Sign in with
Sign up | Sign in
Your question

Help - 32-bit Vs 64-bit

Tags:
Last response: in CPUs
Share
Anonymous
a b à CPUs
June 9, 2005 11:30:09 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

I am assisting my IT Director in doing an analysis for my company about
switching to 64 bit machines. My company is a large group with many
business divisions - some small and small big. Each operates
independently. There various different spplications that are run in the
company (not all in the same machine) - ERP, custom built database
applications, CAD/CAM applications, number crunching modelling
software, MS office, email, etc.

Some of the questions that I am addressing are:

1. Should we be buying 64-bit now or wait? What does it mean to the
company? To the IT department?

2. A lot of people say because there are not many 64 bit applications
and there is no point in migrating to 64-bit. Is this true?

3. Terminal Services - how many more users will 64-bit TS support? How
does 64-bit TS compare to 32-bit TS?

4. Are there any compelling reasons for us to consider 64 bit
processors?

If there is anything significant you want to add, please do. I would
also appreciate links to useful information about these issues.

Thanks in advance.

Sowmya

More about : bit bit

Anonymous
a b à CPUs
June 9, 2005 12:38:08 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

Seems like you need to do more research within your company. You
mention all the specialized software that you are using, this in its
self would be a reason not to switch.

Also has someone done a study in your company about Open Source, seems
you could save more money switching to an Open Source solution then
moving to 64 bits.

You also do not mention your support needs, is it done in house? Will
your needed hardware have the needed driver support?

My suggestion is hire a consultant, that specializes in this type of
evaluation. If the right person was hired and did a good job, you
could save millions of dollars, and only upgrade what is needed. Most
likely the only hardware that would really benefit would be the
servers, and those few applications that can really use the extra
memory, and registers that 64 bit can provide.

Everything else would be some what of a waste, sure every one wants a
new system, but do you really think the office worker who only types
documents, other basic office applications will need 64 bit, probably
not.

As I suggested earlier you probably save more money with Open Source
then a company wide switch to 64 bit.

Rthoreau
June 10, 2005 2:04:25 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Thu, 09 Jun 2005 08:38:08 -0700, Rthoreau wrote:

> Seems like you need to do more research within your company. You
> mention all the specialized software that you are using, this in its
> self would be a reason not to switch.

Not a reason not to either. 64bit platforms may improve productivity
significantly, or may not. An internal "beta", both for servers and
workstations would be the direction I'd look. That is, there is no reason
to commit either.

> Also has someone done a study in your company about Open Source, seems
> you could save more money switching to an Open Source solution then
> moving to 64 bits.

Perhaps, but these are two very different issues. There is less risk in
going 64bit than throwing over an OS for "open-source". Perhaps not as
much to gain, but *far* less risk. 64bit systems work just fine along
side of 32bit systems. The same can't be said for open-source next to
WinWhatever.

> You also do not mention your support needs, is it done in house? Will
> your needed hardware have the needed driver support?

Red-herring. Driver support is a minor issue. Again, a small "beta"
program will quickly ring out any problems with minimal impact. Drivers
for "open-source" is far more problematic.

> My suggestion is hire a consultant, that specializes in this type of
> evaluation. If the right person was hired and did a good job, you could
> save millions of dollars, and only upgrade what is needed. Most likely
> the only hardware that would really benefit would be the servers, and
> those few applications that can really use the extra memory, and
> registers that 64 bit can provide.

OTOH, the cost of the hardware is infintessimal. I'd try it and see. I
certainly would *not* try "open-source" without a rather in-depth study.

> Everything else would be some what of a waste, sure every one wants a
> new system, but do you really think the office worker who only types
> documents, other basic office applications will need 64 bit, probably
> not.

Likely not. Does it really matter though? The cost of that person is
*so* much larger than the hardware it's a nit. Indeed I'd think "laptop
and dock" before either, for professionals. Hourlys, I'd go with
solid desktops (no, not 64b).

> As I suggested earlier you probably save more money with Open Source
> then a company wide switch to 64 bit.

