Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iM 4,1-14,4 2011 iMac Graphics Card Upgrade
- Thread starter MichaelDT
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
This is not true any longer! It was slow just because we did not know in that time how to configure the cards (PolarisBoost). And we had only one case where a card was put into a Dell Laptop. Do not make a general claim out of this.
The point is readers of this thread buy cards from Chineses resellers (Dell and HP spare cards), claim the card is broken, resend it back to China just the let the seller discover it is working perfectly fine in the Dell/HP Laptop.
If I would resell cards I would not accept any return of a user trying to put such a card in a system not designed for this. You move all the risks of such a deal over to the seller, although only the (unexperienced) buyer is fully responsible for the outcome of such an experiment.
Please do not post more rumours or speculations about this special topic. Please report success or non success with a picture of the card. The problem with the non working AMD cards exists for the WX4130/WX4150, too. So it cannot be just related to the difference of two (or more) WX7100 card versions.
i don't realy spred a rumour, i just said that even AMD don't know some simple stuf. realy only saying some thing to perhaps find a source to get good cards. its still everyone's own responsibility YES ! AGREE. to get and buy a card.
thats why i choice to not post where i get the card.
i could have said.. i buy it here.. also you buy it here..
i choice to be smart and don't do that this time.
like i said
And on my part. The seller said, and know I going to put it in a iMac. So also his responsibility I assume.
He could have said. No we don't sell cards to be put in imac's
thats why i choice to not post where i get the card.
i could have said.. i buy it here.. also you buy it here..
i choice to be smart and don't do that this time.
like i said
And on my part. The seller said, and know I going to put it in a iMac. So also his responsibility I assume.
He could have said. No we don't sell cards to be put in imac's
Last edited:
As far as I can tell from reading through this thread, there are 2 issues with the AMD WX4130 and WX4150: there are (at least) 2 versions, one (for HP laptops) with an "00" suffix on the part number printed on the board, and one (for Dell laptops) with an "00D" suffix on the board, and the Dell version is more compatible? Also, the cards are sometimes not seen at all by a 2011 iMac?
So, I have a 2009 27" iMac, and I see an AMD WX4130 card with the "00D" suffix (full part number is 109-C98581-00D-02), it should work?
Thanks!
So, I have a 2009 27" iMac, and I see an AMD WX4130 card with the "00D" suffix (full part number is 109-C98581-00D-02), it should work?
Thanks!
I have to agree on @Ausdauersportler in this one. He said once ..
No you can't be sure. You better can go with the cards that have native support.
The AMD cards are a huge Rusia roulette
Thank you for the quick reply. I have searched the entire thread for posts with "4130" and "4150". All of the problems I have seen were from 2011 iMacs and people with the HP version of the card. Have you seen any post from a person having a problem with the "00D" card on a 2009 27 inch iMac?
From page 1: "There are multiple versions of this card - only the "Dell" branded versions currently work. "
But just below this, it also says " Even worse, some models are not detected in any iMac model from late 2009 to mid 2011." From these 2 statements, it's not clear if the Dells always work, or if even the Dells sometimes don't work. Not sure if "some models" means the non-Dell models...
Long story short:
We burned a lot of money on buying some different cards and ended up with a majority of not working ones - even cards with the same numbers sometimes do and do not running. No evidence found for a general green light.
Buy such a card, take the risk, check it out, and report back! This thread needs more contributors
The config part has been done for you to play with (latest OpenCore image with PolarisBoost).
I would also love to see some sort of very rough approximation of each graphics card's speed, maybe as a ratio with the 4850 as a baseline. I have been working through the benchmark page, and I am very happy to see that 2 people put up benchmarks for the 4850 that came stock in my machine. Its Unigine Valley score is around 650 and FPS is around 15 in an i7-860. For comparison, the WX4130 is clocking in at around 1700 at 40 fps. The K3100M with the same i7-860 CPU is turning in similar numbers, so this suggests that the WX4130 and K3100M are both 2 or 2.5x as fast as the stock 4850. The GTX780M is at around 2000 and 48 fps, so around 3x as fast as the 4850. This is another helpful factor in deciding which card is best for me.
You can do this calculation on your own. This is not a buyers guide here.
This is a technical thread discussing the work needed to be done to get other cards working in iMacs 2009-2011.
Yes, thank you. I AM doing this calculation on my own. I thought others walking down this path might appreciate seeing those numbers. You may say this is not a buyer's guide, but as far as I can tell, this thread is the only resource online for people wanting to upgrade their iMac video cards.
