HP zv6000 (Compaq r4000) BIOS update released… will your pc boot afterwards?

October 30, 2006

HP has recently released a BIOS update for the zv6000/r4000 series of notebook pcs. This update, called F1.C, supposedly resolves an issue where users may get a Blue Screen (BSOD) error when using the USB ports and Cardbus (Express54) slot. I, for one, have never experienced this issue therefore I cannot comment on whether or not the issue is improved upon or resolved.

A side note-

I have recieved more than one email stating that after installing this BIOS update your system may no longer boot into Windows. I have not personally seen this issue. However, I have had difficulty in the past when attempting to update the BIOS using the Winflash utility — to the extent that my system was effectively ‘bricked’ and had to be sent to HP for warranty repair.

Typically, a BIOS flash can cause a PC not to boot into Windows if one of the following settings is reverted to default after being changed… or simply changes: APIC mode, ACPI settings, Power Management (S1/S3/s1&s3), and Cool & Quiet. If you are one of the ones affected by this update check these settings first.

My experience with laptops has been the following. As a general rule, **do not** update your BIOS unless your system is specifically affected by the issue that the new BIOS addresses, *especially* if your notebook is no longer under warranty. I have found through my own experiences that flashing the BIOS on a laptop is much more troublesome (and problematic) than doing so on a desktop. There are several more “technical” reasons for this, but the one I will hammer home here is this one: Most laptops do not have a stand-alone OS disk that does anything less than a complete re-format and fresh install of Windows (plus associated “bloatware”) and will not do a repair install.

I have said my piece and you’ve been warned. Now here’s the link:

http://h10025.www1.hp.com/ewfrf/wc/softwareDownloadIndex?softwareitem=ob-43144-1&lc=en&cc=us&dlc=en&product=453185&os=228&lang=en

Happy Flashing ;)


Blame it on the X300, Part One

October 14, 2006

I have recieved several messages confirming that this problem also occurs on the Xpress 1100m IGPs as well as the 200m IGPs.

A couple of recent comments have specifically referenced the Acer 5102WLMI laptop which employs the newer (and supposedly FIXED) ATI Radeon Xpress 1100m. The 1100m is, for the most part, a rebranded 200m with higher clock frequencies. They are fundamentally the same, both based on the horribly flawed (aka sorry excuse for a GPU) Radeon X300SE. I have my reasons for feeling this way, simply read on.

I recently acquired a Radeon X300SE pciE video card to evaluate so I could get a good baseline as to what to reasonably expect performance-wise out of the 200m/1100m chips. Logically, a Xpress 200m/1100m is a X300 core with half the pipes, reduced clock frequencies (both memory and core) and a reduced memory bandwidth (64bit instead of 128bit, although some X300SE models only come with 64bit) On paper alone, the 200m/1100m should be somewhere between 1/3 to 1/2 of what the X300 is in terms of performance. 

Through my testing unfortunately, the picture has not turned out to be very pretty at all. The X300 all in all, is a horrible GPU. Where it appears to be most flawed is in OpenGL operations. In many benchmarks frame rates drop to as low as 3 frames per second where in comparable DirectX benches the frame rates are at or around 20.

One thing comes to mind here. Is it any real wonder why the X300 card was the only card of its generation to NOT be ported to a FireGL variant?

I am sorry that this has taken so long. When my research is completed I will post real numbers for all to see and evaluate for yourselves.


A BIG thank you goes out to Killersneak of DNA-Drivers for his effort

September 26, 2006

Recently I submitted several posts to people who specialize in “homebrew” ATI video drivers. Specifically, Omega, DNA, and DHZeroPoint.

Killersneak, of DNA drivers was kind enough to respond to my post and lend me some assistance on this issue. He was nice enough to re-release a reworked Catalyst 4.6.6.8 Beta. It has some memory optimizations specifically for the xpress 200m. I have always liked his drivers, and run them on more than one system.

You can access the forum post and the Beta driver link from here:

http://www.dna-drivers.com/forums.php?m=posts&q=1508

After installing his drivers, I did notice a performance substancial improvement, but nothing that can make KOTOR or any of the other OpenGL apps remotely usable. Killersneak did do a fantastic job, however, as performance improvements can be measured to be anywhere from 100-200 percent better in certain tests. Unfortunately for the xpress 200m, a one to two hundred percent increase is only about 1-2 additional frames per second. Given how bad the frame rates are, the performance gains need to be in the 1000s of percent range to make any real difference. I went into this fairly convinced that this is entirely a hardware limitation (aka Video FIRMWARE problem), but I wanted that final nail in the proverbial coffin.