It's a *lot* riskier too. There may be a gain, but it's not clear. I'd
certainly not recommend this path to anyone who doesn't want to deal with
a lot of pain. By comparison, 64bit is simple. It simply works, even if
it's not used.

--
Keith
Related resources
Anonymous
a b à CPUs
June 10, 2005 12:54:39 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On 9 Jun 2005 07:30:09 -0700, "it_executive" <sowmya.eshwar@gmail.com>
wrote:

>I am assisting my IT Director in doing an analysis for my company about
>switching to 64 bit machines. My company is a large group with many
>business divisions - some small and small big. Each operates
>independently. There various different spplications that are run in the
>company (not all in the same machine) - ERP, custom built database
>applications, CAD/CAM applications, number crunching modelling
>software, MS office, email, etc.
>
>Some of the questions that I am addressing are:
>
>1. Should we be buying 64-bit now or wait? What does it mean to the
>company? To the IT department?

I see no reason to avoid 64-bit hardware right now - it might be kinda
difficult at the top end of x86, since AMD64 and EM64T are in both CPU
lines. ERP, CAD/CAM and numerically intensive modelling are all going to
benefit from either the increased address space -- per process and across
the system -- or the increased register count of x86-64. Talk to your
software vendors about timetables for availability - might give them some
incentive to move things along; see what in-house developers think about
availability of tools and the scope of the task.

>2. A lot of people say because there are not many 64 bit applications
>and there is no point in migrating to 64-bit. Is this true?

The beauty of x86-64 is that you don't have to take the plunge across the
board - if you need hardware anyway, get the systems which can run 64-bit
and run them as 32-bit initially. Take a couple for experimenting and try
loading up a 64-bit OS and see how they run your 32-bit apps - if
everything works well, you're in good shape to add 64-bit apps as they
appear or are developed in-house.

>3. Terminal Services - how many more users will 64-bit TS support? How
>does 64-bit TS compare to 32-bit TS?

I've no experience with TS but I'd think the large virtual space and the
larger physical memory would allow more TS sessions to co-exist without
paging/swapping disk I/O overhead.

>4. Are there any compelling reasons for us to consider 64 bit
>processors?

Are there any reasons to avoid 64-bit capable processors from either AMD or
Intel? They don't cost more.

>If there is anything significant you want to add, please do. I would
>also appreciate links to useful information about these issues.

This is an AMD marketing site http://72.3.250.42/amd_05e1/index.html but
you might find some useful info and maybe reasons: choose "AMD In the
Enterprise" and then click on "Enterprise Success Stories".

--
Rgds, George Macdonald
Anonymous
a b à CPUs
June 10, 2005 5:29:32 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

George Macdonald wrote:
> On 9 Jun 2005 07:30:09 -0700, "it_executive" <sowmya.eshwar@gmail.com>
> wrote:
>
>
>>I am assisting my IT Director in doing an analysis for my company about
>>switching to 64 bit machines. My company is a large group with many
>>business divisions - some small and small big. Each operates
>>independently. There various different spplications that are run in the
>>company (not all in the same machine) - ERP, custom built database
>>applications, CAD/CAM applications, number crunching modelling
>>software, MS office, email, etc.
>>
>>Some of the questions that I am addressing are:
>>
>>1. Should we be buying 64-bit now or wait? What does it mean to the
>>company? To the IT department?
>
>
> I see no reason to avoid 64-bit hardware right now - it might be kinda
> difficult at the top end of x86, since AMD64 and EM64T are in both CPU
> lines.

I would be a little more specific: there is no reason to avoid
x86-64 *except* for Intel dual-core chips. There are simply too
many reports of heat/power related problems with the Intel
dual-core chips.

Personally, I would avoid P4 and Xeon for all but a few specific
tasks, such as the oft cited video encoding, regardless of
whether or not they are EMT64 versions. For most tasks they use
a lot more power to do less work than their AMD counterparts.

