Looking for DS Fat Firmware variants (V1, V2, V3) for my preservation project.

reimolaev · Apr 16, 2026

Hi everyone! I’m working on a personal collection/archive of Nintendo DS firmwares. I’ve managed to find most Lite versions and even some prototypes (040615), but I’m missing the early DS Fat revisions (V1, V2, V3) and the elusive iQue Fat dump.

If anyone has these old consoles and is willing to help with a dump (using DSFWDumper), or knows where these specific versions are archived, please let me know. I'm looking for clean, unpatched (no FlashMe) files. Thanks in advance!

rvtr · Apr 17, 2026

reimolaev said:
Hi everyone! I’m working on a personal collection/archive of Nintendo DS firmwares. I’ve managed to find most Lite versions and even some prototypes (040615), but I’m missing the early DS Fat revisions (V1, V2, V3) and the elusive iQue Fat dump.

If anyone has these old consoles and is willing to help with a dump (using DSFWDumper), or knows where these specific versions are archived, please let me know. I'm looking for clean, unpatched (no FlashMe) files. Thanks in advance!

Already preserved, but I'd suggest mirroring my archive. It's just this single web page and all the dumps.
https://docs.randommeaninglesscharacters.com/dsfirmwares.html

reimolaev · Apr 18, 2026

rvtr said:
Already preserved, but I'd suggest mirroring my archive. It's just this single web page and all the dumps.
https://docs.randommeaninglesscharacters.com/dsfirmwares.html

Thanks! Now all I have to do is find v6 and v7...

rvtr · Apr 19, 2026

reimolaev said:
Thanks! Now all I have to do is find v6 and v7...

v6 and v7 are flashme versions, not official builds. See the bottom section of the page for those.

reimolaev · Apr 19, 2026

rvtr said:
v6 and v7 are flashme versions, not official builds. See the bottom section of the page for those.
View attachment 568850

Okay. It's just that all sources (for example, the ConsoleMods Wiki) say there are only seven official firmware versions.

rvtr · Apr 20, 2026

reimolaev said:
Okay. It's just that all sources (for example, the ConsoleMods Wiki) say there are only seven official firmware versions.

There's a lot of incorrect information out there about firmware versions. I've purchased and dumped hundreds of wifi boards, so you can trust what I say. I've never seen anything newer than 2006/03/08 11:19. Small bit of my collection

Wokann · Apr 20, 2026

reimolaev said:
Okay. It's just that all sources (for example, the ConsoleMods Wiki) say there are only seven official firmware versions.

yes.
In retail versions of ds (not factory or debug version), there are 5 version for international ds phat, 2 version for international ds lite, 1 verison for chinese iQue ds phat, 1version for chinese iQue ds lite, 1 version for korean ds lite.
In community, we call the 5 international ds phat firmware as v1~v5, while nintendo called v4~v5 as Ver4.0~Ver5.0 with NTR tag (region named WORLD WIDE), which i discovered it from leaked offical F-writer. So v1~v3 has a high possiblilty to be named as Ver1.0~Ver3.0 by nintendo.
We call the 2 international ds lite firmware as v6~v7, while nintendo called v6~v7 as Ver2.0~Ver3.0 with USG tag (region named WORLD WIDE).
USG-WORLD WIDE-Ver1.0 hasn't been dumped, and not appeared in retail ds lite. A member in discord claimed her friend has owned this debuger/factory console with USG-WORLD WIDE-Ver1.0 and shown the screenshot for it. So it does exist, but not for retail sale.
iQue and Kor do not have a well-known community names. nintendo call iQue phat firmware as NTR Ver1.0 (region named CHN [iQue]), iQue lite firmware as USG Ver1.0C (region named CHN [iQue]), Kor lite firmware as USG Ver1.0K (region named KOR).

