Usually, you would use the Flitedeck utility from VIA. However - this one is deprecated, and won't work on the new plattforms.
You can use a USB-Floppy drive now. Seems the BIOS has been improved - compared to the one on the M10000. Make sure the Write-Protection is removed: remove the jumper WP accordingly for the time to flash the BIOS.

Create a bootable dos-Disk - using Caldera-DOS 6.x here - put all required files on it. You can find some in www.bootdisk.com. Make sure you have a new version of the flash-tool (using version 829 here).
Boot from the floppy - and once done - type in the following command:

awfl829.exe BIOS.bin /py/sn/cd/cp

and you should be done. Note that this should work with future and old versions of BIOS's. Do not use the option /tiny on that one. Won't work.  

 In the module-configuration file for lm_sensors - e.g. under Mandrake 10.1 /etc/sysconfig/lm_sensors
Gentoo /etc/conf.d/lm_sensors
make sure the following modules are configured to be loaded. You'll require lm_sensors-2.9.1 at least for the display of sensor informations to be correct.

MODULE_0=i2c-viapro
MODULE_1=i2c-isa
MODULE_2=eeprom
MODULE_3=vt1211

The /etc/sensors.conf file has to be as in the attached sensors.conf file.  

  Check out the manufacturer homepage for technical details: Via Homepage

Some more guides can be found in the download section
* Epia SP8000E Operating Guidelines
* Epia SP8000E User Guide 

 The Epia SP8000E uses a 800MHz Nemiah core CPU. By loading the longhaul module - you can throttle the CPU (Tested with 2.6.11.6). Please note that there are many patches to the cpufrequency control - and many are unstable. If your System locks-Hard - you have a bad implementation :)
So - all you need to do is enable the cpu frequency scaling by loading the longhaul module and some governors:

 modprobe longhaul
 modprobe cpufreq_powersave

Check using the dmesg command to see if it loaded correctly.
If this works - make sure the longhaul module gets loaded at every boot. For this - just add longhaul and cpufreq_powersave into the /etc/modprobe.preload file.
Using longhaul - make sure your kernel was configured with the userspace-governor enabled by default - then change to:

root # cd /sys/devices/system/cpu/cpu0/cpufreq/
cpufreq # ls 
affected_ cpuinfo_cur_freq
cpuinfo_max_freq
cpuinfo_min_freq
scaling_available_frequencies
scaling_available_governors
scaling_cur_freq
scaling_driver
scaling_governor
scaling_max_freq
scaling_min_freq
scaling_setspeed
cpufreq # cat scaling_cur_freq
798000
cpufreq # cat scaling_available_frequencies
532000 598000 731000 798000 665000
cpufreq # echo 532000 > scaling_setspeed
cpufreq # cat scaling_cur_freq
532000

With the userspace governor activated

cpufreq # cat scaling_available_governors
userspace

The System will set the speed up to 800MHz in this special case - if required - and when idle again - go down with the frequency. In short - it will dynamically monitor the system load and regulate the CPU Frequency accordingly. If you want to stick the system into powersave mode - load the cpufreq_powersave.ko module

cpufreq # modprobe cpufreq_powersave
cpufreq # cat scaling_available_governors
powersave userspace
echo powersave > scaling_governor

This will make the CPU-Frequency sticky at the lowest speed your CPU can run at - and hopefully lower the overall system power consumption and heat dissipation This is what I have added to the system's /etc/rc.local file on my server - so it gets activated at every boot.

sleep 60
echo powersave > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor

For those having a RedHat based system - you can use the following startup-script cpufreq.tar.gz. 2 Files are in there:

/etc/init.d/cpufreq
/etc/sysconfig/cpufreq

Edit the /etc/sysconfig/cpufreq and set your preferences - then:

init.d # ./cpufreq status
CPU Frequency Scaling deactivated !
stargate init.d # ./cpufreq start
Loading modules: longhaul, cpufreq_powersave
Setting gowernor to: powersave
Actual powerstatus: longhaul(powersave) @ 399000KHz

If this works - register it to the underlaying system issuing:

chkconfig --add cpufreq
chkconfig --list cpufreq
cpufreq         0:off   1:off   2:on    3:on    4:on    5:on    6:off

This will start the script at every startup.
Note that you can start the exact order by adapting the chkconfig-line in the init.d/cpufreq script itself. It is set to be started after udev and iptables startup.

# chkconfig: 2345 05 90

If chkconfig does not exist on your system - you can of course create symlinks - as back in old times. 

 Check the following files:

• Epios Distribution
  • For Epios System - Via own Drivers xorg.conf

 Here it is:  

 The main reason is simple. If you use an old PC laying around - you would require up to 100W Power in average - makes 2.4KWh per day power to use.
An Epia SP8000E System, with 2x2.5" HD Drives uses ~28W Power in full utilization mode - makes approx 672Wh in a day - ~3 Times less money to pay ... Note than in IDLE mode - I have a constant 18W = 432Wh power consumption.

However - other features come in I based my choice on: USB 2.0, Firewire, Padlock-capability (e.g. AES encryption engine), Passive cooling, S-ATA connectors and a well supported cpufrequency-scaling. Check out the Epia Power Simulator for computing the powerconsumption of a system you might want to set up.  

 The VIA CN400 Chipset is designed to enable high quality digital video-streaming and DVD Playback in fanless operations mode.

Features are a Embedded UniChrome PRO 2D/3D Mpege-2 Decoder/MPEG-4 Accelerator, DDR400 support, motion compensation and duo-view support. Also - USB 2.0, Serial-ATA, 10/100 MBIt LAN and ATA133/100 IDE.

The Overall Layout looks like this:
 

 The very standard kernel from kernel.org, adding the vt1211-patch - as stated in the related links - and use the attached config.

For a Mandrake 10.1/10.2 System

• config_2.6.13.EpiaN
  config_2.6.13.2.EpiaN scsi and raid compiled in to get 2 S-ATA Drives running. Note that the longhaul module causes the system to crash, and using 2 SATA disks in mirror mode - makes it somewhat unstable. The VT8237 Chip seems to not like high load on the bus.
• config_2.6.11.5-epiaN scsi and raid compiled into the kernel - as 2 S-ATA Drives running on it. No initrd required to boot. This kernel is much more stable when using the cfq scheduler then any other kernel. Under heavy traffic & High load though - crashes can still occure.

 I would have loved someone to answer me that question before I bought that case. The Short answer is: No
The case is nice - though the CPU temperature is going up to 70/80C. That might be OK for the CPU and motherboard - however - we have to take into account the other components in the system. The Harddisk or CDRom Drives will suffer from it. Especially harddisks start to fail when the temperature is over 60C - so at least the harddrive will require a FAN to operate.

Solution: Instead of the location for the 2.5" harddrive - put a 50mm FAN on the side which blows the air out of the case - you'll have to make a big hole on the side though. This keeps the System temperature below 45C which is sustainable for all the hardware inside.  

 The simple answer is: Yes

Using the Serener GD-L01 Case - this is definitly not an issue. It is passively cooled using Heatpipe technolgy, and keeps the system warm only. Not hot.