Windows 7 32-bit with full 4 GB or 8 GB RAM support


Recommended Posts

The people who keep saying to upgrade to 64 bit and that 32 bit is outdated really need to learn to read.

This thread is about a PROOF OF CONCEPT and NOT about everyday usage of a 32bit OS with >4GB RAM.

What people are saying is that it is POSSIBLE with SOME hardware to use more than 4GB on a 32bit OS. That is all.

The basis for that claim is that many of Microsoft's 32bit Server operating systems support more than 4GB of RAM. It is on Microsoft's own website, this is not debatable. PAE (Physical Address Extension) was designed for exactly this reason.

The feasibility, stability, software support, etc is NOT being argued. This will not work on a lot of hardware, and I'm sure some software will have problems. This is the reason that Microsoft disabled mapping more than 4GB RAM when PAE on the consumer operating systems.

The Sever and Client operating systems now share the same kernel, so if it is possible in the server OS, then it should be possible to enable it in the client counterparts. This patch removes the artificial limitations whether the hardware supports it or not. If not, the system will crash, if it does, more than 4GB of RAM will be detected and used.

Get it? So stop ignoring the whole point of the thread and saying to upgrade to 64 bit. Some people just like to experiment for fun. They just want to see if they can get the 32 consumer Windows OS to use more than 4GB of RAM. Some might use it for everyday usage, but they know it is an unofficial patch and any issues that arise are just their problem.

Edited by mrp04
Wow some people (ShamRocker1) are just so dense!

If I wasn't so lazy I would try this out just to show ShamRocker that he is wrong.

No. I only work in IT. I'm a total idiot. Go troll somewhere else.

The people who keep saying to upgrade to 64 bit and that 32 bit is outdated really need to learn to read.

This thread is about a PROOF OF CONCEPT and NOT about everyday usage of a 32bit OS with >4GB RAM.

What people are saying is that it is POSSIBLE with SOME hardware to use more than 4GB on a 32bit OS. That is all.

The feasibility, stability, software support, etc is NOT being argued. This will not work on a lot of hardware, and I'm sure some software will have problems. This is the reason that Microsoft disabled mapping more than 4GB RAM when PAE on the consumer operating systems.

This patch removed the artificial limitations whether the hardware supports it or not. If not, the system will crash, if it does, more than 4GB of RAM will be detected and used.

Get it? So stop ignoring the whole point of the thread and saying to upgrade to 64 bit. Some people just like to experiment for fun. They just want to see if they can get the 32 consumer Windows OS to use more than 4GB of RAM. Some might use it for everyday usage, but they know it is an unofficial patch and any issues that arise are just their problem.

I rest my case.

No. I only work in IT. I'm a total idiot. Go troll somewhere else.

I rest my case.

What do you mean you rest your case? That is a big piece of text you quoted.

Brandon, who I believe is a Microsoft Employee said it is possible but disabled on the client editions due to compatibility issues with a lot (but not ALL) hardware.

Wow ... who cares. LOL! Doesn't anyone else here think this one is out of hand? Some insults flying!

Obviously some people do, which is why this thread was made. :rolleyes:

I agree with you but this thread was just about it being possible. People do lots of pointless things because they find it interesting. If I had a spare hard drive with nothing on it I would probably try it out just to see if it works. I currently run 64 bit Windows 7 on two machines with 4GB of RAM.

Also the OP apparently has a 32bit machine with 4GB of RAM. He can try it out to see if it works. If he finds it stable, why not use it? He knows full well that it will be in an unsupported state, but that is his decision.

Don't forget that most-if-not-all 32 bit applications are physically limited to 2gb of memory addressing. So unless you have a lot of applications running simultaneously, you'll get no benefit of course.

This doesn't apply to server applications obviously, as they're written to take advantage of PAE.

Don't forget that most-if-not-all 32 bit applications are physically limited to 2gb of memory addressing. So unless you have a lot of applications running simultaneously, you'll get no benefit of course.

This doesn't apply to server applications obviously, as they're written to take advantage of PAE.

I did not address that issue. Thank You for reminding me. :) Anyway. This thread has been rendered pointless I do believe.

I did not address that issue. Thank You for reminding me. :) Anyway. This thread has been rendered pointless I do believe.

I don't know if I have ever had one process use more than 2GB of RAM but having a lot of RAM lets me run many programs at once, each using a good chunk. But anyways, yes, I am done with this thread.

I don't know if I have ever had one process use more than 2GB of RAM but having a lot of RAM lets me run many programs at once, each using a good chunk. But anyways, yes, I am done with this thread.

I've had a few myself, but I handled okay on 32 bit with 3.2gb of ram (limited). Since then went to 64bit for FULL compatiblity anyway. :)

Only if you hack Windows to allow >4GB addressing via PAE and have hardware that supports remapping of device addresses and supports >4GB of addressing. Finding such hardware is rare.