I have two machines that need new cards...when the 4850 died on my son's 2009 27" imac the first time, Apple replaced it cheaply at the Apple store. When it failed the second time, I was ready to replace it but found another identical iMac for sale nearby, and the selling price was about the same as the price to buy a used video card online. So the computer with the bad card went into the closet. Now, several years later, the "new" iMac is having video card issues, so I have 2 machines with video cards to replace. I also have a 2011 27 inch iMac (my machine, not my son's), and I personally replaced the 6970 card on that machine once...it is still working, but when it fails, I will try one of the options from this thread instead.
In the other post above, where I ask about the WX4130, you invite me to buy one and see if it works. I would like to do this and help solve the mystery of why some 4130/4150 cards work and others don't. This is a great idea, but for my first repair I think it's smart to start with the simplest replacement. Then, when I am more of an expert, I will attempt a more daring upgrade.
I will visit my local computer recycling shop tomorrow to see if they have any of the cards on the Page 1 list...
Great work!! Asking for the 765 users amongst us. Please, please, please, pretty please????
I've tried both with Ubuntu Live CD and with Xanderson USB Stick Method also with a chip programmer. Nothing worked.
Maybe by reading some posts you came to the conclusion that upgrading imacs isn‘t trivial. Thats not only due to some sort of secret experts knowledge people in ghis forum don‘t want to share. It has more to do with an OS which does not like HW upgrade ad long as the don‘t come from Apple itself and cost billions, on the other hand the buggy AMD IS X drivers, lacking Nvidis support by Apple, hardware varisbility on Card side etc etc.
Regarding performance the same holds true -> there is no dimple and short answer you might be seeking. On the other hand the first post on this thread is quite comprehensive. I‘ll gce you a hint: there is a link to a google spreadsheet where people post their Benchmark results...
I have a suspicion, but didn‘t verify it yet as using OpenCore and injecting a System ID broke my GraphPaf Prism (statistics) license. My assumption is: If you do groupwise comparison of all card results by lets say t-test or better one-way ANOVA, the difference between the reported cards will likely be nonsignificant. But I might be wrong, it‘s just a guess
So much still to learn...
Regards,
roscho
Hello, after upgraded my iMac 21.5 mid2011 now I wish to upgrade my OS From High Sierra to Catalina with OpenCore.
Just to give you complete informations I have two disk (SSD and HD), on the first there is High Sierra and on the second there is Win7 installed with Bootcamp.
Sorry for the "stupid question": is OpenCore Catalina an upgrading of my actual system or is a new system installation that I can install on a separate partition disk ?
Thanks!
Just to give you complete informations I have two disk (SSD and HD), on the first there is High Sierra and on the second there is Win7 installed with Bootcamp.
Sorry for the "stupid question": is OpenCore Catalina an upgrading of my actual system or is a new system installation that I can install on a separate partition disk ?
Thanks!
Hi!
You can do both, but I would add a new APFS container on your main boot disk and install Catalina using the @dosdude1 patcher into this new container.
Every question about installing Catalina on unsupported Macs will be discussed in this thread....
Using OpenCore with the Catalina Loader is a completely separate story targeting on the Brightness Control. Take a look at post #1 on page 1 got the thread, please read it, you will find docs linked there.
You can do both, but I would add a new APFS container on your main boot disk and install Catalina using the @dosdude1 patcher into this new container.
Every question about installing Catalina on unsupported Macs will be discussed in this thread....
Using OpenCore with the Catalina Loader is a completely separate story targeting on the Brightness Control. Take a look at post #1 on page 1 got the thread, please read it, you will find docs linked there.
Wow thanks Ausdauersportler, only now I've understood that they are two separate way (dosdude1 and OpenCore) for different scopes !
My idea to change with OpenCore Catalina (or either High Sierra) is because I've problem during upgrading with my "standard" installation of High Sierra....
The Catalina Loader IS NOT Catalina! It is just a piece of software to give you back the brightness control with the K2100M.Wow thanks Ausdauersportler, only now I've understood that they are two separate way (dosdude1 and OpenCore) for different scopes !
My idea to change with OpenCore Catalina (or either High Sierra) is because I've problem during upgrading with my "standard" installation of High Sierra....
While upgrading High Sierra you might boot directly into High Sierra using alt/option, do the upgrade and use the Catalina Loader after completing the installation.
Do the same while installing Catalina - the installer gets confused by OpenCore/Catalina Loader. Do my a favour and start reading in post #1, again!
@nikey22 I bought, flashed and installed a k2100m (2gb) after there was a vbios for it which worked and brought back the boot screen (No brightness control), and Im so happy to be able to use it and use Catalina.
