AMD64-based laptops from HP/Compaq are crippled

[J. Clarke]

Has it not been made abundantly clear to you what chips are used in the
laptops in question?

No it has not. Athlon 64s are made in Socket 754 and Socket 939
variants--just saying "Athlon(TM) 64" tells you nothing whatsoever about
the socket.

It's disabled in BIOS. HP admits as much. The motherboard is socket 939.
Why don't you read the site and maybe think for a moment.

I read the site and it looked like a bunch of people whingeing. Searching
the page for "939" got no hits at all.

Further research indicates that apparently for some bizarre reason HP built
a laptop that cannot possibly use a mobile processor. I'd be more worried
about that than about whether it has the second memory channel enabled.

 

[NoNoBadDog!]

What do Turions have to do with it?


For some tasks. For others, dual-channel will run twice faster.

Dual Channel will not run "twice faster". There is overhead that will
prevent getting a doubling of the throughput. While the difference can be
dramatic, it will not be "twice as fast".

Bobby

 

[alex goldman]

I read the site and it looked like a bunch of people whingeing. Searching
the page for "939" got no hits at all.

Go to http://www.zv6000forums.com/search.php , type 939 into the first form,
click search.

I'm getting lots of hits, most of them relevant, I'm also getting tired of
your "helplessness" .

 

[alex goldman]

My dual channel AMD gets a bit less than twice the memory speed in
memtest86 when running dual channel.

No AMD runs DDR2.

Why they would disable dual channel I don't know. Since they can
completely configure the pc with compatible memory, it doesn't make sense,
unless they randomly throw whatever cheap memory they can find into it.
Then it figures.

What beats me is why NoNoBadDog would talk about things he's totally
clueless about in a way to suggest he's the ultimate "white paper" -reading
expert.

I also have a strong suspicion a few others in this thread are sock puppets
of the same physical person. The similarties in the use of obscene
language, self-contradiction, talking about things one does not know much
about while insinuating expertise, and helplessness in looking up simple
public-domain facts are too striking to be random.

I will therefore excuse myself from this thread. Hope it was helpful to
someone.

 

[General Schvantzkoph]

I've been researching my next laptop purchase, and whether it will run
Linux, and completely accidentally stumbled upon an alarming issue that
people need to be more aware about.

Apparently (and if my understanding is correct), HP/Compaq disables
dual-channel memory access in Socket 939 motherboards, making CPU <-> RAM
communication slower, and overall performance less than what people would
expect when buying a system with the outward characteristics of what they
think they are getting.

http://zv6000forums.com/viewtopic.php?t=148

OK, read some of the posts. Here is the likely reason that they turned off
dual channel memory, they ****ed up the design of machine and it won't run
reliably in dual channel mode. Assuming that they aren't complete idiots
what happened is that they intended to use both channels. If they were
planning for it to be a single channel machine they would have used a 754
pin part, which is what they did in the nx6125 or the ze2000z. When they
got to the lab it didn't work when both memory channels were running at
the same time. Rather then re-layout the pc board they disabled dual
channels in the BIOS and shipped it. They probably have a pc board spin in
the works, when it's ready they'll change the model number and ship it as
a new machine. The current ones are never going to run in dual channel
mode, it's not a bug in the BIOS it's a bug in the printed circuit board
layout. If you are going to buy an HP laptop today buy the nx6125 or the
ze2000z, those are available with Turions which are lower power then the
939 A64. Also most Turions have 1M of cache, the only exception
is the ML32 so stay away from that one. As I said in an earlier post cache
matters, dual channel doesn't. I know you think that dual channel is
important but it's not. I challenge you to name one real world program
where it makes even a 10% difference. I did some extensive benchmarking
doing heavy duty simulations, FPGA place and routes, GCC makes. The 3400+
with 1M of cache and a single data channel was never slower then the 3800+
with two channels, a faster clock, but only 1/2M of cache. On Verilog
simulations the 3400+ was twice as fast as the 3800+.