Not so rare. As I wrote I have tested the patch on 4 different PCs. On all these PCs the patch works fine. I would say: If a CPU and motherboard support PAE, then also memory remapping is supported and 4 GB (or more) is supported.

And even if you do find it, you may still have devices which don't play well with this configuration

Have you tested the patch ? Or do you only believe it ?

so your machine may be unstable.

Have you tested the patch? Or do you only believe it ?

Not to mention that hacking your kernel / HAL to allow it to even happen... which itself can destabilze your system

Have you tested the patch? Or do you only believe it ?

cause updates to fail, require re-patching it when updates are made, etc.

This is he only point I haven't tested yet. But how many times Microsoft updates the kernel file "ntkrnlpa.exe" ?

and even netbooks now will run 64bit just fine

Here you are completely wrong.

  1. What Netbook has a 64 bit CPU ? I don't know any.
  2. On most Netbooks "Windows 7 Starter" edition is preinstalled. And Windows 7 Starter edition only exist as 32 bit version.

(But discussing about Netbooks is senseless: What Netbook has 4 GB ore more installed ?)

I don't know if I have ever had one process use more than 2GB of RAM

If I want to run a virtual machine with 2 GB of RAM then I am very fast at this point.

I have 7 Ultimate on my daughter's netbook, but it is 32-bit because the first generation Atom CPUs did not support 64-bit. The new ones do.

Seriously, though, who cares? If you want more than 4 GB of RAM get a 64-bit OS, especially if you're running Windows 7. With the exception of the Atom, nearly every CPU to come out in about the last 7-8 years is 64-bit capable. If you don't have a 64-bit capable CPU then you probably should upgrade your severely outdated system.

russians wanting to patch your kernels ? ahahahahaha all your base are belong to us :shifty:

I have installed the patch and it works fine - even with 8 GB RAM.

I do a virus scan from a Linux live CD with three up-to-date virus scanner: Avira Antivirus, Bitdefender and Kaspersky Anti Virus (also a Russian program !). Up-to-date virus signatures were loaded from Internet. And I don't find any harmful.

  • 4 weeks later...

As I (and others) have stated, this patch is most likely making a hybrid kernel that has the memory support from the server software patched into the kernel for the client OS.

I can't speak for the Russian patch, but I suspect it's another of the many that have automated my directions. On the assumption that I am the original source, let me see if I cannot clear up some confusion. The patch does not create any sort of hybrid kernel. In Windows Vista, the "memory support from the server software" is present already. Indeed, the kernel from Windows Vista SP1 is exactly the kernel from the original Windows Server 2008.

Windows Vista is the first 32-bit client version that has all the code for using memory above 4GB but does not permit you to use memory above 4GB. Windows XP SP2 and SP3 have the constraint but not the code. Earlier versions, back to Windows 2000, have the code but not the constraint.

What the patch does is simulate the provision of valid license data that differs from the ordinary solely by not limiting the use of memory. If Microsoft would provide license data that does not limit memory use, then you would not need the patch. It's that simple. It changes barely a dozen bytes of code.

For some background on the complications which resulted in >4GB of addressing using PAE being disabled for client editions of Windows:

Although these complications are certainly enough for disabling memory above 4GB by default when PAE is enabled, they in no way imply that memory above 4GB must be disabled forcefully for all users. Let's please be clear that "which resulted in" does not mean it was necessitated by technical reasons alone. Microsoft had choices about dealing with the complications. Someone chose not just to disable memory above 4GB in client versions but to do it such that OEMs and users would not be permitted to enable that memory.

But you probably wouldn't connect a 1TB disk to an I/O controller which only support 750GB disks.

But if you discovered that the I/O controller had all the working circuitry to support 1TB but came with a tamper-proofed DIP switch that was set to limit its support to 750GB, you might not be happy about it.

Only if you hack Windows to allow >4GB addressing via PAE

Or if you get from Microsoft some valid license data that differs from the ordinary only by not limiting the maximum physical address. Let's be clear that this is all that is required from the Windows side. Where you talk of hacking Windows, you wouldn't need to change one byte of code except that the license data that sets the limit is tamper-protected.

and have hardware that supports remapping of device addresses and supports >4GB of addressing. Finding such hardware is rare.

Really? I don't know for sure, but I'd have expected that any computer offered for sale in the last 3-4 years with 4GB of RAM and 64-bit Windows pre-installed or available as an option would do this remapping. I can see only two alternatives. One is that all the device memory is above 4GB so that the 4GB of RAM below the 4GB address is all usable to 64-bit Windows. However, I can't see OEMs wanting to configure machines this way any time soon, since it would likely make trouble for installing a 32-bit Windows. The machine might even have to be sold with a warning that it is for 64-bit Windows only. The other alternative is that the machine is sold with a warning that not all 4GB will be usable - even with 64-bit Windows.

