Yuppiee!!! I finally fixed my Event ID 18 CPU Errors in Windows 7!

I have been getting an Event ID 18 Error since a few months. I complained to DELL then they sent me a new motherboard, new CPU, didn't help, then they changed my laptop into the one you see in my sig with an upgraded GPU and CPU (780M GTX + 4900MQ) but the problem was still there.

The problem doesn't occur if I am using Windows 8.1, so every time I would give up and install Windows 8.1, use it for a couple of days, then get some quirks here and there and degraded performance in everything and every benchmark so I switch back to Windows 7, then back to 8.1, and so on, I have done over 40 formats trying every single driver combination, registry tweaks (Enable LargeSystemCache for example), changing the order, and I have finally figured it out!!

It seems to have been related to the Intel Chipset Software + IMEI (strangely)

Finally, no more Event ID 18 errors and I can live peacefully knowing that my laptop is working 100%.

These are the drivers I used and the exact order I installed them in:

01-Intel Lynx Point HM87 Driver 9.4.0.1027

Intel Dynamic Platform & Thermal Framework Driver for Extreme CPU 7.1.0.2104 (only if you have an MX CPU)

02-Intel Management Engine Interface Driver 9.5.15.1730

03-Intel Rapid Storage Technology Driver 12.0.7.1002

Intel Rapid Start Technology Utility 3.0.0.1056 (only if you have an mSATA SSD and use it as a caching drive)

04-Intel USB 3.0 eXtensible Host Controller Driver 3.0.2.54

05-Realtek RTS5209 Card Reader Driver 6.2.9200.28144

06-Intel HD Graphics Driver 10.18.10.3939

07-Alienware On Screen Display Application 0.33.0.11C

After the initial reboot, you need to press FN + F5 on your keyboard to switch to your dedicated GPU then install its driver

You will get a message saying:

08-Alienware nVIDIA Graphics Driver 332.70 or Alienware AMD Graphics 13.251.3.5 if your laptop had an AMD Radeon Graphics card.

If you prefer to use the latest nVIDIA Driver for better performance, then bear in mind that nVIDIA has disabled CUDA support in the latest drivers, to get CUDA support in the latest drivers, follow the instructions in this thread: Restore CUDA in nVIDIA's latest drivers

09-Realtek ALC3661 Audio Codec Driver 6.0.1.7260 (The reason this driver points to an external download link from Media Fire is because I uploaded it there with the PCEE4 folder which DELL forgot to include in the latest driver. Without that folder, the Dolby Digital application will not be installed thus not giving you the extra audio enhancements that Dolby provides)

10-Qualcomm LAN Killer E2200 Driver 1.0.36.1067

It is advisable not run the full setup as it will install the Killer Bandwidth Control which may cause BSODs sometimes. If you only want the INF driver to install via device manager, you can get it here: Download LAN Killer E2200 Driver

11-WLAN + Bluetooth Drivers

Alienware laptops come with different WLAN + Bluetooth cards so select the driver according to what you have in your laptop:

Broadcom BCM4352 Wireless Bluetooth 4.0 Driver 6.5.1

Dell Wireless 1550 WiFi / Bluetooth Driver 6.5.1

Intel Dual Band Wireless AC7260 Bluetooth Driver 17.0.1401.428 + Patch

Intel Centrino Wireless-N 7260 AC Driver 17.0.1

Qualcomm Killer Wireless-N1202 Driver 10.0.1.263

12-Synaptics Touchpad Driver 17.0.8.2

13-Command Center 3.6.4.0

07-ST Microelectronics Free Fall Sensor Driver 4.12.0046

14-Alien Autopsy (Optional)

 

So there we have it guys, this is the reason why Alienware killed the AW18. Changing CPU and motherboard, even the bloody notebook,when it was driver related lol.

With a company so big as Dell, I cant wonder if the money would be better spent on training the tech teams better instead of throwing it out the window like this

 

I bet it was a cheaper motherboard compared to one with the CPU and GPU welded to the motherboard.

 

Maybe the reason to go soldered was to save time?
You know, just change the whole thing with a motherboard swap now since they know the tech team would eventually have gone through everything on it anyway?

 

Dude, when I told the call center agent my M17X R4 webcam was not working, they didn't tell me I had to press FN + F9 on the keyboard, so they sent a technician to replace the webcam! when the tech came, by that time I had an advice from someone on this forum to try the FN key and to my surprise it worked! so i sent him back! they would've wasted money blindly replacing stuff and it still wouldn't have worked.

now due to this serious error of Even ID 18, they replaced a motherboard, a CPU, and then a whole new laptop with an upgraded 4900MQ vs my old 4800MQ and upgraded GPU vs my old 770GTX and it was all a driver and order installation issue which is all wrong on their FAQ links

 

That's just crazy. Glad you got it sorted.

It worked out nice for you, but that kind of technical incompetence may be partially responsible for this new stuff. Kind of like what Cloudfire said. I bet they have lost millions on using the shotgun method to repairs and system exchanges.

