XP boot times

[ochadd]

After several hours of tweaking with no success I'm begging for ideas.

Old system boot time: 40 seconds
X2 3800 @ 2.3
2GB DDR500
DFI NF4 Ultra
WD800JB 7200rpm 8MB cache IDE

New system boot time: 58 seconds
e6750 @ 3ghz
2GB DDR2 900
Asus P5ke
WD2500AAKS 7200rpm 16MB SATA

Both have XP SP2 and is fully updated.
New system GUI boot is disabled in boot.ini and set the wait times to 0. Old system is a standard setup.
Both have the same services enabled.
Both load AVG and Zonealarm on startup.
Both have the most recent BIOS.
No sounds played at startup on new system. Old system standard XP sound theme

I've been running the system for a few weeks with no problems aside from slow boot times. I had not tweaked XP until last night and expected massive improvements as my last system did. It improved a total of 7 seconds.

The three issues I can think of are a screwed XP install, AMDs better memory performance, or somehow the SATA drive is doing something on the backend the IDE drive didn't do. Due to no other problems I'm thinking XP is fine.

Any ideas?

 

[loup garou]

Originally posted by: ochadd
After several hours of tweaking with no success I'm begging for ideas.

Start using sleep/hibernate and stop worrying over boot times.

 

[ochadd]

Originally posted by: loup garou

Originally posted by: ochadd
After several hours of tweaking with no success I'm begging for ideas.

Start using sleep/hibernate and stop worrying over boot times.

That is not an acceptable solution. It will do a good job at hiding the problem in the morning/night but the underlying issue is still there.

 

[nerp]

Maybe it's not really an issue? You might be able to use bootvis or something to see if there's a snag or delay for some reason, but you might be staring at an architectural reality that can't be changed by any sort of tweaking. Maybe newer drivers over the next year will fix it if it happens to be a NIC or device of some kind that hasn't had its drivers fully optimized.

 

[MrChad]

Originally posted by: ochadd

Originally posted by: loup garou

Originally posted by: ochadd
After several hours of tweaking with no success I'm begging for ideas.

Start using sleep/hibernate and stop worrying over boot times.

That is not an acceptable solution. It will do a good job at hiding the problem in the morning/night but the underlying issue is still there.

What's the problem? A difference of 18 seconds? Do both systems perform acceptably otherwise?

 

[ochadd]

Originally posted by: nerp
Maybe it's not really an issue? You might be able to use bootvis or something to see if there's a snag or delay for some reason, but you might be staring at an architectural reality that can't be changed by any sort of tweaking. Maybe newer drivers over the next year will fix it if it happens to be a NIC or device of some kind that hasn't had its drivers fully optimized.

That's what I'm thinking as well. I have no experience with decent Intel chipsets so immediately the trouble seems to come from the platform. The P35 boards have been out for some time so I'd imagine most bugs would be worked out.

 

[ochadd]

Originally posted by: MrChad

Originally posted by: ochadd

Originally posted by: loup garou

Originally posted by: ochadd
After several hours of tweaking with no success I'm begging for ideas.

Start using sleep/hibernate and stop worrying over boot times.

That is not an acceptable solution. It will do a good job at hiding the problem in the morning/night but the underlying issue is still there.

What's the problem? A difference of 18 seconds? Do both systems perform acceptably otherwise?

The Intel system doesn't feel as snappy but if it's actually slower or just perception I don't know. The problem is also the 18 seconds. If it were 5 seconds I could deal with it. If it were 20 seconds and I knew where those 20 seconds were going it would be fine as well. To have my 3 year old machine be perceptively faster.. Should have bought a raptor.

 

[corkyg]

You did a 'no-no' by posting the same thread in two forums. So, I'll give you the same answer.

Dupe

 

[ochadd]

Originally posted by: corkyg
You did a 'no-no' by posting the same thread in two forums. So, I'll give you the same answer.

Dupe

I figure hardware people and software people have different outlooks on this particular problem.

 

[nerp]

Hmm. Now that I think about it, if your new machine "feels" slower in general than your old AMD box, I'd be concerned. It's true that basic windows functions can only go so fast -- it's hard to tell a difference between my celeron M 440 laptop and opteron box when it comes to surfing web pages with IRC open. But. . . what about game performance and synthetic benchmarks? The Intel system should dominate. Does it?

 

[ochadd]

Yes it tears the old system a new one in WoW, DiRT, Unreal III, and Crysis. The upgrade wasn't as mind blowing though as I upgraded the video card upgraded in my old system from a 7800gt to a 3870 before trying a new platform. Unreal III, Dirt, and Crysis all gained more than a couple FPS with the platform change.

 

[SoundTheSurrender]

It just could be the way the hardware handles booting and is faster afterwards.

 

[lxskllr]

Originally posted by: SoundTheSurrender
It just could be the way the hardware handles booting and is faster afterwards.

That's my guess. There may be a difference between DDR and DDR2 systems. My system in sig boots up slower than my old P4 system did, but once it's in Windows there's no contest. My new rig annihilates all of my benches from my old rig.

Old rig:
P4 3.06 @ 3.50ghz
1gb DDR
160gb hd 8mb cache
GF 6800

 

[Canterwood]

I've found a difference between my old AMD X2 and my C2D in Windows. The AMD is much smoother.

There's a definate lag with the C2D opening folders with lots of different files, that doesn't exist with the AMD.

The C2D wins in apps and games, but the lag in Windows is kinda annoying.