8 replies to this topic

[MedEvil]

The last couple of days, i have been fighting with PE to turn UDMA on for the BootDrive (CD).

It wasn't till i started to question the testing tools, that i started to see some light.

The tests done on machines with a proper XP, lead those tools to declare all sorts of things.
The only thing they could actually agree upon was, to disagree. Every Time!

So today i settled for indirect testing by doing speed tests. And that's where things got interesting!

Each CD / DVD-Drive always starts reading the innermost circle of the CD with the same speed no matter PIO or UDMA.
It isn't till about 150 - 170MB that a difference occurs!
In PIO mode the speed becomes static, with UDMA the speed keeps increasing.

The first 150MB there is no difference between PIO and UDMA and that's where the system files reside!


One more?


I have a Laptop which boots NaughtyPE to Desktop in less than 3'30", while the big machines take an average of 7'30".

So i always thought, how can i make the drives in the desktop machines as fast as the laptop one?

Turns out the Laptop drive is not only actually slower, but it also runs slower!

While the desktop drives start out with 10/14x, the Laptop starts with only 8x.
While the desktop top (PIO) speed is almost 16x/19x, the Laptop doesn't even reach 13x.

The only advantage the Laptop drive seems to have is, that it takes up considerably less CPU Power on the 1000MHz PIII, than the desktop drives take from the 2+Ghz machines.



btw. The fact that all HDDs get max.UDMA, while all CD/DVD-ROM drives get only PIO despite using the same driver, suggests that there is no technical problem that prevents the Boot-CD from getting UDMA, just some code.




Conclusion:
-----------
Though i would love to have at least once UDMA for the CD-Drive, to proove that it is useless.
I have to say that this is a pointless quest. Nothing is to be gained from success.

[Wonko the Sane]

I have some difficulties in following your report.

"Speed tests" performed HOW?

If the "laptop CD" is not only PIO instead of UDMA, but also "slower" how comes it loads faster?

Isn't the CD drive initially driven through it's BIOS driver (thus a different BIOS code or a different firmware inside the drive may make a difference)?

Have you checked if here:
http://winhlp.com/node/10
there isn't anything useful for your chase?

Can you try making a CD with a (dummy) 150 Mb file at the beginning and with actually booted file past this initial area?
Does it make a difference?

Wonko

[MedEvil]

"Speed tests" performed HOW?

With a Nero Tool called CDSpeed. It does with CD/DVD drives what HDTune does with HDD. Checking the raw read speed.

If the "laptop CD" is not only PIO instead of UDMA, but also "slower" how comes it loads faster?

Beats me! It has the lower raw read speed, but the booting goes faster.

Isn't the CD drive initially driven through it's BIOS driver (thus a different BIOS code or a different firmware inside the drive may make a difference)?

Are you suggesting that the BIOS runs the drive with UDMA, but XP switches back to PIO mode?
But even if it would work like that, at the beginning of the CD the raw read speed is below max PIO and it wouldn't matter.

Can you try making a CD with a (dummy) 150 Mb file at the beginning and with actually booted file past this initial area?
Does it make a difference?

What should that be good for?
Once the drive reaches max PIO the raw read speed becomes static.

[MedEvil]

Have you checked if here:
http://winhlp.com/node/10
there isn't anything useful for your chase?

I have always suspected, that you post links witout even reading them. Now i have proof!

If the program found any ATA channel to reset, reboot your computer and test all drives.

[Wonko the Sane]

Never mind.

I'll go and find some more links without reading them, only, this time, I'll also avoid pointing them out for your interest.

I have always suspected that any single minute spent in trying to understand and help you is perfectly and utterly wasted, now I have definite proof.

Wonko

[MedEvil]

Ever thought about posting something that is actually helpful or at least on topic?

[Wonko the Sane]

Ever thought about posting something that is actually helpful or at least on topic?

Sure. Too bad you failed to understand it EVERY single TIME I did.

D@mn, another minute of my life completely wasted with you.

Wonko

[MedEvil]

That's impossible! Every minute spend with me is an unforgettable experience!

[MedEvil]

I'm now able to explain, why exactly a PE starts faster from a DVD than from a CD, despite PIO mode.

Since all or at least most drives today use CAV, they have a fixed starting transfer rate at the beginning of the media (in the center, if you don't know).

For my drives the min CD speed is 8x 10x and 14x. Where 1x = 150KB/s
They do not reach top PIO speed up until 150MB into the CD at which point the PE is already loaded.

My drives start however DVD with a min speed of 2,10x. Where 1x = 1353 KB/s
If you do your math right, you will notice that DVDs are read right from the start with top PIO speed.

The advantage of a DVD is in the beginning a 2 times greater readspeed than a CD. The advantage than becomes less and less untill both media reach the same transfer speed for the rest of the media.

UDMA for a BootCD will probably not give you any advantage, without drastic changes to the layout, but UDMA for a BootDVD would be pretty impressive!
Raising the starting speed from 2,13x to 3,34x alone. Which is a speed increase of almost 57% over PIO.






PS: I told you that CD and DVD are read from the hole to the rim.
Did you knew that on double layer DVD the second layer is read the other way around, from the rim to the hole?