So in conclude, for retail ds phat and ds lite console, the info of firmwares are:
WORLD WIDE: 5 NTR (Ver1.0~Ver5.0) + 2 USG (Ver2.0~Ver3.0) (community call v1~v5,v6~v7)
CHN [iQue]: 1 NTR (Ver1.0) + 1 USG (Ver1.0C) (i named them iQue_v1, iQue_v2)
KOR: 1 USG (Ver1.0K) (i named it Kor_v1)
10 retail firmware in total.

v1: 2004-10-05 11:07———————-NTR—WORLD WIDE—Ver1.0 (official ver name guess)
v2: 2004-11-26 09:51———————-NTR—WORLD WIDE—Ver2.0 (official ver name guess)
v3: 2005-02-28 08:51———————-NTR—WORLD WIDE—Ver3.0 (official ver name guess)
v4: 2005-06-06 14:48———————-NTR—WORLD WIDE—Ver4.0 (confirmed)
v5: 2005-12-07 14:45———————-NTR—WORLD WIDE—Ver5.0 (confirmed)
v6: 2006-02-05 21:33 (Lite)————--USG—WORLD WIDE—Ver2.0 (confirmed)
v7: 2006-03-08 11:19 (Lite)————--USG—WORLD WIDE—Ver3.0 (confirmed)
iQue_v1: 2005-06-09 21:15————--NTR—CHN [iQue]—Ver1.0 (confirmed)
iQue_v2: 2006-04-26 15:35 (Lite)——USG—CHN [iQue]—Ver1.0C (confirmed)
Kor_v1: 2006-11-09 21:30 (Lite)——--USG—KOR—Ver1.0K (confirmed)

As for Flashme, the version name only shows the version changes of flashme itself, not combine with offical version.
Before flashme v7, flashme only has custom hack for ds phat. From v7~v8a, flashme has both verions for ds phat and ds lite.
Flashme for ds phat are made based on official NTR-WORLD WIDE-Ver3.0 (community call v3). Flashme for ds lite are made based on official USG-WORLD WIDE-Ver2.0 (community call v6).
So if you see flashme v7, it doesn't mean it was made from community call v7 (official call USG-WORLD WIDE-Ver3.0), it was made from v3 and v6 for ds phat and ds lite.

I've collected and organized these firmware files, but gbatemp probably isn't suitable for directly uploading copyrighted content. Myrient used to store these firmware files but has since shut them down. However, there's an easy way to obtain the international versions v1-v7: the latest noflashme bundle for flashme, which contains the v1-v7 firmware data to restore devices flashed with flashme back to the original firmware. You can get v1～v7 firmware in noflashme.

Added: The version numbers (v3, v5, etc.) listed in the link above are incorrect. There's a method to determine the official version by observing the color of the screen when ds console crashes (v5 is Magenta) after removing the cartridge in the Picochat chat room. However, later versions like v6 and v7 are also Magenta and they're all incorrectly listed as v5. You can check the correct version using the v* and timestamp mapping I provided. I've also listed a more detailed table on the TCRF website, which you can check out.

reimolaev · Apr 20, 2026

Thanks! I didn't even realize it was so confusing

Wokann · Apr 20, 2026

reimolaev said:
Thanks! I didn't even realize it was so confusing

In short, the link provided by rvtr has already contained the 10 retail firmware versions you want to collect.

rvtr said:
Already preserved, but I'd suggest mirroring my archive. It's just this single web page and all the dumps.
https://docs.randommeaninglesscharacters.com/dsfirmwares.html

Althrough the version names aren't perfectly accurate, the firmware dates are correct, so you just need to match the dates to find them.

rvtr said:
I've never seen anything newer than 2006/03/08 11:19.

The version mentioned by rvtr is the last retail version of WORLD WIDE, USG Ver3.0 (community call v7 firmware)

rvtr · Apr 20, 2026

Wokann said:
Althrough the version names aren't perfectly accurate, the firmware dates are correct, so you just need to match the dates to find them.

The versions were based off of older gbatemp posts. I put the blame on those. The community had no standards even with their custom versions. I hate is so much. v7 could mean iQue, or it could mean flashme, or it could mean the final USG firmware, or it meant nothing at all. Like these tests:

Firmware v6: Dark blue screen will appear.
Firmware v7: Nothing happens, Pictochat functions as normal.

Dark blue is the community's v2, and nothing happening just doesn't happen (except on flashme iirc?), so a community v7 doesn't officially exist.