Just to give you a notion of what kind of improvement is needed here, take the 1-2 frames per second performance we get now as a baseline. At 1fps. 100% better is 2fps. 1000% better = 10fps. 2000% better = 20fps = barely playable) On average an optimized driver can improve performance but definately not at the level that is needed here. This proves beyond most measure of doubt, that no amount of software optimizations or tweaks can help this problem. This means if it looks, smells, and walks like a duck… it is, well, a duck. Duck, in this case, meaning a Hardware issue.

Feel free to download the driver, some OpenGL benchmarking programs and give it a whirl for yourself. I have come to the conclusion that this is without a doubt, a video firmware issue. However, even with this amount of proof I doubt HP, or any other hardware manufacturer, will do much about it.


HP zv6000 owners unite

September 3, 2006

There is a website located at the following address. http://www.notebookanalysis.com

It is specifically dedicated to the HP Pavilion dv6000 series and it’s compaq equivalent, the R4000 series. I urge anyone who owns these (or other HP/Compaq models) who suffer this issue to join up. At some point soon a petition will be started. I am currently shopping my technical findings to as “reputable” of a source as I can. I have even debated sending my findings off to Tom’s Hardware to see if maybe they’ll bite on it.


Status Update

August 29, 2006

Here are a few updates as to what has been going on with my attempt at getting the OpenGL/Xpress 200m issue resolved.

1) It is my theory that it is a Video Firmware Issue. I am still awaiting some confirmation as to my findings, but most people I have submitted my test data to are in agreement.

2) HP, for the most part, is ignoring my repeated trouble ticket submissions.

 Here are the driver sets that I’ve used:

ATI Catalyst 5.5 through 6.1, when ATI dropped Xpress 200m support

HP driver for zv6000, old and new — new being 11/2005

Omega Drivers

Zeropoint Drivers

DNA Drivers

Catalyst 6.8 drivers (ATI re-added support for the Xpress 200m with this release)


Research on the Xpress 200m

August 7, 2006

During the time my laptop was on its way to HP for repair I compiled a lot of research on laptops running the ATI Radeon Xpress 200m graphics. Here is some things that I’ve noticed:

There are some newer lines of AMD Turion laptops that are shipping with Xpress 200m graphics that are apparently running KOTOR, Quake, and other OpenGL 1.4 (or above) apps the way they should be. I even tested this theory on a new Compaq Presario floor model at Circuit City. I asked my friend who works there if I could install the Quake 4 Demo on it and run a few tests — after educating him about my problem. I proceeded to do this… and sure enough the Demo ran. Granted this was not great performance… but hell even 15-20 fps is still 14x better than what I’m getting now. I was quite amazed yet frustrated at the same time.

 I began to dig further.

From what I’ve found certain runs of the ATI Xpress 200m chips came with Video Firmwares that are somewhat “bugged” and do not allow for proper OpenGL performance. These chips offload most of the Vertex and T&L operations to the CPU and brings it (no matter its speed) to its knees. For the record, my chip identifier is “0×5955″, with a BIOS version of VER008.025l.004.0000.

I did not note the BIOS identifer on this floor model I tested, but did look around HP’s website for some other information relevant to this issue…

Many of the newer Compaq and HP Laptop models have a revised BIOS downloadable. This BIOS lists in its fixes that it “Resolves a Video Firmware Issue”. Unfortunately, my model of laptop is not supported by this download. The link to it is this: http://h10025.www1.hp.com/ewfrf/wc/softwareDownloadIndex?lc=en&cc=us&dlc=en&os=228&product=1818493&lang=en&softwareitem=ob-39058-1

If anyone can who is having this issue, please send me your BIOS identifiers so I can compile a list of chips that have this problem versus chips that do not. I plan on re-opening the ticket with HP in an attempt to get them to release a new video firmware for the HP zv6000 (mine) and Compaq M2000 (my wife’s) line of laptops.


The Saga Continues

August 7, 2006

Here’s an update:

While working with HP regarding the 200m OpenGL issue they had me do a BIOS update. In any case the laptop locked up during this and made the computer unusable.

 To their credit, I told HP about this and they promptly paid for overnight shipping to them and then back to me. Thing works great now… but still the same issue with OpenGL.


Follow

Get every new post delivered to your Inbox.