Why is there an x86 emu if a processor is x86-64?

Discussion in 'Windows 64bit' started by =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=, Jul 23, 2006.

  1. I don't quite understand the approach Microsoft has taken with x64. I have an
    AMD Athlon 64 3500+, and it supports both 64bit instructions and the 32bit.
    So why is x64 emulating the 32bit instructions while my CPU can interpret
    them natively? Wouldn't that give me better program/driver support? If this
    is something that would require some long explanation (like if its something
    that involves the kernel's inner workings) then don't worry about it much and
    just say "its the kernel" or somethin, I'll get ya.

    Thanks a lot!
    Elliot
    =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=, Jul 23, 2006
    #1
    1. Advertising

  2. The kernel. ;)

    Seriously, it _does_ take advantage of the native support. But there still
    needs to have a very thin layer that translates between the 64 bit OS and
    the 32-bit application. This layer really doesn't impose any overhead, and
    many 32-bit programs actually appear faster in 64bit, though it really
    depends a lot on the program. The 64bit I/O subsystem is definitely more
    efficient, however, so programs that use a lot of I/O can sometimes be
    faster in 64 bit windows even though they are 32bit programs and they run in
    the WOW64 subsystem. There's a good graphic of this in the whitepaper I
    wrote for Microsoft:
    http://download.microsoft.com/downl..._XP_Professional_x64_Edition_Right_for_Me.doc

    You might find it of interest.

    --
    Charlie.
    http://msmvps.com/xperts64


    Elliot Hudgins wrote:
    > I don't quite understand the approach Microsoft has taken with x64. I
    > have an AMD Athlon 64 3500+, and it supports both 64bit instructions and
    > the 32bit. So why is x64 emulating the 32bit instructions while my CPU
    > can interpret them natively? Wouldn't that give me better program/driver
    > support? If this is something that would require some long explanation
    > (like if its something that involves the kernel's inner workings) then
    > don't worry about it much and just say "its the kernel" or somethin, I'll
    > get ya.
    >
    > Thanks a lot!
    > Elliot
    Charlie Russel - MVP, Jul 23, 2006
    #2
    1. Advertising

  3. Wow! Your paper is really, really thought out and explained pretty much
    everything I needed to know! Thanks a lot for the info. I guess that I just
    wished for 32 bit drivers to work in 64, I miss my tuner :p

    Let's all hope that driver support gets better once Vista comes around :D
    Oh, and do you know if it will have some type of megalo-awesome emulation
    layer that WILL run 32-bit drivers? If you say it will, I'll love you.
    Forever. Not in the man to man way, but I'll give you a hug in public and not
    think twice. :p

    Elliot

    "Charlie Russel - MVP" wrote:

    > The kernel. ;)
    >
    > Seriously, it _does_ take advantage of the native support. But there still
    > needs to have a very thin layer that translates between the 64 bit OS and
    > the 32-bit application. This layer really doesn't impose any overhead, and
    > many 32-bit programs actually appear faster in 64bit, though it really
    > depends a lot on the program. The 64bit I/O subsystem is definitely more
    > efficient, however, so programs that use a lot of I/O can sometimes be
    > faster in 64 bit windows even though they are 32bit programs and they run in
    > the WOW64 subsystem. There's a good graphic of this in the whitepaper I
    > wrote for Microsoft:
    > http://download.microsoft.com/downl..._XP_Professional_x64_Edition_Right_for_Me.doc
    >
    > You might find it of interest.
    >
    > --
    > Charlie.
    > http://msmvps.com/xperts64
    >
    >
    > Elliot Hudgins wrote:
    > > I don't quite understand the approach Microsoft has taken with x64. I
    > > have an AMD Athlon 64 3500+, and it supports both 64bit instructions and
    > > the 32bit. So why is x64 emulating the 32bit instructions while my CPU
    > > can interpret them natively? Wouldn't that give me better program/driver
    > > support? If this is something that would require some long explanation
    > > (like if its something that involves the kernel's inner workings) then
    > > don't worry about it much and just say "its the kernel" or somethin, I'll
    > > get ya.
    > >
    > > Thanks a lot!
    > > Elliot

    >
    >
    >
    =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=, Jul 23, 2006
    #3
  4. No, no 32bit system level drivers. In fact, ALL drivers in 64bit Vista must
    be signed. Personally, I'm glad MS is taking this step, even though I know
    it may, in the short term, make things slightly worse. In the long term,
    we'll all benefit. And with the logo requirements for Vista, I think we'll
    see the driver situation get a lot better.

    Glad you liked the paper.