Certainly, the last two machines I bought from Dell with 4GB installed have remapping done automatically.

Whether it is rare or not, users can check whether it is presently done on their machine by running the driver from www.geoffchappell.com/studies/windows/km/hal/api/x86bios/fwmemmap.htm.

And even if you do find it, you may still have devices which don't play well with this configuration, so your machine may be unstable.

You may, but the chances must have been nearly nothing for years now. It's really very difficult to construct a bug in a 32-bit driver that wouldn't have to be fixed when porting to 64 bits. Indeed, the only examples I know of are so contrived that I don't believe they would ever appear in a real-world driver. Maybe you know some plausible examples.

There certainly are 32-bit drivers that "don't play well" with memory above 4GB. Some, not Microsoft's, are even shipped on the Windows discs. So, yes, users who have suitable hardware ought not test the use of memory above 4GB until all their (hardware) drivers are up to date and preferably WHQL-certified.

Not to mention that hacking your kernel / HAL to allow it to even happen... which itself can destabilze your system, cause updates to fail, require re-patching it when updates are made, etc.

I myself do not recommend it for ordinary use. I'm actually quite astonished at how many want to, but that's another story. Anyone who's concerned enough to want the everyday use of all 4GB on 32-bit Windows should press Microsoft for a license upgrade.

With that out of the way, I must say you go too far. Even without a license upgrade from Microsoft, they do not have to hack the HAL. (With the license upgrade, they wouldn't have to hack the kernel, either.) Assuming they patch the kernel correctly, it cannot "itself" destabilise the system. Assuming the patch is applied to a copy, as directed, it cannot cause updates to fail. But yes, you are right at the end: the copy would have to be repatched if ever your NTKRNLPA.EXE gets updated.

But if you discovered that the I/O controller had all the working circuitry to support 1TB but came with a tamper-proofed DIP switch that was set to limit its support to 750GB, you might not be happy about it.

'Tis why people buying a PC should do research to see if the machine will meet their immediate and/or future needs. To assume that any one particular piece of hardware will "just work" in any PC, no matter what, is a bit shortsighted.

Although these complications are certainly enough for disabling memory above 4GB by default when PAE is enabled, they in no way imply that memory above 4GB must be disabled forcefully for all users. Let's please be clear that "which resulted in" does not mean it was necessitated by technical reasons alone. Microsoft had choices about dealing with the complications. Someone chose not just to disable memory above 4GB in client versions but to do it such that OEMs and users would not be permitted to enable that memory.

Perhaps because OEMs and users are not qualified to make this decision, as evidenced by this thread.

But if you discovered that the I/O controller had all the working circuitry to support 1TB but came with a tamper-proofed DIP switch that was set to limit its support to 750GB, you might not be happy about it.

Not a fair comparison. A fair comparison is:

There is no DIP switch, the I/O controller just doesn't enable >750GB support because it has not been tested for that case and there are known compatibility / stability concerns. The manufacturer offers a free exchange for a version of the controller which does support >750GB without the caveats, but for some reason you're deciding that you'd rather try to hack the thing to do something unsupported.

Or if you get from Microsoft some valid license data that differs from the ordinary only by not limiting the maximum physical address. Let's be clear that this is all that is required from the Windows side. Where you talk of hacking Windows, you wouldn't need to change one byte of code except that the license data that sets the limit is tamper-protected.

The "license data" is part of the product code. Changing that is unsupported and violates your license agreement. The effects of doing so (for instance, on your ability to install future updates) are hard to predict.

Really?

Really.

I don't know for sure, but I'd have expected that any computer offered for sale in the last 3-4 years with 4GB of RAM and 64-bit Windows pre-installed or available as an option would do this remapping. I can see only two alternatives. One is that all the device memory is above 4GB so that the 4GB of RAM below the 4GB address is all usable to 64-bit Windows. However, I can't see OEMs wanting to configure machines this way any time soon, since it would likely make trouble for installing a 32-bit Windows. The machine might even have to be sold with a warning that it is for 64-bit Windows only. The other alternative is that the machine is sold with a warning that not all 4GB will be usable - even with 64-bit Windows.

Computers in the last 3-4 years with 64-bit OSes and 4+ GB of memory are rare. That set is probably okay in this regard, but since they already have a 64-bit OS, the point is moot. This discussion is about machines which came with 32-bit OSes.