 

[Dan]

There are too many ignorant dumbasses on this board, that refuse to admit
they are wrong and half the statistics and facts they generate and just
random numbers from their head!

NoNObadDOG. You either have to be trying to annoy people on this board or
you're in Guinness Book OWRs under the record living human with the least
amount of brain cells

 

[Wes Newell]

Where did you get this idea?

From my considerable knowledge of how they work. And then there are
thounsands of benchmarks to prove it all over the web. find a benchmark of
apps that have a 3400+ single channel cpu and a 3500+ dual channel listed
in the same benchmark. Now guess which one wins most of the benchmark
test. if you think the 3500+ you'd be wrong. Now compare the 3400+ to a
3800+. Both have the same clockspeed of 2.4GHz, and the same 512K cache.
Only difference is the dual channel ram. Does the 3800+ win all
benchmarks? No. And the reason is pretty simple. The overhead of the dual
channel slows it down when high bandwidth isn't needed. And the onboard
very low latency memory controller allows the single channel to service
the data request very fast compared to old chipset controller that had to
send the data over a shared FSB. The lack of a major performance increase
for dual channel is probably why HP decided not to use it if that's what
they did. I still think leaving it out was stupid though. But in reality,
it's not a big deal or advantage. Dual cores may prove to be a different
story since they will have much more demand. Time will tell.

 

[Russell Shaw]

From my considerable knowledge of how they work. And then there are
thounsands of benchmarks to prove it all over the web. find a benchmark of
apps that have a 3400+ single channel cpu and a 3500+ dual channel listed
in the same benchmark. Now guess which one wins most of the benchmark
test. if you think the 3500+ you'd be wrong. Now compare the 3400+ to a
3800+. Both have the same clockspeed of 2.4GHz, and the same 512K cache.
Only difference is the dual channel ram. Does the 3800+ win all
benchmarks? No. And the reason is pretty simple. The overhead of the dual
channel slows it down when high bandwidth isn't needed. And the onboard
very low latency memory controller allows the single channel to service
the data request very fast compared to old chipset controller that had to
send the data over a shared FSB. The lack of a major performance increase
for dual channel is probably why HP decided not to use it if that's what
they did. I still think leaving it out was stupid though. But in reality,
it's not a big deal or advantage. Dual cores may prove to be a different
story since they will have much more demand. Time will tell.

It may lead to a greater increase in battery lifetime compared to
speed lost.

 

[Doug Lynn]

Hi, aren't Turions all socket 754...???

Thats what I read...

 

[General Schvantzkoph]

The laptops use the Turion chip. According to AMD, the Turion's
integrated memory controller can use PC1600, PC2100, PC2700 or PC3200
unbuffered DDR SO-DIMMs and has a maximum throughput of 3.2 GB/s.

So, explain to me what engineering in dual channel support is going to
buy you. BTW, does any vendor manufacture dual channel compatible
SO-DIMMs?

HP has a number of different A64 notebooks, at least one (R4000) uses a
939 pin processor which could have supported dual channel memory if they
had chosen to do so. All laptops in this range have two DIMM sockets so
dual channel memory is entirely doable assuming that the processor has
support for it (there is no such thing as a dual channel DIMM, dual
channel memory systems use DIMMs in pairs).

The latest laptops all use the Turion which appears to be a single
channel processor. However the Turion is available with a 1M cache which
more then makes up for the lack of dual memory channels.

BTW I have a Compaq R3000z with an 3400+ in it. It's screaming fast and
I've had no trouble running Fedora Core 3, 32 and 64 bit on it. The only
thing that's a pain is the built in Broadcom wireless chip which lacks a
native Linux driver. It can be made to work using Ndiswrapper but that's a
lot less convenient then having a wireless card with a native Linux
driver. If you get an HP/Compaq and they are still using the Broadcom
wireless chip (I bought mine a year ago) I'd recommend that you delete
that option and buy a Linux compatible PCMCIA card.

