Albuquerque

"to get the processors full potential" - by this you mean the absolute maximum where you're stressing the device beyond it's designed limits. Good luck getting a useful lifetime out of that piece of hardware. I have overclocked and over-volted essentially every CPU and GPU that I have purchased with my own money. And you know what? I've never broken a piece of hardware in it's overvolted and overclocked state either. 486DX33 -> 50mhz with some additional soldering for ~10% voltage bump and a jumper for ~50% more FSB. That chip cost me $380 and it lived for three years. 486DX4/133 -> 200mhz using same exact setup as above. AMD made this processor, it was AWESOME and performed phenominally. I had this box for almost five years with ZERO issues. Used the stock heatsink with a bigger fan. SIS 864p VLB video card. Overvolted ~5% by some soldering, overclocked by ~35% via software. Had this video card on the above-mentioned 486 system all eight years without issues. No heatsink either Pentium 66 -> 100. Overvolted by 15% and overclocked by 50% (multiplier). Only had this for about a year, so who knows? Used the stock HS. Pentium 166 -> 250 Overvolted by probably 20% and overclocked by ~15% FSB and multiplier. Worked awesome for about six years, used only a bigger fan and some ducting on top of the stock HS. Voodoo Original 50mhz -> 72mhz with about a 20% increase in voltage, a big heatsink from a Pentium 133 box and some software. I still have this card somewhere, ran on several different boxes in this config without ever having a problem. NVIDIA Riva 128 original 4mb -> Actually, I don't think I voltmodded this, just an overclock. Meh. *insert long pause* Celeron 350A -> 504. More voltage, more FSB. Stock HSF, and lots of fun! I still have this box running a Telnet server... NVIDIA GF3/Ti200. About 15% more voltage gave me Ti500 speeds. Stock HSF too... P3/550 -> 684. Lots more voltage and more FSB. Stock HSF but bigger fan. Ran really well in a dual-processor config. My dad still has this box as his 24/7 rig, it's quite old too, has the above-mentioned GF3 in it. P4/1.8A (Northwood) -> 2.7ghz. This was my first really awesome overclock since the Celly 350A. Used 1.85v to get there, but who cared? Sold it to a friend and it's still his primary box. ATI 9500np softmodded to all 8 pipes and voltmodded + OC'd to 440mhz. I eventually killed this, but it was due to me being a dumbass and shorting it out with a huge steel washer that was holding the heatsink down. Still lived for better than two years. ATI 9800np OV'd and OC'd to ~465mhz. This one cooked too because the aftermarket fan I strapped to it suddenly stopped one day. D'oh ATI x800pro Vivo -> XTPE Vivo at 555/1166 with a pencil mod and some water. Ran awesome for three years, sold it to a guy at work, it's still his primary vid card and he loves it. P4 2.4C -> 3.4ghz. OV'd to 1.65 with a big-a** copper heatsink. Loved it! This box is now my mom's... P4 3.0 Prescott -> 4.2Ghz. Overvolted to 1.525 and watercooled. This is my current rig, it's been running for about a year. I love this rig... NVIDIA 7900GT-on-AGP (aka the Gainward 7800GS+ 512mb). Actually I haven't overvolted this yet, but mostly because it isn't the bottleneck still Overclocked to 540/1525 tho...

Maybe you should tell us what you've done to use 13GB of storage for Vista? Because a brand new 100% default install of Ultimate takes ~4gb. That's with all the media, backgrounds, fonts, language MUI's and other nonsense that it comes with. Err have you tried that yourself? I find it very hard to believe as my Ultimate install with everything cut out by Vlite takes ~4 GB without the pagefile or hibernation file. A quick Google shows that other people with Ultimate are reporting sizes of 8->12 GB. As I have eight licenses for four different versions, I must answer yes, I have tried it myself. Edit: I feel the need to post something more relevant to the original thread, so here goes: I like Windows, I like *nix. I have four machines at home, and all of them either solely boot to a Windows OS or boot to Windows as the default in a dual-boot environment. My job function is entirely based on Windows OS deployment and SMS tools, and as such I use *nix simply out of my own curiosity and only at home. I'll say this: The "geek" side of me loves Linux of the various flavors. I have more experience with Solaris myself, but I've also played around with RedHat, Suse 10 and FreeBSD. But the learning curve is quite steep; I've been neck-eedp in low-level computer functionality since doing assembly programming on an Apple ][e. Stepping into *Nix is difficult and often confusing, and even more so when you start playing with the multitudes of "flavors" that are available. The "get stuff done", "play games" and "general day to day tinker" sides of me likes Windows and will probably continue for a long time. Vista adds a lot more "geek" functionality in my opinion, specifically with the advent of powershell (available on XP, granted) and a ton of other stuff under the hood like performance event logs and the like. Further, while I could certainly build a *nix desktop for my mom to use 1000 miles away, I wouldn't. The learning curve for her would be too severe, even though she has no need or use to get technical with it. She knows where Windows puts stuff, and she understands "My Documents" and she uses plug-n-play functionality with her various devices (printers, flash drives, MP3 players, cameras, scanners). That kind of work would open up a level of confusion for her that would only result in phone calls to ME, and that would just irritate me. Same goes for my GF. Maybe that relegates Windows to the computer n00bs and idiots of the world? Maybe so. But there's a lot more n00bs and idiots out there, which means so long as that's the case, Windows will still prevail. Now had this discussion been about Macs, I might have a different end-result...

