issue where the CMOS mfg defaults do not match the DFT spec.
  • Added version 1.1 of EBU to the BIOS Build.
  • Changed the Timer 0 from mode 2 (rate generator) to mode 3 (square wave generator). This is the default AT setting defined by IBM.
  • Disallow AGP 4X if card is operating at 3.3 Volts


  • Fixed an issue where the system would hang if the system has more than 128KB of total option ROM space required.
  • VBIOS 5.1 Production Candidate 2.0 2231
  • Fix ACPI S3 state that caused Win98 and W2K shutdown problems.


  • GC01 candidate for the Express BIOS Update(EBU) application.
  • Added support for the New Flash Product(ICH2NS) for the Non-Secure BIOS update.
  • Define FLASH_PRODUCT as ICH2NS for the customer flash updates on ICH2 products. Added a BIOS work-around for the Ich2 bug, where the read/write transactions were not considering the A16 inversion, when DTE bit is set.
  • Added code to lock SDRAM Read/Write Throttle Control Register (Set Bit 3 Reg. 72h). Since the value has already been locked in stone, this bit can now be set. Once set, we have to spin the BIOS to change the throttling values.
  • Fixed issue with USB2 Legacy not working in DOS.
  • Limit the amount of Memory Loads to 4 for PC133. For 5 or 6 rows, it will always run at PC 100.
  • Fixed an issue where the processor programming for unlock 66 MHz processors are not set correctly. The reason of the failure is when we went to the updated CPU Mux programming method, the CMOS values corresponding to the MUX Values is incorrect.
  • Added code to restore all segment register after returning from option rom execution.
  • Added code to look at the number of floppies and hard drives in the BDA before and after option ROM execution to determine if a drive was added. Previously we looked to see if INT13h/40h had been chained.
  • Added code to address erratic results going from different Sleep States(S1 or S3) after installing the system in S3. The resulting behaviors were: After setting to S1, Win2K will report that the computer is not ACPI-compliant. After setting to S1, the system will not go to S1 but to S3 but the system will not wake up from PS/2 Before accessing any data through SMBUS, ICH2 required the TCO Timer to be halted so as not to interrupt the SMBUS Transaction.
  • Removed code that adjusted the Memory Buffer Strength to be 0.7x for light load PC133.
  • Added work-around to fix the following issue: Understand why the CMOS BB Fault Tolerant Flag is set to NOT Normal after a Flash Update. This will make the BIOS code restore A16 inversion when it thinks that the Flash Update did not complete. Understand why RTC, WOL , PME will cause the SMI handler in the BB to hang.
  • Added fix will temporarily allow the BIOS to clear RTC, PME, WOL Status early in POST to prevent a hang at 0xE8 while the system is trying to reset.
  • Adjusted the Memory Buffer Strength and PMOS/NMOS for PC 133. SMD is always set to 1.0X, SMA is 0.7X for loads<=4, 1.0 for loads between 8 and 32, and 1.7 for loads >=36. SCS is always set to 1.0X. PMOS/NMOS is the same for row<=2, but PMOS is incremented by 2 for row>=3 and NMOS is incremented by 1 for row>=3.
  • Add new Video BIOS Production 05 Build 2197.
  • The code uses port 64 bit 2 to determine if the system is coming out of Soft or Hard Reset.

This site uses cookies. Cookies are simple text files stored on the user's computer. They are used for adding features and security to this site. Read the privacy policy.