As I revisit this thread now I see that you have made some awesome progress and achieved so much, incredible work! The only question I have is if it is possible to get the same functionality for the k2100m (perhaps the k1100m too)? Making it available to more mxm-a, which the iMac 21.5” is limited to - Or is there some technical obstacles that makes it impossible?
As I revisit this thread now I see that you have made some awesome progress and achieved so much, incredible work! The only question I have is if it is possible to get the same functionality for the k2100m (perhaps the k1100m too)? Making it available to more mxm-a, which the iMac 21.5” is limited to - Or is there some technical obstacles that makes it impossible?
Yes, I know that at page #1 there are all necessary information about. But for my upgrade problems with SecUpd2020-004HighSierra I can't find answer into this forum...
You have not described the nature of you problem! It works pretty well and I do it once a week on test systems ....
Thank you for replying.
Yes, that is where the numbers in my post came from.
I have a different guess, depending on your definition of "nonsignificant". I assume that the Unigine scores are linear (a score of 2000 is twice as fast as a score of 1000)? I don't have access to the statistical software you use, or a statistics background, so this is not a rigorous statistical analysis, but when I looked at the benchmarks page and sorted by year and CPU, I see differences that seem fairly dramatic. For example, within the group of 2011 27 inch iMacs with i5-2400 CPUs, the Unigine scores range from 463 to 2238. For my machine (2009 27 inch with an i5-750), scores range from 413 to 1677 and FPS varies from 9.9 to 40. The only machines on the spreadsheet with my stock GPU (4850) have an i7 processor, and are showing scores of around 650 with 15 fps. If I had seen this thread BEFORE my video cards died, I would have run Unigine Valley on it to see the benchmark for my configuration. Now the card is not working, but it seems safe to assume that the i5 has somewhat lower scores than the i7. From the spreadsheet, it appears that some cards (K610M) could be even slower the stock 4850, while others (the very attractive, but unpredictable WX4130 and 4150) offer 3x or 4x the speed.
I do have a 4850 that I bought a year ago on eBay to repair the first broken machine. For the moment, maybe I will just put in that GPU to quickly get the machine running again for my kids and to get a "stock" benchmark score, and then I can use my other machine to experiment.
If you ever do get your software working to do a proper analysis, duplicates in the data may be a challenge. It seems very likely that these are all the same machine, for example:
Yes, that is where the numbers in my post came from.
I have a suspicion, but didn‘t verify it yet as using OpenCore and injecting a System ID broke my GraphPaf Prism (statistics) license. My assumption is: If you do groupwise comparison of all card results by lets say t-test or better one-way ANOVA, the difference between the reported cards will likely be nonsignificant. But I might be wrong, it‘s just a guess
I have a different guess, depending on your definition of "nonsignificant".
I assume that the Unigine scores are linear (a score of 2000 is twice as fast as a score of 1000)? I don't have access to the statistical software you use, or a statistics background, so this is not a rigorous statistical analysis, but when I looked at the benchmarks page and sorted by year and CPU, I see differences that seem fairly dramatic. For example, within the group of 2011 27 inch iMacs with i5-2400 CPUs, the Unigine scores range from 463 to 2238. For my machine (2009 27 inch with an i5-750), scores range from 413 to 1677 and FPS varies from 9.9 to 40. The only machines on the spreadsheet with my stock GPU (4850) have an i7 processor, and are showing scores of around 650 with 15 fps. If I had seen this thread BEFORE my video cards died, I would have run Unigine Valley on it to see the benchmark for my configuration. Now the card is not working, but it seems safe to assume that the i5 has somewhat lower scores than the i7. From the spreadsheet, it appears that some cards (K610M) could be even slower the stock 4850, while others (the very attractive, but unpredictable WX4130 and 4150) offer 3x or 4x the speed. I do have a 4850 that I bought a year ago on eBay to repair the first broken machine. For the moment, maybe I will just put in that GPU to quickly get the machine running again for my kids and to get a "stock" benchmark score, and then I can use my other machine to experiment.
If you ever do get your software working to do a proper analysis, duplicates in the data may be a challenge. It seems very likely that these are all the same machine, for example:
| 5/18/2020 16:44:04 | 2009 | 27" | i7-860 | WX4130 | 1683 | 40.2 |
| 5/22/2020 16:17:55 | 2009 | 27" | i7-860 | WX4130 | 1683 | 40.2 |
| 5/22/2020 16:56:18 | 2009 | 27" | i7-860 | WX4130 | 1683 | 40.2 |
| 5/24/2020 5:35:49 | 2009 | 27" | i7-860 | WX4130 | 1683 | 40.2 |