> ERP, CAD/CAM and numerically intensive modelling are all going to
> benefit from either the increased address space -- per process and across
> the system -- or the increased register count of x86-64. Talk to your
> software vendors about timetables for availability - might give them some
> incentive to move things along; see what in-house developers think about
> availability of tools and the scope of the task.
>
>
>>2. A lot of people say because there are not many 64 bit applications
>>and there is no point in migrating to 64-bit. Is this true?
>
>
> The beauty of x86-64 is that you don't have to take the plunge across the
> board - if you need hardware anyway, get the systems which can run 64-bit
> and run them as 32-bit initially. Take a couple for experimenting and try
> loading up a 64-bit OS and see how they run your 32-bit apps - if
> everything works well, you're in good shape to add 64-bit apps as they
> appear or are developed in-house.
>
>
>>3. Terminal Services - how many more users will 64-bit TS support? How
>>does 64-bit TS compare to 32-bit TS?
>
>
> I've no experience with TS but I'd think the large virtual space and the
> larger physical memory would allow more TS sessions to co-exist without
> paging/swapping disk I/O overhead.
>
>
>>4. Are there any compelling reasons for us to consider 64 bit
>>processors?
>
>
> Are there any reasons to avoid 64-bit capable processors from either AMD or
> Intel? They don't cost more.
>
>
>>If there is anything significant you want to add, please do. I would
>>also appreciate links to useful information about these issues.
>
>
> This is an AMD marketing site http://72.3.250.42/amd_05e1/index.html but
> you might find some useful info and maybe reasons: choose "AMD In the
> Enterprise" and then click on "Enterprise Success Stories".
>
Anonymous
a b à CPUs
June 10, 2005 11:41:16 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Fri, 10 Jun 2005 13:29:32 GMT, Rob Stow <rob.stow@shaw.ca> wrote:

>George Macdonald wrote:
>> On 9 Jun 2005 07:30:09 -0700, "it_executive" <sowmya.eshwar@gmail.com>
>> wrote:
>>
>>
>>>I am assisting my IT Director in doing an analysis for my company about
>>>switching to 64 bit machines. My company is a large group with many
>>>business divisions - some small and small big. Each operates
>>>independently. There various different spplications that are run in the
>>>company (not all in the same machine) - ERP, custom built database
>>>applications, CAD/CAM applications, number crunching modelling
>>>software, MS office, email, etc.
>>>
>>>Some of the questions that I am addressing are:
>>>
>>>1. Should we be buying 64-bit now or wait? What does it mean to the
>>>company? To the IT department?
>>
>>
>> I see no reason to avoid 64-bit hardware right now - it might be kinda
>> difficult at the top end of x86, since AMD64 and EM64T are in both CPU
>> lines.
>
>I would be a little more specific: there is no reason to avoid
>x86-64 *except* for Intel dual-core chips. There are simply too
>many reports of heat/power related problems with the Intel
>dual-core chips.

Yeah I hadn't really been following that so Yousuf's post yesterday was a
bit of a revelation - I'd no idea "i" had made such a hash of it... by
THG's "results" anyway. I'll have to hunt down some more.

>Personally, I would avoid P4 and Xeon for all but a few specific
>tasks, such as the oft cited video encoding, regardless of
>whether or not they are EMT64 versions. For most tasks they use
>a lot more power to do less work than their AMD counterparts.

As an apparently registered AMDroid, I'll probably never have any personal
experience with EM64T anyway.:-) On the subject though, I had the weirdest
angle on it presented by a local small builder recently: dropped in to pick
up a hard disk, in an emergency, and got talking about CPUs to see what he
was currently offering. He tells me that AMD had just tacked 32 bits on to
a 32-bit design to get 64-bits; Intel, however, had done things properly
and designed a "real 64-bit CPU" and that was the reason they ran so hot.

I laughed, shook my head and told him maybe he had it backwards, whereupon
he got quite indignant and wanted to argue - I changed the subject and left
quickly but I've been wondering since, where he got the story:
self-invented folklore... or is this what Intel is telling their vendors?

--
Rgds, George Macdonald
June 12, 2005 1:46:03 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:


> As an apparently registered AMDroid,

Gee George, I thought we were in this together for some time. I can
even come out of the closet at work now, though I haven't exactly
been silent. ;-) My cow-orkers were wondering about my buying an Opteron
a year ago. Well...