Maybe you should tell us what you've done to use 13GB of storage for Vista? Because a brand new 100% default install of Ultimate takes ~4gb. That's with all the media, backgrounds, fonts, language MUI's and other nonsense that it comes with.

Something to keep in mind... All the new IO subsystems in Vista are far-better scheduled so that an app can't steal 100% utilization of an individual device -- which includes your hard drive. This can also make things process slower, especially such as defrag. But there's also an upside: Microsoft did this so your machine doesn't crawl while things like Indexing, Defragging and Virus Scanning are taking place. As such, you should be able to do nearly anything else you want while defrag is running in the background, and at the same time notice almost zero performance degredation. Not completely zero of course but nearly... Also, you shouldn't have to force the defrag -- it will begin auto-defragging everything in the background as you're doing other tasks. You'll notice HD activity occasionally, and if you go look at the new Resource Monitor, you'll sometimes see Defrag running. My first instinct was to kill it, but in the new way of scheduling processes and device usage with Vista, you truly don't notice outside of the flashing disk light and the noise of your heads seeking.

Anything that requires hardware video acceleration will not work. To include 3dflip, aero glass, etc.

I too get around this problem, but I use SysLinux (or IsoLinux for when it's on CDROM) Same general thought process as Grub4Dos, but tastes vary

I just wanted to come in and make sure to clarify: I aboslutely love IBM and would recommend them to anyone. But if the original poster really does want the "UBER" bleeding edge tech, that isn't something that IBM does. And just as a previous poster mentioned, that's not a negative thing for IBM's laptops at all. They are indeed built to be very reliable and in doing so they don't always use the newest and fastest technology. And as others have mentioned multiple times, building your own is probably the best thing you could do. It's a great learning experience, and there's tons of people on here who would be willing to help you out on the way. You could build exactly what you want, getting only the features you need and not having to pay for the crap that you don't.

My answer would have been to put in the BOOTFIX.BIN too; maybe it's some sort of limitation to the USB stick? Odd... I'd wager that if you built it using the ISO Ramdisk method you could get it to work, but I know you don't want to go down that road. Not to ask a dumb question, but have you tried googling something like "USB PE Bootfix.bin" or thereabouts? I'd wager you're not the first person who needs something like this...

Maybe so, but Dell's suck. Overheating, the laptops have that hidden hardware keylogger in the keyboard connector, they configure the OSes to be able to remotely control the customer's PC at any time for 'customer service'. Dells are for people born yesterday, plain and simple. If you have ever looked inside your PC, you are above Dell. Custom Build all the way. I beg to differ. I'm the owner of a Dell laptop, and have never had any issues with it. It has been every bit as reliable as my IBM Thinkpad R50, although with the inferior casing that they use I'm also a bit more gentle with where I stash it. Also, if you're foolish enough to believe the Dell hardware keylogger and/or somehow not even investigate the rumor on your own, then there's quite a bit that could be said about your "birthdate" as well. Let me give you a hint: four letters, starting with "H" and ending in "X". Can you guess? I'll give you a better hint, type "Dell keylogger hardware" in to Google and let me know what the first eleventy brazillion hits are... Second, if you somehow believe that the OS is configured to be remotely controllable at any moment by someone, then "born yesterday" is a severe understatement for your obvious technical understanding of the situation. Is the service even running? (hint: no) Is the Windows firewall allowing that traffic through? (hint: no) Would they have any clue about which machine went to which person, and when that machine is online, and what the name of that machine is for name resolution or what that machines IP address is now that it's out in the wild? (hint: no, no, no, and no) There's nothing wrong with Dell, and that's coming from someone whose second language was assembly at nine years of age. Sure, there are "better" machines you can buy for various reasons and levels, but the generalization you're applying is grotesquely innacurate.