There's a reason I almost exclusively use build dates.
https://gbatemp.net/blogs/unreasonably-angry-rant-about-ds-firmware-versions.19539/

Wokann · Apr 20, 2026

rvtr said:
The versions were based off of older gbatemp posts. I put the blame on those. The community had no standards even with their custom versions. I hate is so much. v7 could mean iQue, or it could mean flashme, or it could mean the final USG firmware, or it meant nothing at all. Like these tests:

Dark blue is the community's v2, and nothing happening doesn't happen, so this test for v7 doesn't exist.

There's a reason I almost exclusively use build dates.
https://gbatemp.net/blogs/unreasonably-angry-rant-about-ds-firmware-versions.19539/

Yes, there was a lot of confusion and misunderstanding about firmware in the community in the early days. I only found various problems with a lot of the information after I actually tested it.
After repeated data testing, examining the FlashMe source code, and analyzing F-writer's internal data, I finally pieced together the information I mentioned above.

Firmware v6: Dark blue screen will appear.
Firmware v7: Nothing happens, Pictochat functions as normal.

As you've tested, both of these claims are incorrect, I've also tested these claims..
For v6, actual testing and comparison of the firmware's internal color hex data showed the correct color is still Magenta, the same as v5.
For v7, my tests confirmed that neither the original v7 nor FlashMe_v7 exhibited this behavior. The original v7, like v5 and v6, uses Magenta, and its internal color hex data is identical. The Flashme v7, as mentioned above, FlashMe Phat is based on v3, and FlashMe Lite is based on v6; they will display green (same as v3) and Magenta (same as v6) colors respectively.

Then what is the standard of version naming from early community?

The corresponding data can be found in the FlashMe source code.

C:

#define NDS_VER1 0x2c7a
#define NDS_VER2 0xe0ce
#define NDS_VER3 0xbfba
#define NDS_VER4 0xdfc7
#define NDS_VER5 0x73b3
#define NDS_VER6 0xe843
#define NDS_VER7 0x0f1f

These u16 data points correspond to CRC16 of part12 located at firmware addresses 0x06-0x07. By comparing these, the specific timestamp version of the firmware can be determined. Furthermore, the noflashme source code directly contains firmware content for versions v1-v7, further verifying the correspondence between community version naming and specific timestamp firmware.
Then the final conclusion for the community version naming to official interantional firmwares is what I said above.

v1: 2004-10-05 11:07
v2: 2004-11-26 09:51
v3: 2005-02-28 08:51
v4: 2005-06-06 14:48
v5: 2005-12-07 14:45
v6: 2006-02-05 21:33 (Lite)
v7: 2006-03-08 11:19 (Lite)

As for the official naming convention like USG-WORLD WIDE-Ver2.0, USG-CHN [iQue]-Ver1.0C, that's data I obtained through research and analysis from F-writer.

In the early days, the community didn't pay much attention to iQue or Kor firmware versions, so the rule of using colors to determine firmware versions was limited to international versions. Therefore, we can't say that the green of an iQue Phat represents v3, or that the Magenta of a Kor lite represents v5.
Custom firmware like FlashMe never created versions specifically for the 3 version of iQue and Kor, only for the international versions v1-v7. As a result, many iQue devices were flashed with FlashMe versions specifically for the international versions, losing the iQue firmware content that included Chinese language support. Even today, you can still find iQue phat/lite on the secondhand market that lack a Chinese language option; this is almost always due to the flashing of FlashME back then. noflashme also can't support restoration for these 3 firmware versions that have been flashed flashme, as it does not include the firmware data content and judgment logic of these 3 firmware versions.
The DSLink flashcart development team once created two versions of Flashme ports for iQue. Compared to the international version of FlashMe, the iQue port of FlashMe was obviously not as widespread, and it has long been lost to history on both World Wide and Chinese internet. I only managed to successfully port the iQue firmware to Flashme again after re-researching and creating a firmware unpacking and repacking toolkit last year.

ak1nok · Apr 30, 2026

Wokann said:
Yes, there was a lot of confusion and misunderstanding about firmware in the community in the early days. I only found various problems with a lot of the information after I actually tested it.

For v6, actual testing and comparison of the firmware's internal color hex data showed the correct color is still Magenta, the same as v5.