> I'll probably never have any personal
> experience with EM64T anyway.:-) On the subject though, I had the weirdest
> angle on it presented by a local small builder recently: dropped in to pick
> up a hard disk, in an emergency, and got talking about CPUs to see what he
> was currently offering. He tells me that AMD had just tacked 32 bits on to
> a 32-bit design to get 64-bits; Intel, however, had done things properly
> and designed a "real 64-bit CPU" and that was the reason they ran so hot.

I see that nonsense all the time. I smile, tell them that they haven't a
clue, and walk away. It's not worth the agravation trying to enlighten
hardware sales-droids. There's no money there, so they hire only
pimply-faced kids who know everything already.

To be fair, my son manages a game store and he tries to hire the best
gamers he can. They haven't a clue about hardware though.

> I laughed, shook my head and told him maybe he had it backwards, whereupon
> he got quite indignant and wanted to argue - I changed the subject and left
> quickly but I've been wondering since, where he got the story:
> self-invented folklore... or is this what Intel is telling their vendors?

Chicken! ;-)

--
Keith
Anonymous
a b à CPUs
June 12, 2005 1:55:00 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

"keith" <krw@att.bizzzz> wrote in message
news:p an.2005.06.12.01.46.02.234291@att.bizzzz...
> On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>
>
>> As an apparently registered AMDroid,
>
> Gee George, I thought we were in this together for some time. I can
> even come out of the closet at work now, though I haven't exactly
> been silent. ;-) My cow-orkers were wondering about my buying an
> Opteron
> a year ago. Well...
>
>> I'll probably never have any personal
>> experience with EM64T anyway.:-) On the subject though, I had the
>> weirdest
>> angle on it presented by a local small builder recently: dropped in to
>> pick
>> up a hard disk, in an emergency, and got talking about CPUs to see
>> what he
>> was currently offering. He tells me that AMD had just tacked 32 bits
>> on to
>> a 32-bit design to get 64-bits; Intel, however, had done things
>> properly
>> and designed a "real 64-bit CPU" and that was the reason they ran so
>> hot.
>
> I see that nonsense all the time. I smile, tell them that they haven't
> a
> clue, and walk away. It's not worth the agravation trying to enlighten
> hardware sales-droids. There's no money there, so they hire only
> pimply-faced kids who know everything already.
>
> To be fair, my son manages a game store and he tries to hire the best
> gamers he can. They haven't a clue about hardware though.
>
Actually the story is totally correct. It is just that a few things got
left out. Intel did go off and do a 64 bit arch. first, from scratch,
clean paper, all that. Set out to do it "right". Y'all remember Itanium
don't you? Sure you do.

AMD just took the kludgy X86 architecture that had started out as a
pocket calculator and gave it 64 bit addressing and a few more registers
and some other stuff, "pasted on".

The part that got left out was that Itanium, supposedly the thing more
advanced and done "right" didn't meet customer needs as well as the thing
AMD "pasted together" out of the kludgy X86 and a few leftover address
bits. :-)

Sounds pretty close to me....

del
>> I laughed, shook my head and told him maybe he had it backwards,
>> whereupon
>> he got quite indignant and wanted to argue - I changed the subject and
>> left
>> quickly but I've been wondering since, where he got the story:
>> self-invented folklore... or is this what Intel is telling their
>> vendors?
>
> Chicken! ;-)
>
> --
> Keith
>
June 12, 2005 3:01:36 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sat, 11 Jun 2005 21:55:00 -0500, Del Cecchi wrote:

