strange

[anikhtos]

the other day i played with my old computer with
athlon xp 2600+ at nforce2 mainboard
i installed xp to it (i have there 98 to play old games that dislike xp)
so what i noticed is the super fast boot time.
all the bench sites show that atom 330 has 2 times the power over that cpu but again in real life i do not see it. did not have time too look into it but the boot time at athlon xp was far far faster. well it is faster than the athlon II 250e machine i have.
so that means
1)windows xp not utilize some features of modern cpu
2)modern cpu are not that strong at older feautures??

 

[skipsneeky2]

Used to have a pentium 4 that would boot quicker then my core 2 duo e6750 would.
To weird

 

[Blitz KriegeR]

If its a brand new OS install with minimal drives/apps to load of course it is going to boot faster regardless of CPU.

Also, the biggest factor in boot time is actually the HDD/SSD. Only if the CPU is really shit does it become the limiting factor.

 

[anikhtos]

If its a brand new OS install with minimal drives/apps to load of course it is going to boot faster regardless of CPU.

Also, the biggest factor in boot time is actually the HDD/SSD. Only if the CPU is really shit does it become the limiting factor.

clean install on both of them
same hard disk on both machines (mooved the ssd between systems)
hdd is not the issue at booting
even after some drivers install a few programs the boot speed of 2600+ remain fast.
because of testing at my current system the athlon II 250e sucks even at clean install at boot time.
so why such an old cpu boot so fast faster than a newly one?

 

[Blitz KriegeR]

What I said is still relevant, but the only thing I can add is that Athlon XP was a mainstream desktop platform, while Atom is a low-power ultra-portable. Don't forget the differences between such platforms.

EDIT: Sorry misread you are using Athlon II 250e, not Atom.

 

[Patrick Wolf]

Edit: Nvm, the original post is too confusing to even comment.

 

[Blitz KriegeR]

Nvm, the original post is too confusing to even comment.

+1

 

[anikhtos]

Edit: Nvm, the original post is too confusing to even comment.

what is so strange at the original?
according to benchmark tools the atom cpu is 2 times faster than the athlon xp 2600+ cpu.
but yet the athlon xp beats atom at boot time.
okey as someone will say atom is a cripled cpu to draw little watts.
so i checked my other system which is athlon II 250e and still the athlonII 250e has the fatest boot.
the hard disk was the same in all 3 configurations.
ram 2,5gb for athlon XP, 4gb ram for the rest 2
windows XP was installed in all 3 configuration
either clean install or the installation with drivers and programs and still
athlon XP was far far faster.
even with drivers and programs athlon xp made almost the same boot time
while athlon II 250e was a notisable 2-3 sec slower than clean install.

so i wonder is it streange or not for such an old cpu to outsmoke thosemoderns cpu in boot time??? since in some times the boot time is used as a bencmark tool for the cpu.

obiously the architecture is far diferent to this cpu.
with the newer has more technologies supporting.
so in order for the cpu to support more technologies the manufaucture has halved older technologies??? thus making the newer processor slower ???

 

[Zap]

since in some times the boot time is used as a bencmark tool for the cpu.

I would never use boot time as a CPU benchmark. Heck, I've had systems take a half minute to finish POST.

I think it may have to do with how many things are integrated into the platform and how many/big drivers are loaded.

 

[sm625]

I have a windows 95 system with a cyrix 2xx something mhz that i guarantee will boot faster than any modern system out there. From the time you turn on the power switch, its like 15 seconds to the desktop. It has nothing at all to do with cpu speed. It is chipset. It is drivers. More specifically, the amount of hardcoded delays when the drivers initialize. Computers could boot a LOT faster than they do, its just that hardware vendors are lazy and insert all sorts of 50 mSec sleeps in their code, and things have to be done in order. The delays add up. This has always aggravated me because its just plain retarded, and microsoft should have done something about it 10 years ago.

 

[darkswordsman17]

I would never use boot time as a CPU benchmark. Heck, I've had systems take a half minute to finish POST.

I think it may have to do with how many things are integrated into the platform and how many/big drivers are loaded.

This.

As for extrapolating that about OS and CPU features, well OSes aren't really the driving force behind CPU development so I wouldn't really say too much about things like that. Keep in mind that Atom isn't really more modern than an Athlon XP either.