    --
    Charlie.
    http://msmvps.com/xperts64


    Elliot Hudgins wrote:
    > Wow! Your paper is really, really thought out and explained pretty much
    > everything I needed to know! Thanks a lot for the info. I guess that I
    > just wished for 32 bit drivers to work in 64, I miss my tuner :p
    >
    > Let's all hope that driver support gets better once Vista comes around :D
    > Oh, and do you know if it will have some type of megalo-awesome emulation
    > layer that WILL run 32-bit drivers? If you say it will, I'll love you.
    > Forever. Not in the man to man way, but I'll give you a hug in public and
    > not think twice. :p
    >
    > Elliot
    >
    > "Charlie Russel - MVP" wrote:
    >
    >> The kernel. ;)
    >>
    >> Seriously, it _does_ take advantage of the native support. But there
    >> still needs to have a very thin layer that translates between the 64 bit
    >> OS and the 32-bit application. This layer really doesn't impose any
    >> overhead, and many 32-bit programs actually appear faster in 64bit,
    >> though it really depends a lot on the program. The 64bit I/O subsystem
    >> is definitely more efficient, however, so programs that use a lot of I/O
    >> can sometimes be faster in 64 bit windows even though they are 32bit
    >> programs and they run in the WOW64 subsystem. There's a good graphic of
    >> this in the whitepaper I wrote for Microsoft:
    >> http://download.microsoft.com/downl..._XP_Professional_x64_Edition_Right_for_Me.doc
    >>
    >> You might find it of interest.
    >>
    >> --
    >> Charlie.
    >> http://msmvps.com/xperts64
    >>
    >>
    >> Elliot Hudgins wrote:
    >>> I don't quite understand the approach Microsoft has taken with x64. I
    >>> have an AMD Athlon 64 3500+, and it supports both 64bit instructions and
    >>> the 32bit. So why is x64 emulating the 32bit instructions while my CPU
    >>> can interpret them natively? Wouldn't that give me better program/driver
    >>> support? If this is something that would require some long explanation
    >>> (like if its something that involves the kernel's inner workings) then
    >>> don't worry about it much and just say "its the kernel" or somethin,
    >>> I'll get ya.
    >>>
    >>> Thanks a lot!
    >>> Elliot
    Charlie Russel - MVP, Jul 23, 2006
    #4
  5. =?Utf-8?B?RWxsaW90IEh1ZGdpbnM=?=

    RonK Guest

    Waht make and model is your Tuner Card ?
    You might find a usable driver at PlanetAMD64:
    http://www.planetamd64.com/


    "Elliot Hudgins" <> wrote in message
    news:...
    > Wow! Your paper is really, really thought out and explained pretty much
    > everything I needed to know! Thanks a lot for the info. I guess that I
    > just
    > wished for 32 bit drivers to work in 64, I miss my tuner :p
    >
    > Let's all hope that driver support gets better once Vista comes around :D
    > Oh, and do you know if it will have some type of megalo-awesome emulation
    > layer that WILL run 32-bit drivers? If you say it will, I'll love you.
    > Forever. Not in the man to man way, but I'll give you a hug in public and
    > not
    > think twice. :p
    >
    > Elliot
    >
    > "Charlie Russel - MVP" wrote:
    >
    >> The kernel. ;)
    >>
    >> Seriously, it _does_ take advantage of the native support. But there
    >> still
    >> needs to have a very thin layer that translates between the 64 bit OS and
    >> the 32-bit application. This layer really doesn't impose any overhead,
    >> and
    >> many 32-bit programs actually appear faster in 64bit, though it really
    >> depends a lot on the program. The 64bit I/O subsystem is definitely more
    >> efficient, however, so programs that use a lot of I/O can sometimes be
    >> faster in 64 bit windows even though they are 32bit programs and they run
    >> in
    >> the WOW64 subsystem. There's a good graphic of this in the whitepaper I
    >> wrote for Microsoft:
    >> http://download.microsoft.com/downl..._XP_Professional_x64_Edition_Right_for_Me.doc
    >>
    >> You might find it of interest.
    >>
    >> --
    >> Charlie.
    >> http://msmvps.com/xperts64
    >>
    >>
    >> Elliot Hudgins wrote:
    >> > I don't quite understand the approach Microsoft has taken with x64. I
    >> > have an AMD Athlon 64 3500+, and it supports both 64bit instructions
    >> > and
    >> > the 32bit. So why is x64 emulating the 32bit instructions while my CPU
    >> > can interpret them natively? Wouldn't that give me better
    >> > program/driver
    >> > support? If this is something that would require some long explanation
    >> > (like if its something that involves the kernel's inner workings) then
    >> > don't worry about it much and just say "its the kernel" or somethin,
    >> > I'll
    >> > get ya.
    >> >
    >> > Thanks a lot!
    >> > Elliot

    >>
    >>
    >>
    RonK, Jul 24, 2006
    #5
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Dim
    Replies:
    6
    Views:
    855
    Oxford Systems
    Jun 21, 2004
  2. The Frozen Canuck
    Replies:
    1
    Views:
    787
    Chuck Walbourn [MSFT]
    Jan 16, 2006
  3. Daniel
    Replies:
    11
    Views:
    605
  4. markm75
    Replies:
    7
    Views:
    1,082
    S.SubZero
    Jan 9, 2008
  5. Corsair
    Replies:
    6
    Views:
    1,031
    Corsair
    Jul 4, 2009
Loading...

Share This Page