>
> "keith" <krw@att.bizzzz> wrote in message
> news:p an.2005.06.12.01.46.02.234291@att.bizzzz...
>> On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>>
>>
>>> As an apparently registered AMDroid,
>>
>> Gee George, I thought we were in this together for some time. I can
>> even come out of the closet at work now, though I haven't exactly
>> been silent. ;-) My cow-orkers were wondering about my buying an
>> Opteron
>> a year ago. Well...
>>
>>> I'll probably never have any personal
>>> experience with EM64T anyway.:-) On the subject though, I had the
>>> weirdest
>>> angle on it presented by a local small builder recently: dropped in to
>>> pick
>>> up a hard disk, in an emergency, and got talking about CPUs to see
>>> what he
>>> was currently offering. He tells me that AMD had just tacked 32 bits
>>> on to
>>> a 32-bit design to get 64-bits; Intel, however, had done things
>>> properly
>>> and designed a "real 64-bit CPU" and that was the reason they ran so
>>> hot.
>>
>> I see that nonsense all the time. I smile, tell them that they haven't
>> a
>> clue, and walk away. It's not worth the agravation trying to enlighten
>> hardware sales-droids. There's no money there, so they hire only
>> pimply-faced kids who know everything already.
>>
>> To be fair, my son manages a game store and he tries to hire the best
>> gamers he can. They haven't a clue about hardware though.
>>
> Actually the story is totally correct. It is just that a few things got
> left out. Intel did go off and do a 64 bit arch. first, from scratch,
> clean paper, all that. Set out to do it "right". Y'all remember Itanium
> don't you? Sure you do.

No, umm... Really?!

> AMD just took the kludgy X86 architecture that had started out as a
> pocket calculator and gave it 64 bit addressing and a few more registers
> and some other stuff, "pasted on".

Sure, but isn't bubble-gum what holds the world together?

> The part that got left out was that Itanium, supposedly the thing more
> advanced and done "right" didn't meet customer needs as well as the thing
> AMD "pasted together" out of the kludgy X86 and a few leftover address
> bits. :-)
>
> Sounds pretty close to me....

Well, "right" is a loaded term too. "Right" for whom?

It seems it hasn't even beed "right" for Intel. Too bad...

--
Keith
Anonymous
a b à CPUs
June 12, 2005 7:16:09 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

keith wrote:
> On Sat, 11 Jun 2005 21:55:00 -0500, Del Cecchi wrote:
>

>>AMD just took the kludgy X86 architecture that had started out as a
>>pocket calculator and gave it 64 bit addressing and a few more registers
>>and some other stuff, "pasted on".
>
>
> Sure, but isn't bubble-gum what holds the world together?

No. Duct tape and WD-40 are what makes everything work.
If it moves and it shouldn't, use duct tape.
If it doesn't move and it should, use WD-40.
June 12, 2005 7:16:10 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sun, 12 Jun 2005 03:16:09 +0000, Rob Stow wrote:

> keith wrote:
>> On Sat, 11 Jun 2005 21:55:00 -0500, Del Cecchi wrote:
>>
>
>>>AMD just took the kludgy X86 architecture that had started out as a
>>>pocket calculator and gave it 64 bit addressing and a few more registers
>>>and some other stuff, "pasted on".
>>
>>
>> Sure, but isn't bubble-gum what holds the world together?
>
> No. Duct tape and WD-40 are what makes everything work.
> If it moves and it shouldn't, use duct tape.
> If it doesn't move and it should, use WD-40.

I like Red Green too. It's proof that the Canuckistanis aren't
totally useless. ;-)

--
Keith
Anonymous
a b à CPUs
June 12, 2005 7:40:33 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sat, 11 Jun 2005 21:46:03 -0400, keith <krw@att.bizzzz> wrote:

>On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>
>
>> As an apparently registered AMDroid,
>
>Gee George, I thought we were in this together for some time.

Oh sure - there are even some who think this NG is a cabal of AMD
subversives... when all that's really happened is that the "i" crowd have
nothing to crow about... for the moment.:-)

> I can
>even come out of the closet at work now, though I haven't exactly
>been silent. ;-) My cow-orkers were wondering about my buying an Opteron
>a year ago. Well...

Hmm, I'd have thought they might have appreciated a "partner" corp's
success based on the adoption of "their" technology. I guess the reported
$46M of Nov '02 was small potatoes??