All the same peripherals as well? What about video card, sound card, network chip, etc? One small thing can be the difference you're seeing, which is why boot time is such an unreliable benchmark tool. It makes sense for testing HDDs and SSDs, where you're keeping the system consistent and just changing that, but comparing different systems (which is what you're doing, as each system uses different CPU/mobo chipset, etc) doesn't work as well. There's too many variables to take into account, as just changing motherboards (even ones that use the same chipset) could make a noticeable difference.

 

[GammaLaser]

I have a windows 95 system with a cyrix 2xx something mhz that i guarantee will boot faster than any modern system out there. From the time you turn on the power switch, its like 15 seconds to the desktop. It has nothing at all to do with cpu speed. It is chipset. It is drivers. More specifically, the amount of hardcoded delays when the drivers initialize. Computers could boot a LOT faster than they do, its just that hardware vendors are lazy and insert all sorts of 50 mSec sleeps in their code, and things have to be done in order. The delays add up. This has always aggravated me because its just plain retarded, and microsoft should have done something about it 10 years ago.

+1. Besides the CPUs, there are other differences such as the mobo/chipset/bios/drivers which cause this. CPUs themselves have little to do with the boot time.

 

[anikhtos]

This.

As for extrapolating that about OS and CPU features, well OSes aren't really the driving force behind CPU development so I wouldn't really say too much about things like that. Keep in mind that Atom isn't really more modern than an Athlon XP either.

All the same peripherals as well? What about video card, sound card, network chip, etc? One small thing can be the difference you're seeing, which is why boot time is such an unreliable benchmark tool. It makes sense for testing HDDs and SSDs, where you're keeping the system consistent and just changing that, but comparing different systems (which is what you're doing, as each system uses different CPU/mobo chipset, etc) doesn't work as well. There's too many variables to take into account, as just changing motherboards (even ones that use the same chipset) could make a noticeable difference.

so even if we count time after the os selection screen
there would be boot time diference because of the drivers and the chipsets involved. okey thanks for clarify (i knew the boot is not that reliable but since the cpu power between athlon xp and athlon II 250e is enormous i found strange to see such a speed favor for athlon xp)

 

[darkswordsman17]

so even if we count time after the os selection screen
there would be boot time diference because of the drivers and the chipsets involved. okey thanks for clarify (i knew the boot is not that reliable but since the cpu power between athlon xp and athlon II 250e is enormous i found strange to see such a speed favor for athlon xp)

Yeah, pretty much, although that would remove one factor of the bootup so it would be a bit better for comparison.

I actually wouldn't mind at least some boot time feedback, as while its not a big issue, it can matter. Especially for say laptops or OEM systems.

 

[Deanodarlo]

I'm going from memory here, so could be completely wrong but...

...apart from multitasking and multi-threaded apps, and those that use later SSE extensions, is the Atom 330 really much faster than a Athlon XP 2600+?

Plus isn't the Atom an in-order CPU, when the Athlon XP is a traditional out-of-order? The Athlon XP may indeed be much faster at certain tasks and handling single threads.

The nforce2 and nforce3 chipset do seem to have lightening boot times by the way. My XP and Athlon 64 retro PC's boot faster than my modern Intel rigs - probably because of really tightly developed chipset drivers for XP provided by nvidia.

The nforce2 was an awesome chipset - still one of my favourites with the soundstorm.

 

[Rvenger]

I'm going from memory here, so could be completely wrong but...

...apart from multitasking and multi-threaded apps, and those that use later SSE extensions, is the Atom 330 really much faster than a Athlon XP 2600+?

Plus isn't the Atom an in-order CPU, when the Athlon XP is a traditional out-of-order? The Athlon XP may indeed be much faster at certain tasks and handling single threads.

The nforce2 and nforce3 chipset do seem to have lightening boot times by the way. My XP and Athlon 64 retro PC's boot faster than my modern Intel rigs - probably because of really tightly developed chipset drivers for XP provided by nvidia.

The nforce2 was an awesome chipset - still one of my favourites with the soundstorm.

I miss my A7N8X Deluxe and Barton 2500+. The sound was amazing on that board. Bios glitch - not so much.

 

[anikhtos]

Yeah, pretty much, although that would remove one factor of the bootup so it would be a bit better for comparison.

I actually wouldn't mind at least some boot time feedback, as while its not a big issue, it can matter. Especially for say laptops or OEM systems.

the boot up i checked it visually at windows logon screen
at xp it moved from that screen to desktop
with 250e made 2passes
with atom 5-6 passes.