Maybe you shouldn't be so quick to post such a trite response? If you can pull your head out of the sand momentarily, I bet you might be able to realize that I just answered your question: has anyone experienced this and have the solution? The answers are: YES, others have experienced it, posted about it and discussed it in the very recent and distant past. NO, there has never been an answer found here, on the 911 forums, in any other accessible fashion by way of Google's search engine, and even on Intel's or Microsoft's premier knowledgebase. The real kicker? It doesn't even sound like Intel knows about the issue, which is why they probably haven't fixed it. So someone who is affected by this problem needs to let Intel and/or Microsoft know about it. And that person needs to have access likely to an Enterprise SA with Microsoft or some similar support contract with Intel in order to not be charged for it. If you can't get that through your grubby little head, too bad. My s*** works, and I'm using the driver AND the Intel nic card. I should pull a "steelhead" and just say that it sucks to be you and you just need to get over it and do it a different way. Good riddance.

This question seems to come up for every single person I've ever heard of who PXE-boots a machine with an Intel Pro/1000 nic. The Pro/1000 drivers work exactly as advertised in a boot-from-anything-but-PXE scenario, but the first time you boot from PXE it goes to hell. I'd suggest calling Intel if that's an option in your environment. Myself and countless others are using the same Intel ProSet 10.3 and 10.4 drivers to run our WinPE implementations without issue -- but those of us who have it working are also not using it for PXE booting.

You say "most powerful" with biggest harddrives and ram. An IBM Thinkpad, as much as I love both of mine, will not do that. IBM (now Lenovo) does not place their laptops on the bleeding edge -- you aren't going to find the biggest fastest hard drives, fastest video cards and highest-end processors in an IBM laptop. It's a shame too, because they're arguably the best on the market -- maybe that's why? If you want the biggest and fastest of everything and money is no object, then I too would suggest Dell. Alienware was just purchased by Dell, so all the uber-high-end Desktop rigs with the best dual core processors and quad-SLI setups will be available through them. Want a laptop instead? There are some options, but if you just want bragging rights about having the biggest and best, go get the new Dell "luggable" $4000 model that has the integrated 20" LCD panel, eight-speaker audio with subwoofer, detachable bluetooth keyboard and mouse and the works. It's insane. However, the real answer for building the best-of-the-best desktop is right above you: Build your own. You can buy all kinds of things over the counter, but it will still pale in comparison to what someone could do with their own parts. Just like you can buy a 400HP crate motor over the counter, but it will pale in comparison to what's possible when someone purchases the parts specific to what they need and hand-builds a 1100HP monster. I've only bought two fully-assembled computers in my life -- both were laptops. In a desktop environment, there's no reason not to build your own. And if you think you can't? The first computer I fully assembled myself was a 286-16 with math-co-processor back when I was 13. Age has nothing to do with it -- intent and patience does. It took a while, but it came out just fine.

Indeed. Funnily enough, the x1300 had no problems running Aero Glass at all, but my current 7300Go seems really choppy and "hitchy" with it -- even though the 7300 has more physical memory, more memory bandwidth and operates at a higher operational capacity (clockspeed * ROPs * ALU's). Drivers can be a big impediment...

I'm not sure why 400mbit isn't fast enough... Can someone elaborate on that? That's an effective transfer rate of 50 megabytes per second, and while it's possible that a 7200RPM drive can exceed that transfer capacity in rare cases, it isn't anything typical and will not impede your performance unless you're streaming a single very large file from the very outer edges of the drive track. Both serial and parallel ATA standards can theoretically transfer quite a bit more data than 50 megabytes per second, the actual problem is your physical drive cannot. Spindle speed and data density on the platters does not even saturate the current "old" ATA-133 mark right now, even on the uber-highest-end SATA drives. Most consumer-grade drives (with exception to the Raptors) will hit between 50 and 65mb/min of sustainable sequential-pattern read speed at an absolute maximum limit. Unfortunately, almost all of your boot process and application process has a considerable amount of random-seek pattern to it, you never actually hit that mark in any real world environment outside of a carefully crafted benchmark scenario. Cliff notes? USB2 is as fast as you need for a standard bootable 7200RPM hard drive of *any* size. The 134mb limitation comes from a firmware limitation in the drive enclosure you choose, so don't get an old or ungodly cheap one -- make sure somewhere it mentions support for 48-bit LBA and you're fine.

If you don't want to get that crazy, you can include a small freeware utility named "hider.exe" and have it run as the very first line in your batch script... Usage: Hider.exe hide command prompt Later in the script when you need it to be visible: Usage: hider.exe show command prompt I am not the writer of the software nor does it gain me anything if you use it... Linkage: http://www.humus.ca/index.php?page=Cool&ID=2

I had problems on a mobility x1300 that *if* the little Windows Experience Index thing didn't measure at least a 2.0 on the Graphics mark then Aero Glass would be disabled. Turned out I needed an updated driver...