>> I'll probably never have any personal
>> experience with EM64T anyway.:-) On the subject though, I had the weirdest
>> angle on it presented by a local small builder recently: dropped in to pick
>> up a hard disk, in an emergency, and got talking about CPUs to see what he
>> was currently offering. He tells me that AMD had just tacked 32 bits on to
>> a 32-bit design to get 64-bits; Intel, however, had done things properly
>> and designed a "real 64-bit CPU" and that was the reason they ran so hot.
>
>I see that nonsense all the time. I smile, tell them that they haven't a
>clue, and walk away. It's not worth the agravation trying to enlighten
>hardware sales-droids. There's no money there, so they hire only
>pimply-faced kids who know everything already.
>
>To be fair, my son manages a game store and he tries to hire the best
>gamers he can. They haven't a clue about hardware though.

This guy is no gamer - he's had a successful small builder store for years,
selling to the likes of AT&T back in the day. He used to be quite rational
about it all but I get the feeling he's been nobbled.

Mind you the gamers are the ones who are clued in about AMD - some of the
forums are just like AMD sites now... with overclocking stories which are
hair-raising.

--
Rgds, George Macdonald
Anonymous
a b à CPUs
June 12, 2005 7:40:33 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sat, 11 Jun 2005 21:55:00 -0500, "Del Cecchi" <dcecchi.nospam@att.net>
wrote:

>
>"keith" <krw@att.bizzzz> wrote in message
>news:p an.2005.06.12.01.46.02.234291@att.bizzzz...
>> On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>>
>>
>>> As an apparently registered AMDroid,
>>
>> Gee George, I thought we were in this together for some time. I can
>> even come out of the closet at work now, though I haven't exactly
>> been silent. ;-) My cow-orkers were wondering about my buying an
>> Opteron
>> a year ago. Well...
>>
>>> I'll probably never have any personal
>>> experience with EM64T anyway.:-) On the subject though, I had the
>>> weirdest
>>> angle on it presented by a local small builder recently: dropped in to
>>> pick
>>> up a hard disk, in an emergency, and got talking about CPUs to see
>>> what he
>>> was currently offering. He tells me that AMD had just tacked 32 bits
>>> on to
>>> a 32-bit design to get 64-bits; Intel, however, had done things
>>> properly
>>> and designed a "real 64-bit CPU" and that was the reason they ran so
>>> hot.
>>
>> I see that nonsense all the time. I smile, tell them that they haven't
>> a
>> clue, and walk away. It's not worth the agravation trying to enlighten
>> hardware sales-droids. There's no money there, so they hire only
>> pimply-faced kids who know everything already.
>>
>> To be fair, my son manages a game store and he tries to hire the best
>> gamers he can. They haven't a clue about hardware though.
>>
>Actually the story is totally correct. It is just that a few things got
>left out. Intel did go off and do a 64 bit arch. first, from scratch,
>clean paper, all that. Set out to do it "right". Y'all remember Itanium
>don't you? Sure you do.
>
>AMD just took the kludgy X86 architecture that had started out as a
>pocket calculator and gave it 64 bit addressing and a few more registers
>and some other stuff, "pasted on".
>
>The part that got left out was that Itanium, supposedly the thing more
>advanced and done "right" didn't meet customer needs as well as the thing
>AMD "pasted together" out of the kludgy X86 and a few leftover address
>bits. :-)
>
>Sounds pretty close to me....

Yeah but this was in particular to do with EM64T being a "real 64-bit chip"
- pretty mangled version of the story. In fact if you believe the analysis
and pics at www.chip-architect.com EM64T is a 32-bit "bit-slice" design.

--
Rgds, George Macdonald
June 12, 2005 12:51:11 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sun, 12 Jun 2005 03:40:33 -0400, George Macdonald wrote:

> On Sat, 11 Jun 2005 21:46:03 -0400, keith <krw@att.bizzzz> wrote:
>
>>On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>>
>>
>>> As an apparently registered AMDroid,
>>
>>Gee George, I thought we were in this together for some time.
>
> Oh sure - there are even some who think this NG is a cabal of AMD
> subversives... when all that's really happened is that the "i" crowd have
> nothing to crow about... for the moment.:-)

Subversive? Me?

>> I can
>>even come out of the closet at work now, though I haven't exactly
>>been silent. ;-) My cow-orkers were wondering about my buying an Opteron
>>a year ago. Well...
>
> Hmm, I'd have thought they might have appreciated a "partner" corp's
> success based on the adoption of "their" technology. I guess the reported
> $46M of Nov '02 was small potatoes??