Today, one of my viewers left a comment under my video about Nintendo DS hacking, saying that the dark-blue v6 firmware does not exist, and linked me to this thread. I decided to look into it more carefully, because I have a Nintendo DS ML / refurbished Nintendo DS Fat with firmware that is clearly from the 4+ generation: the console does not work with PassMe 1, but it does work with PassMe 2.
At the same time, when I do the PictoChat test, the screen color does not look like the usual bright magenta. It looks much closer to dark blue. I showed this moment in my video here:

Youtube Video

After that, I checked the actual firmware more precisely. It turned out to be:
Phat — 2005/12/07 14:45 — v5 — Magenta
So this is not some mysterious “v6 on DS Fat”. According to the firmware data, it is the normal World Phat v5 / Magenta firmware.
However, I still think the old mention of a dark-blue color probably did not appear out of nowhere. On my console, this “Magenta” really does look much closer to dark blue, especially when looking at the actual old DS Fat screen rather than just reading the color name in a table. Maybe on dim, aged, or replacement DS Fat screens, magenta can be perceived as dark blue.
And this creates a particularly interesting problem. If we rely only on the old color-based classification and simply remove “dark blue” as a non-existent option, then my console could easily be misidentified as v2 / Grey-blue. Visually, the screen color really does look closer to blue than to magenta.

So there is a double trap here:
- a separate “dark blue v6” firmware for DS Fat probably does not exist;
- but if we simply remove “dark blue” from the old color chart, we can make another mistake: misidentify a late v5 / Magenta firmware, which looks blue on a real screen, as an early v2 / Grey-blue firmware.

In other words, the issue may not be a lost firmware, but the old visual classification by screen color. It was simply too imprecise. The color test alone is not reliable enough and should be confirmed by a firmware dump, build date, or at least by PassMe 1 / PassMe 2 compatibility.

Wokann · Apr 30, 2026

ak1nok said:
And this creates a particularly interesting problem. If we rely only on the old color-based classification and simply remove “dark blue” as a non-existent option, then my console could easily be misidentified as v2 / Grey-blue. Visually, the screen color really does look closer to blue than to magenta.

Yes, I think so too.
The reason why some people believed the Picochat color detection in v6 corresponded to dark blue was likely a display illusion caused by the semi-reflective screen of the DS Phat console. After all, the DS Phat's screen wasn't as backlit as the Lite version to display colors very clearly.
By examining historical records, it can be reasonably inferred that this is the cause.
The earliest discussion about v6 and deep blue appeared on page 8 of the "I've figured out a way to check your FW version!" post on gbadev.org.

A user named Emil007 mentioned that after sending his DS to Nintendo for repair, the staff informed him that they had upgraded the firmware. Emil007 tested it with PicoChat and determined it to be dark blue. Unfortunately, the image link wasn't saved, but it can be inferred that his console was a DS Phat, not a DS Lite (due to screen differences, the Phat is prone to color misjudgment, while the Lite is not).

Emil007's comment was posted on March 20, 2006.
Emil007 purchased a gray DS on its first day of release in Europe.
The DS Phat's release date in Europe was March 11, 2005. The DS Lite's release date in Europe was June 23, 2006.
The official v3 firmware was created on February 28, 2005. The official v5 firmware was created on December 7, 2005. The official v6 (lite) firmware was created on February 5, 2006.

Therefore, even though Emil007's pictures are no longer viewable, we can still deduce the following:
Emil007 purchased a gray DS Phat with firmware possibly ranging from v1 to v3 on March 11, 2005. One day, for repair purposes, he sent the console to Nintendo. The technician upgraded the firmware (most likely to block players from using PassMe1 and WiFiMe). The upgraded DS Phat appeared a deep blue color in the Picochat test.

We now know that the highest firmware version for the DS Phat was v5, while v6 and v7 firmware were specifically for the Lite version. Although the Phat and Lite firmware can be flashed interchangeably, flashing the Lite firmware onto a non-CPU20 DS Phat will result in the backlight not being able to be turned off, and flashing the DS Phat firmware onto a DS Lite will result in the loss of four levels of dimming, leaving only two modes: a completely black screen and a fixed brightness mode.
Normally, official staff wouldn't do something that could lead to user complaints. Emil007's subsequent comments didn't mention this issue either, suggesting that the last firmware upgrade for this DS Phat was v5, meaning the actual displayed color is magenta, and the dark blue is merely a misinterpretation caused by the screen display.
----------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------
There are also other pieces of evidence before and after the Emil007's comment that could support the DS Phat's visual illusion.