D'oh -- seems I have my terminology mixed up. Disregard

Agreed. NLite is perfect for those advanced home users who know exactly what they want (and don't want) from their OS. In a corporate environment where you run the gammut of clients from those who can barely operate their DVD player to those who develop the newest point-of-sale equipment, you can't use a product like NLite to build an OS (and certainly not distribute it in a controlled fashion) There are much larger things to worry about in a corporate deployment, such as retaining client data, client applications and preferences, full OS functionality across all hardware platforms, etc. You don't want to start "lightening" the OS for fear of someone needing a function that you somehow second-guessed that nobody would use. BDD is the next iteration of the half-decent SMS OS Deplyment Kit. The functions available are quite good and give you a significant amount of flexability in installation options and client data management. In an active directory environment, there are enough features and options to manage a wide-scale deployment down to the individual check mark ticks on a specific model of laptop belonging to a specific user.

Well, not to be a complete and utter buttcheese, but BDD is how you do it. It does work, more than a few of us here on the forum are actively using it now for test-deployments on our respective test environments. What specific part are you having problems with? What pieces have you sucessfully accomplished on the road to building your deployment via BDD?

New and old; I don't like either one of them. I also don't appreciate all the IQ shortcuts that NVIDIA selects "on your behalf" that you have to manually go through and disable... Trilinear optimization, mipmap optimization, anisotropic filtering optimization, texture LOD optimizations, texel clamp optimizations. And when you're done turning all the optimizations off, you still get texture aliasing that you have to manuall jack with the LOD to half-assedly correct it. I like the speed, and right now even with the 1950AGP out on the market, the 1500-production run of the AGP-connected 7900GT made by Gainward is still the fastest AGP card on the market. And again, I do like the speed, but if I were purchasing a card today, I'd be right back at ATI again. Hopefully when NVIDIA's G80 comes out in a few more months, they'll have corrected the IQ shortcuts that they initially caused in ther FX line so long ago. And maybe the extra time will give them an opportunity to continue working on the driver interface. I'd still rather have the ATI CCC then NVIDIA's new offering. Take note that both company's "new" offerings are the only ones being supported in Vista, meaning relying on the prior control panels is relying on obsolesence. ATI's had much more time refining their new control panel methodology than NVIDIA has, and it blatently shows in my opinion.

I'm exactly the opposite of the previous two posters... I've owned an ATI Radeon since the softmoddable 128mb 9500 non-pro's several years back; all my previous cards were NVIDIA. Originally I wasn't sure what to think of ATI's drivers, but even back then I found them more than competent. And I upgraded from a 9500np to a 9800XT 256mb, then an x800xt vivo and continued to love it. Had no problems whatsoever with the drivers, performance or stability, and had no problems with the graphics quality or performance either. I'm now an owner of one of the ultra-rare only-1500-produced 7900GT-on-AGP cards from overclockers.co.uk. Full 24 pipes, full 7900GT speeds, and 512mb of ram. It performs considerably faster than my x800, but I'm not happy with the driver interface at all. I'm also not happy with the texture quality issues that I see in many games. I still like having the extra speed of the 7900 GT to be sure. But given a preference, I'd rather have one of the newly released 1950Pro AGP cards instead and take the rather insignificant performance hit.

Not a good idea. Windows periodically reads/writes to the drive even when idle, so that could create a conflict. The best image is still created from a disk that is otherwise not in use, e.g. from DOS or live CD boot. Probably finally using the "Volume ShadowCopy Service" from XP / 2003 that I never saw fully utilized earlier. Allows you to make background copies of files even if they're still in use, even if they're still open for editing for that matter. Good to see someone putting it to use in an interesting way.

True DX8 class hardware (ATI R200, NV20, Matrox G300) doesn't support aero glass features.

Looks like a perfect config for a headless box. Good job Often I find myself looking at someone's hardware list for such a device and wondering what they think they'll get with such overabundance of ram, processor capacity, etc. The last time I had a headless device was for a NAT, firewall, DHCP, DNS cache, FTP and Telnet box with some file sharing functionality. I used an HP Vectra XA/166 -- Pentium original 166mmx underclocked to 90mhz (the 166mhz chip ran on less voltage than the 90mhz chip did, ha!), 96mb of ram, a pair of 3C905b NICs and an old Adaptec 1540 PCI SCSI card with an outboard storage cube. It came with an 85W power supply, and was about the size of a VCR give or take. I had uptime in excess of 400 days on that dinky little box loaded with WinXP and the standard firewall software. Actually the OS would have help up for quite a bit longer, but my software RAID-0 stripe set died when one of the SCSI drives in the external storage thing puked. I had to reboot the box to reconfig it all. Using a very low power (and not great performing) processor is perfect for these kinds of tasks. Saves money too

If you're not smart enough to test your pelt before attaching it, you shouldn't be using one And the only way to really get full use of a pelt is to watercool it (IMO), and again, my opinion of testing your watercooling before use is the same as above