We're a little too far from that deal for it to make any difference. I've
never even been on a tour of a kitchen, much less cooked a pizza.
However, since our partner went to the dark side...

>>> I'll probably never have any personal experience with EM64T anyway.:-)
>>> On the subject though, I had the weirdest angle on it presented by a
>>> local small builder recently: dropped in to pick up a hard disk, in an
>>> emergency, and got talking about CPUs to see what he was currently
>>> offering. He tells me that AMD had just tacked 32 bits on to a 32-bit
>>> design to get 64-bits; Intel, however, had done things properly and
>>> designed a "real 64-bit CPU" and that was the reason they ran so hot.
>>
>>I see that nonsense all the time. I smile, tell them that they haven't a
>>clue, and walk away. It's not worth the agravation trying to enlighten
>>hardware sales-droids. There's no money there, so they hire only
>>pimply-faced kids who know everything already.
>>
>>To be fair, my son manages a game store and he tries to hire the best
>>gamers he can. They haven't a clue about hardware though.
>
> This guy is no gamer - he's had a successful small builder store for
> years, selling to the likes of AT&T back in the day. He used to be
> quite rational about it all but I get the feeling he's been nobbled.
>
> Mind you the gamers are the ones who are clued in about AMD - some of
> the forums are just like AMD sites now... with overclocking stories
> which are hair-raising.

They're not timmid, that's for sure!

--
Keith
Anonymous
a b à CPUs
June 12, 2005 8:04:25 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

keith wrote:
> On Sun, 12 Jun 2005 03:16:09 +0000, Rob Stow wrote:
>
>
>>keith wrote:
>>
>>>On Sat, 11 Jun 2005 21:55:00 -0500, Del Cecchi wrote:
>>>
>>
>>>>AMD just took the kludgy X86 architecture that had started out as a
>>>>pocket calculator and gave it 64 bit addressing and a few more registers
>>>>and some other stuff, "pasted on".
>>>
>>>
>>>Sure, but isn't bubble-gum what holds the world together?
>>
>>No. Duct tape and WD-40 are what makes everything work.
>>If it moves and it shouldn't, use duct tape.
>>If it doesn't move and it should, use WD-40.
>
>
> I like Red Green too. It's proof that the Canuckistanis aren't
> totally useless. ;-)
>

I thought about providing the proper attribution but I was sure
it would just go over the heads of all you Yankees. ;-)
Anonymous
a b à CPUs
June 13, 2005 12:18:35 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sun, 12 Jun 2005 08:51:11 -0400, keith <krw@att.bizzzz> wrote:

>On Sun, 12 Jun 2005 03:40:33 -0400, George Macdonald wrote:
>
>> On Sat, 11 Jun 2005 21:46:03 -0400, keith <krw@att.bizzzz> wrote:
>>
>>>On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>>>
>>>
>>>> As an apparently registered AMDroid,
>>>
>>>Gee George, I thought we were in this together for some time.
>>
>> Oh sure - there are even some who think this NG is a cabal of AMD
>> subversives... when all that's really happened is that the "i" crowd have
>> nothing to crow about... for the moment.:-)
>
>Subversive? Me?

AMD revisionist?:-)

>>> I can
>>>even come out of the closet at work now, though I haven't exactly
>>>been silent. ;-) My cow-orkers were wondering about my buying an Opteron
>>>a year ago. Well...
>>
>> Hmm, I'd have thought they might have appreciated a "partner" corp's
>> success based on the adoption of "their" technology. I guess the reported
>> $46M of Nov '02 was small potatoes??
>
>We're a little too far from that deal for it to make any difference. I've
>never even been on a tour of a kitchen, much less cooked a pizza.
>However, since our partner went to the dark side...

Dark side? I was talking about the AMD rescue op. for SOI.

