Windows via Boot Camp is Slow ;-)

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

TaliesinSoft - 12 Jun 2006 20:51 GMT

I've installed Boot Camp and Windows XP Pro so that I can run a specific
application development environment that is not available in a Macintosh
native form.

Today I ran a set of benchmarks furnished by the environments developer and,
lo-and-behold, they ran almost twice as fast on my 2.16 GHz MacBook Pro as
they do on a 3 GHz + Pentium.

Signature

James Leo Ryan ..... Austin, Texas ..... taliesinsoft@mac.com

Reply to this Message

Richard E Maine - 12 Jun 2006 21:31 GMT

> I've installed Boot Camp and Windows XP Pro so that I can run a specific
> application development environment that is not available in a Macintosh
[quoted text clipped - 3 lines]
> lo-and-behold, they ran almost twice as fast on my 2.16 GHz MacBook Pro as
> they do on a 3 GHz + Pentium.

That might be the dual core you are seeing, if the app can make use of
both cores in parallel.

Or, if the app happens to be graphics intensive, it could be the
graphics chip/card. The x1600 graphics chip on the MacBook Pro is no
slouch at all. Sure, it isn't the very top end of screaming (pretty much
literally, with their fans), hot (definitely literally), $500+ cards
that the hard-core gamers go for, but it isn't that far below them.
Turns out is is better than any of the video cards that I happen to have
in my Windows boxes.

Signature

Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain| experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain

Reply to this Message

NRen2k5 - 13 Jun 2006 06:21 GMT

>> I've installed Boot Camp and Windows XP Pro so that I can run a specific
>> application development environment that is not available in a Macintosh
[quoted text clipped - 14 lines]
> Turns out is is better than any of the video cards that I happen to have
> in my Windows boxes.

True, that. The x1600 is top-of-the-line.

- NRen2k5

Reply to this Message

NRen2k5 - 13 Jun 2006 06:17 GMT

I think that your Pentium system in need of a cleanup.

- NRen2k5

Reply to this Message

Mike - 13 Jun 2006 16:50 GMT

> I've installed Boot Camp and Windows XP Pro so that I can run a specific
> application development environment that is not available in a Macintosh
[quoted text clipped - 4 lines]
> lo-and-behold, they ran almost twice as fast on my 2.16 GHz MacBook Pro as
> they do on a 3 GHz + Pentium.

Why is that surprising? You were expecting a single CPU machine to be
faster than a dual CPU machine????

Mike

Reply to this Message

James Glidewell - 14 Jun 2006 17:03 GMT

>> I've installed Boot Camp and Windows XP Pro so that I can run a specific
>> application development environment that is not available in a Macintosh
[quoted text clipped - 8 lines]
> Why is that surprising? You were expecting a single CPU machine to be
> faster than a dual CPU machine????

I figure that he was surprised that a _Mac_ *laptop* beat a reasonably
high-end _PC_ *desktop* for running any Windows app.

(Note that it wasn't explicitly stated that the Pentium system was a
desktop, but that would indeed make the results all the more "surprising")

A more discerning fellow such as yourself is no doubt unsurprised by
the superiority of the Apple product's design and performance.

BTW, "dual core" != "dual CPU" in terms of performance. In most cases,
close, but the gap widens as contention for the paths off the chip
increases.

Reply to this Message

TaliesinSoft - 14 Jun 2006 18:48 GMT

> Mike wrote:
>>
[quoted text clipped - 21 lines]
> close, but the gap widens as contention for the paths off the chip
> increases.

Following up on my original posting........

The Pentium system was indeed a desktop.

I have now also installed Windows XP Pro under Parallels. The benchmarks,
much to my surprise, ran about 10% faster under Parallels than under Boot
Camp.

Signature

James Leo Ryan ..... Austin, Texas ..... taliesinsoft@mac.com

Reply to this Message

Richard E Maine - 14 Jun 2006 19:41 GMT

> I have now also installed Windows XP Pro under Parallels. The benchmarks,
> much to my surprise, ran about 10% faster under Parallels than under Boot
> Camp.

Ok. Now that one I have a hard time understanding.