As for buying an HP/Compaq. The HPs and the Compaqs are identical except
for color. On any give day the price on http://www.hpshopping.com for
identical models can vary significantly between the HP and Compaq versions
depending on the specials of the day so check both models. HP also sells
systems without Windows, you can get Linux or Freedos instead. They don't
offer these systems on hpshopping, you have to go to http://www.hp.com and
look up business systems. I don't know if this is actually cheaper then
paying the Windows tax on hpshopping.com, you'll have to do the
comparisons for the model that you are interested in yourself.

 

[Marek Williams]

As for buying an HP/Compaq. The HPs and the Compaqs are identical except
for color. On any give day the price on http://www.hpshopping.com for
identical models can vary significantly between the HP and Compaq versions
depending on the specials of the day so check both models. HP also sells
systems without Windows, you can get Linux or Freedos instead. They don't
offer these systems on hpshopping, you have to go to http://www.hp.com and
look up business systems. I don't know if this is actually cheaper then
paying the Windows tax on hpshopping.com, you'll have to do the
comparisons for the model that you are interested in yourself.

I had an R4000 before I sent it back under HP's "no hassle" return
policy. Now I have an R3240, which is much better. I don't see any
speed difference in the R3240. Notice that I said I don't see any
speed difference. Benchmark tests might tell a different story. But
from the standpoint of the user, if the difference isn't noticeable,
then who cares?

It took a long time (because I'm a n00b), but I finally got the wifi
working with ndiswrapper. There's still a problem though. It works
fine, but ever since getting it working the bootup process hangs for
three minutes on "configuring network interfaces," or whatever that
line says. I just hit Ctrl-c to continue. But then the wireless is not
functioning after it finishes booting. I have to do ifup wlan0, and
that takes awhile to finish too. I went to a local Linux clinic for
beginners and one of the gurus got it working without the delay. But a
few days later it was back to the delay. Unfortunately, what he did
took him a while to figure out, and I didn't keep notes. (OK, not only
am I a n00b, I'm kinda dumb.)

As for HP v. Compaq, the Compaq has one feature that I dearly love --
the on-off button for the touchpad. I want the touchpad functional, in
case my mouse battery goes dead or I get to the university and
discover I left the mouse at home. But otherwise, it is a PITA. You're
typing along and your thumb hits it and zooom! Your cursor is suddenly
across the screen.

And as for the laptops sans Windows, I tried everything to get a
laptop from HP without Windows before buying the R4000. HP insists
there is no such thing as a laptop without Windows. I tried the
corporate department and every other department. I think they might
have offered a laptop without Windows once, but as far as I know that
option is history.

And for the record, the main problem with the R4000 was the ATI video.
I could never get it out of VESA 1024 x 768. In fact, to date I still
haven't heard anyone getting it successfully into a decent resolution.
In contrast, when I installed 64-bit Ubuntu 5.04 on my R3240 it
automatically found the video and set it up at 1600 x 1080.

 

[Unruh]

It took a long time (because I'm a n00b), but I finally got the wifi
working with ndiswrapper. There's still a problem though. It works
fine, but ever since getting it working the bootup process hangs for
three minutes on "configuring network interfaces," or whatever that
line says. I just hit Ctrl-c to continue. But then the wireless is not

It is the dhcp getting the ip address from the access point. Now I also do
not know why it takes so long, but it usually does.
If you would rather not wait, put
ONBOOT=no
into /etc/sysconfig/network-script/ifcfg-eth1
(assuming your wireless is eth1)

 

[General Schvantzkoph]

Here are some links where they offer notebooks with FreeDOS, although I
don't see how to select FreeDOS instead of Windows.

http://h10010.www1.hp.com/wwpc/us/en/sm/WF04a/321957-64295-89315-321838-f33.html

Here are their Linux certified laptops.
http://h10018.www1.hp.com/wwsolutions/linux/products/clients/clientscert-redhat.html

 

[Bill Kraski]

Here are their Linux certified laptops.

http://h10018.www1.hp.com/wwsolutions/linux/products/clients/clientscert-redhat.html

At least with SUSE & Mandriva, they're a couple of revisions behind. Are
these old pages or is certification that slow?