--
Rgds, George Macdonald
June 13, 2005 1:13:27 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sun, 12 Jun 2005 16:04:25 +0000, Rob Stow wrote:

> keith wrote:
>> On Sun, 12 Jun 2005 03:16:09 +0000, Rob Stow wrote:
>>
>>
>>>keith wrote:
>>>
>>>>On Sat, 11 Jun 2005 21:55:00 -0500, Del Cecchi wrote:
>>>>
>>>
>>>>>AMD just took the kludgy X86 architecture that had started out as a
>>>>>pocket calculator and gave it 64 bit addressing and a few more registers
>>>>>and some other stuff, "pasted on".
>>>>
>>>>
>>>>Sure, but isn't bubble-gum what holds the world together?
>>>
>>>No. Duct tape and WD-40 are what makes everything work.
>>>If it moves and it shouldn't, use duct tape.
>>>If it doesn't move and it should, use WD-40.
>>
>>
>> I like Red Green too. It's proof that the Canuckistanis aren't
>> totally useless. ;-)
>>
>
> I thought about providing the proper attribution but I was sure
> it would just go over the heads of all you Yankees. ;-)

Well, I record him on my DVR/DVD-recorder every weekday on CBC out of
Montreal. My son is PO'd that he can't get him in NH. DVDs will travel! ;-)

OTOH, my wife doesn't think he's funny at all. ...go figure wimmin.

--
Keith
June 13, 2005 1:22:53 AM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

On Sun, 12 Jun 2005 20:18:35 -0400, George Macdonald wrote:

> On Sun, 12 Jun 2005 08:51:11 -0400, keith <krw@att.bizzzz> wrote:
>
>>On Sun, 12 Jun 2005 03:40:33 -0400, George Macdonald wrote:
>>
>>> On Sat, 11 Jun 2005 21:46:03 -0400, keith <krw@att.bizzzz> wrote:
>>>
>>>>On Fri, 10 Jun 2005 19:41:16 -0400, George Macdonald wrote:
>>>>
>>>>
>>>>> As an apparently registered AMDroid,
>>>>
>>>>Gee George, I thought we were in this together for some time.
>>>
>>> Oh sure - there are even some who think this NG is a cabal of AMD
>>> subversives... when all that's really happened is that the "i" crowd have
>>> nothing to crow about... for the moment.:-)
>>
>>Subversive? Me?
>
> AMD revisionist?:-)

Revisionist? I was doing AMD K6s at work when we were working for the
"other guy". Subversive? Revisionist? I don't think so. ...perhaps
"renegade". ;-)

>>>> I can
>>>>even come out of the closet at work now, though I haven't exactly been
>>>>silent. ;-) My cow-orkers were wondering about my buying an Opteron
>>>>a year ago. Well...
>>>
>>> Hmm, I'd have thought they might have appreciated a "partner" corp's
>>> success based on the adoption of "their" technology. I guess the
>>> reported $46M of Nov '02 was small potatoes??
>>
>>We're a little too far from that deal for it to make any difference.
>>I've never even been on a tour of a kitchen, much less cooked a pizza.
>>However, since our partner went to the dark side...
>
> Dark side? I was talking about the AMD rescue op. for SOI.

I wasn't (though I know a little about that too; I never liked in
history when I was in grade-school, but I was wrong ;) . I was thinking
*closer* to home. Yes, *DARK* side. I thought you'd figured that part out
long ago. You know my email address.

--
Keith
Anonymous
a b à CPUs
June 13, 2005 3:14:19 PM

Archived from groups: comp.sys.ibm.pc.hardware.chips (More info?)

Del Cecchi wrote:
> AMD just took the kludgy X86 architecture that had started out as a
> pocket calculator and gave it 64 bit addressing and a few more registers
> and some other stuff, "pasted on".
>
> The part that got left out was that Itanium, supposedly the thing more
> advanced and done "right" didn't meet customer needs as well as the thing
> AMD "pasted together" out of the kludgy X86 and a few leftover address
> bits. :-)

Well, AMD did take the time to cleanup a lot of cruft from the
instruction set when they extended the instruction set out. Gone are
the segment registers. Gone are the "specialized" general-purpose
registers. Floating point comes now mainly through the randomly
accessible SSE registers rather than the old FPU stack. It's basically
been washed and ironed for the first time.

Yousuf Khan
!