The "obvious" negative of parallels compared to dual booting is that
parallels "ought" to be slower. That's a tradeoff I've been though
before on my Linux/Windows boxes. Running VMWare is hugely more
convenient than dual booting, but it is also slower; the speed cost
varies from modest for most cpu-intensive things, to essentially
infinite for some graphics-intensive things (the infinite is because
some apps literally do not run at all, saying that the emulated graphics
card is not adequate).

The "fancy" features of the Intel dual-core chips ought to be able to
cutdown on the performance penalty, but I have touble imagining how it
could run faster.... I suppose perhaps if there is a problem in the boot
camp video drivers. That's software that you are not using in parallels,
so if something is badly interacting with it, that problem wouldn't show
up in parallels.

I bought a copy of parallels, but haven't yet installed it anywhere. (I
just wanted to lock in the pre-release price).

Signature

Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain| experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain

Reply to this Message

G.T. - 14 Jun 2006 22:12 GMT

> >>> I've installed Boot Camp and Windows XP Pro so that I can run a specific
> >>> application development environment that is not available in a Macintosh
[quoted text clipped - 27 lines]
> much to my surprise, ran about 10% faster under Parallels than under Boot
> Camp.

That's very odd. When you say "under Boot Camp" it's hard to tell if you
realize that Boot Camp is just a boot method, repartitioner, and Windows
drivers. With Boot Camp XP isn't running under anything, it's running direc
tly on the hardware.

One possible scenario with XP running slower on the hardware than under
Parallels is that one piece of hardware that is tested heavily in the
benchmark may have a poor native driver that doesn't work as well as the
virtualized driver in Parallels.

Greg

Signature

"Up in the clouds
ah up in the clouds
down in the dirt
it's so good
so good it hurts" - The Mekons

Reply to this Message

TaliesinSoft - 14 Jun 2006 23:01 GMT

[commenting on my stating that a benchmark ran faster using Parallels Desktop
than it did using Boot Camp]

> That's very odd. When you say "under Boot Camp" it's hard to tell if you
> realize that Boot Camp is just a boot method, repartitioner, and Windows
> drivers. With Boot Camp XP isn't running under anything, it's running
> direc tly on the hardware.

I'll accept that "under Boot Camp" was perhaps a confusing phrase to have
used. The comparison I made was running a benchmark that is interpretively
executed using Dyalog APL which, in this case, was running as a Windows XP
Pro application.

> One possible scenario with XP running slower on the hardware than under
> Parallels is that one piece of hardware that is tested heavily in the
> benchmark may have a poor native driver that doesn't work as well as the
> virtualized driver in Parallels.

Just what hardware and/or native drivers were involved in the benchmark is
something I don't know. I was pleased that both the Parallels Desktop and the
Boot Camp tests produced favorable results. I've opted to go with Parallels
Desktop as I do like to be able to concurrently use both the Windows and Mac
environments.

Signature

James Leo Ryan ..... Austin, Texas ..... taliesinsoft@mac.com

Reply to this Message

G.T. - 14 Jun 2006 23:22 GMT

> [commenting on my stating that a benchmark ran faster using Parallels Desktop
> than it did using Boot Camp]
[quoted text clipped - 19 lines]
> Desktop as I do like to be able to concurrently use both the Windows and Mac
> environments.

Yep, and it rocks that the results are favorable enough for you to use it.

Greg

Reply to this Message

James Glidewell - 15 Jun 2006 00:20 GMT

> Following up on my original posting........
>
[quoted text clipped - 3 lines]
> much to my surprise, ran about 10% faster under Parallels than under Boot
> Camp.

Thanks, James, for posting your results here.

I've pre-purchased Parallels, based on my very positive experiences with
the Beta, but have been reluctant to try Boot Camp, since my experience
with dual boot systems has consistently followed the Murphy's corollary
that "Whatever OS you are currently running is the _wrong_ one for doing
whatever it is you want to do right now". (And I hate sacrificing the
disk space for something I'll use infrequently...)

WinXP under Parallels feels plenty fast, but then so does OSX on my (base
model) MacBook w/1GB memory. Given your performance experiences, it seems
like sticking with Parallels is the right way to go for non-gaming XP apps.

Reply to this Message

Mac Forum / General / Hardware / June 2006

Windows via Boot Camp is Slow ;-)