Courtesy of the Xandros site, here's a link to a vendor who does presently
sell linux certified laptops, usually with your choice of distro (within
which ones they offer) preinstalled:
http://www.linuxcertified.com/xandros-laptop.html

Bill K

 

[Jim]

Hi, aren't Turions all socket 754...???

Thats what I read...

Go-l dot com has/had a /dual/ AMD64 939 laptop in their extreme
performance range. 4GB RAM, 120GB HDx2, a huge battery pack, and a solid
aluminium chassis hinged onto a 20" WXGA panel. Dreamy stuff, if you
want a laptop you could fry half a cow on. I'll take two.

--
Cheers, http://www.dotware.co.uk
Jim http://www.dotware-entertainment.co.uk

Maybe the Internet has the answers I need. It certainly answered my
questions about wang enhancement.

 

[Yadayada]

Go-l dot com has/had a /dual/ AMD64 939 laptop in their extreme
performance range. 4GB RAM, 120GB HDx2, a huge battery pack, and a solid
aluminium chassis hinged onto a 20" WXGA panel. Dreamy stuff, if you want
a laptop you could fry half a cow on. I'll take two.

It's probably still cooler than a P4

 

[Jim]

It's probably still cooler than a P4

they've been having a bit of a rearrange. Just had a look and the
fastest I can find is the Hollywood:

http://www.go-l.com/laptops/hollywood/features/index.htm

This baby has internal space for three hard drives, up to 2GB RAM, P4HT
or EE to 3.4GHz, Radeon Mobility 9700 graphics

Screen size is a bit disappointing though, only 17" widescreen with a
lower resolution than my current powerhouse laptop (Dell Latitude C840
with 2.0 P4m and 15" XGA at 1600x1200 versus the Hollywood with it's 3.4
P4 and 17" WXGA at 1680x1050)

--
Cheers, http://www.dotware.co.uk
Jim http://www.dotware-entertainment.co.uk

Maybe the Internet has the answers I need. It certainly answered my
questions about wang enhancement.

 

[Marek Williams]

It is the dhcp getting the ip address from the access point. Now I also do
not know why it takes so long, but it usually does.
If you would rather not wait, put
ONBOOT=no
into /etc/sysconfig/network-script/ifcfg-eth1
(assuming your wireless is eth1)

Thanks for the suggestion. But before I do it, I suspect it will just
create another problem.

First, the ethernet is eth0 and the wireless is wlan0.

The problem is that at home I have no wireless, but I do have
ethernet. I always need the ethernet because my printers and my
Windows desktop are on it.

Away from home (usually at the university) I have no ethernet, but I
almost always need wireless.

So if I put the ONBOOT line into the network-script for either one,
then that one will not be working. E.g., if I put ONBOOT=no into the
ifcfg-eth1 script, then it should boot fine at the university, but at
home I'll have no ethernet. If I reverse it and put the ONBOOT=no line
into the ifcfg-wlan0 script (I'm assuming such a script exists), then
it will boot fine at home, but at the university the wifi won't be
working.

Note that I made a lot of assumptions above. Did I figure that out
correctly?

 

[Nesmira]

And just to be absolutely clear, only the chips with 128-bit memory
controllers can use dual channel memory and ALL the mobile AMD-64 chips
are listed as being socket 754 and having 64bit memory controllers.


http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_10220_10221^11030,00.html


As for the forum thread you referenced, the posters are just a bunch of
whiny idiots. Since only an idiot would think that a laptop would use
an 89W 939 chip in it. Talk about a hot system, it would have to come
with a nomex pad to keep you from getting 2nd degree burns.[/QUOTE]

I'm one of the whiny idiots and I can tell you have absolutely no clue what you are talking about. The zv6000 and r4000 laptops you referred to have configurations of full blown 939 cpus in them. This has been confirmed. They have 128 bits running. They have everything except the dual channel which was actually not physically there on the board. You're also an idiot because it's not running full bore clock all the time. Laptops have power management to help with heat as well. Before you open your fat mouth, I'd suggest educating yourself before speaking of such matters. Burn....