You may, but the chances must have been nearly nothing for years now. It's really very difficult to construct a bug in a 32-bit driver that wouldn't have to be fixed when porting to 64 bits. Indeed, the only examples I know of are so contrived that I don't believe they would ever appear in a real-world driver. Maybe you know some plausible examples.

Nvidia and ATI video drivers, for example.

I myself do not recommend it for ordinary use. I'm actually quite astonished at how many want to, but that's another story. Anyone who's concerned enough to want the everyday use of all 4GB on 32-bit Windows should press Microsoft for a license upgrade.

Why an upgrade? Your license is already valid for both the 32-bit and 64-bit versions of Windows Vista or Windows 7. If you want a 64-bit (>4GB of memory) system, then use the 64-bit version. No need to "press Microsoft."

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • Most boring game ever. Repetitive, empty, predictable, and full of cliches. Total waste of time and money, IMO.
    • Mafia: The Old Country expansion Man of Honor announced, brings back Salieri from original by Pulasthi Ariyasinghe During Summer Game Fest, 2K and Hanger 13 brought out a new Mafia: The Old Country trailer, revealing the game's first expansion. Named Man of Honor, this is slated to add two new chapters to the Enzo storyline that the game follows. There is an iconic character returning to the series with this expansion, with players set to run into Ennio Salieri, the future Don of the Salieri crime family. Fans of the original Mafia, or its Definitive Edition remake, may remember that name as one of the biggest characters in the storyline. This expansion is set prior to his rise to being the kingpin in the City of Lost Heaven. "Set in Sicily during the winter of 1905, Enzo Favara has proven himself a reliable soldier of the Torrisi crime family in the months since his initiation," says the studio about the new chapters. "Now, the Don entrusts him and Cesare with a delicate assignment of assisting Ennio Salieri, a man of honor recently released from prison and intent on reclaiming what is his." Working at Salieri's side, players will be heading into fresh environments as they return to the role of Enzo as a high-ranking soldato. The studio also promises brand-new weapons, fresh vehicles, and charms to collect in this expansion. Moreover, the expansion will add new content to the updated Free Ride mode. Alongside new collectibles and locations, this will add more challenges to beat alongside Salieri, which are described as runs that will "test the skills of even the most elite mafiosi." The Mafia: The Old Country Man of Honor story expansion will release on August 14, 2026, across PC, Xbox Series X|S, and PlayStation 5. It will cost $10 for owners of the base game to jump into.
    • Try installing Logitech Options
    • Telltale returns with The Wolf Among Us, re-revealing the sequel and a new remaster by Pulasthi Ariyasinghe Telltale Games, the studio known for its episodic games from the 2000s, had a surprise appearance at the Summer Game Fest showcase today. The studio has gone through a collapse, an acquisition, and a revival but has largely gone silent in recent years. Today's sudden announcement was for the popular The Wolf Among Us series, with news about the sequel and a remaster dropping. Set after the events of the 2013-released title, the sequel will bring back Biby Wolf for a new adventure in Fabletown. This town of fairy tale characters is becoming unstable, with Snow White not being able to keep order as before. "When a brutal series of crimes threatens to fracture the fragile balance holding Fabletown together, Bigby Wolf, sheriff of Fabletown, finds himself drawn into a dangerous conspiracy that reaches deep into the city’s criminal underworld," says the studio. "As suspicion spreads across the city and powerful forces work against him, Bigby must uncover the truth and decide what kind of sheriff - and what kind of wolf - he is willing to become." At the same time, Telltale also announced The Wolf Among Us Remastered, bringing back the classic with a fresh coat of paint. This will include all five episodes from the original game plus over an hour of extra content that will offer behind-the-scenes videos, cast interviews, image galleries, and deleted scenes. The remaster will also have updated visuals and improved UI elements and accessibility features, alongside upgrades to the animations, audio, and frame rates. Telltale has been working on this project for almost a decade now, but if the new release schedule holds, fans will be able to jump into The Wolf Among Us 2 sometime in 2027. As for The Wolf Among Us Remastered, this is slated to release during this holiday season.
    • I use a Linux laptop every day. I have tried to switch to Linux on my gaming PC which has a 5070 ti. It was not a great experience
  • Recent Achievements

    • Rookie
      moog19 went up a rank
      Rookie
    • Mentor
      grik went up a rank
      Mentor
    • Dedicated
      JKR earned a badge
      Dedicated
    • One Year In
      CHUNWEI earned a badge
      One Year In
    • Conversation Starter
      FBSPL earned a badge
      Conversation Starter
  • Popular Contributors

    1. 1
      +primortal
      486
    2. 2
      PsYcHoKiLLa
      270
    3. 3
      Skyfrog
      77
    4. 4
      Steven P.
      68
    5. 5
      FloatingFatMan
      62
  • Tell a friend

    Love Neowin? Tell a friend!