Diablohead mentioned on April 24, 2006, that he bought a Silver Mario Kart Bundle DS in the UK on April 19, 2006, which also displayed what appeared to be dark blue. Initially, he thought it was the v6 dark blue mentioned by Emil007, but later considered it might be v2. He also mentioned that the blue and gray display on the DS screen changed depending on the viewing angle.

kieran mentioned on March 9 and 10, 2006, that he bought a silver DS in Ireland in November 2005, and the DS shut down during a Picochat test. However, it was later discovered that the actual color was dark green. The initial test was conducted during the day, and the intense sunlight made the dark green appears as if the device was off on the DS Phat screen. This indicates that it was actually firmware v3.

In conclusion, the rumor that the v6 Picochat test showed a dark blue color was a misinterpretation of the v5 's magenta color due to a combination of the DS Phat screen's physical properties and visual observation.
There is no firmware version higher than v5 on the DS Phat.
The v6 firmware defined by the community later is a firmware specifically for the DS Lite, independent of the DS Phat (excluding the DS Phat with CPU20; the internal identifier of the v6 firmware clearly defines the target console as the DS Lite).

Added:
This can be corroborated from another angle:
The F-writer (the official firmware flashing tool) data shows that the earliest stored firmware data is up to Ver4.0, and Ver1.0-Ver3.0 are not present.

rvtr and I previously had a question: F-writer is used to boot from NDS cartridges and then flash firmwares, meaning that the DS device using F-writer must already have firmware installed in order to boot F-writer. In other words, F-writer can only be used to update firmware, not to inject firmware for an empty firmware chip at the factory.

Combined with Emil007's experience and the history of ds hacking, a reasonable deduction can be made: F-writer can be used by technicians to repair and upgrade the firmware of players' devices. And to block PassMe and WiFiMe from Ver1.0-Ver3.0, it will only install firmware starting from Ver4.0, which already blocks PassMe and WiFiMe.

Within F-writer's internal data, only Ver4.0 and Ver5.0 firmware versions exist for the NTR WORLD WIDE (DS Phat); no higher versions are available. The remaining versions are dark green color for NTR CHN [iQue] Ver1.0 (iQue DS Phat) firmware and magenta color for all region of USG firmware versions used by the DS Lite.
This confirms once again that the highest firmware version of the WORLD WIDE DS Phat is NTR Ver5.0 (communtiy called v5).

ak1nok · Apr 30, 2026

Wokann said:
In conclusion, the rumor that the v6 Picochat test showed a dark blue color was a misinterpretation of the v5 's magenta color due to a combination of the DS Phat screen's physical properties and visual observation.
There is no firmware version higher than v5 on the DS Phat.
The v6 firmware defined by the community later is a firmware specifically for the DS Lite, independent of the DS Phat (excluding the DS Phat with CPU20; the internal identifier of the v6 firmware clearly defines the target console as the DS Lite).

Great job, thank you!

I decided to change the firmware identification guide and make dsbf_dump.nds the main method for checking the firmware version.
I have updated my article accordingly:
https://gbasp.ru/proshivkands-en.html

Thanks again for the research!

Wokann · Apr 30, 2026

ak1nok said:
Great job, thank you!

I decided to change the firmware identification guide and make dsbf_dump.nds the main method for checking the firmware version.
I have updated my article accordingly:
https://gbasp.ru/proshivkands-en.html

Thanks again for the research!

I added some evidence regarding F-writer above just now; you can also add those to your article.

Looking for DS Fat Firmware variants (V1, V2, V3) for my preservation project.

Member

Nintendo DS hoarder

Member

Nintendo DS hoarder

Member

Nintendo DS hoarder

Well-Known Member

Member

Well-Known Member

Nintendo DS hoarder

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Well-Known Member

Popular threads in this forum