 

lol its funny, but its not really funny at the same time.
Just think about the money they are throwing away on this world wide?

Ugh, bleeding money on things that could have been avoided. Not good, not good

 

True! the call center agents love changing stuff!

When I had a previous complaint about screen bleeding, they changed the screen twice, then the LCD cable, then the motherboard, then gave me a new laptop! LL

so that's 2 laptops replaced now, brand new ones

 

*Tinfoil hat on*

What if they discontinued the AW18 just to have fewer parts that could go wrong and need to be replaced?
What if they went soldered because the tech teams replaced CPUs and GPUs when it was socketed, but now cant replace the soldered ones and instead will use more time on software check to ensure its not just software related?
What if they have lost so much money on this worldwide that they try to remove user inflicted errors with upgrades gone wrong or inadept technicians breaking stuff?
What if they instead try to encourage the user to buy the Graphic Amplifier to try to steer the user to using desktop GPUs in that box that Dell dont have to deal with insurance wise?

Interesting stuff...

 

You're making me feel guilty now that I am part of the reason why they stopped the AW18 :laugh:

Seriously, they could avoid so much wasted parts being changed if they hired employees with a 100+ IQ

 

Well you are probably not complaining lol.
But yeah, pretty sick to hear about actually. It cant be a good strategy for a company with so many customers

 

Alienware's demise is all Ferris23's fault. /s

 

Out with the pitchforks. Go get Ferris!

lol no seriously, its their fault not yours. You should get a working notebook. Its theirs responsibility to fix it the best way for both the company and you

 

but......umm.......



:hi2:

True, all I wanted since day one is a 100% working laptop. For the $6500 USD I paid for this config, that's the least I should expect, a laptop that works with the OS that I want (win 7 FTW) and no weird event log errors about a serious problem like CPU Cache Hierarchy Error.......that fux0red up my OCD so much I couldn't sleep

 

Congratulations.
I know that you used irst 12.8.0.1016 earlier but now you use Intel Rapid Storage Technology Driver 12.0.7.1002. Who gives you the best bench result?

 

Man you have a great memory!! You even remember what version of drivers I used!!!!!!!

You be the judge bro

AS SSD Benchmark with IRST 12.0.7.1002 RDx2 (W7) 16K




AS SSD Benchmark with IRST 12.8.0.1016 RDx2 (W7) 16K

 

All that is of great interest, I try to remember you know ..
How do you think this irst driver works with a simple ssd and win 8.1 ?

 

What if they don't want to have any customers left except for stupid people that buy crippled BGA garbage?

 

As long as they buy the new soldered computers and if it sell the same or more than the previous models, we have lost the war Mr Fox.
There is great resistance on this board, but does our opinions and view on the whole matter reflect on the other potential Alienware buyers as well?

Ive been visiting AW facebook page from time to time and most were super enthusiastic about the Alienware 13 and the GA. I think 90% of them are oblivous to fact that the CPU sucks.

I would actually wager that the new AW15 and 17 R2 is super popular without even visiting their facebook page.
In all honesty, I think this is the end for Alienware as we know it. The GA unit may dissappear if it sell badly. The only chance they revert back to using MXM GPUs is if the new soldered strategy ends up costing them more money on service and repairs than previous models, but looking at Ferris23 experience and the way Dell just replaced parts without even having to, I fear it may be a valid strategy for them financially to go soldered. They will also net more money on future sales because now many will be forced to buy a brand new computer from Dell if they want the newest GPUs, where many would have upgraded the GPU earlier.

TL;DR:
We have lost the war

 

Just to sumz up. You told that Intel chipset drivers and Intel ME were the culprit. What was the difference? What you have done wrong and done right now?

 

Heck if I know, I tried every combination and driver install order until this finally worked. Just a side note, I am on BIOS A10 and using the factory OC Level 1

Mind you, the issue was there before with or without the overclock. And when I formatted and installed Windows, the error was still there, every reboot, one entry is added to the list of Event ID 18 Errors : CPU Cache Hierarchy error, even after Windows updates

only until I installed all drivers in the order I mentioned, then the error stopped. Then I cleared the event log completely, and have never seen an event 18 error again

 

I'd create a recovery image then

 

Im glad to hear You fixed it Ferris23. Im on W 8.1 at the moment but i will definitely test Your way in the future.

 

I (we) haven't lost anything except for the name Alienware appearing on the list of potential puchase options. I won't concede to stupid people winning anything. If they find enough people dumb enough to pay money for BGA garbage to keep the lights on, then that's a bad reflection on the ignorance of the average consumer. Some will certainly experience buyer's remorse, but you know there will be some the remain oblivious to the lameness of the product long after the sale. Noobs and derps often have difficulty recognizing themselves as such, but either way it's no skin off my nose.

 

Mr Fox, that's harsh but so true

 

Thanks for the post Matrix Leader 6! I am seeing these WHEA Errors on an Alienware 14 with Windows 7 as well. I haven't found any other users who are experiencing the same issue. These errors appeared all of a sudden in September 2014 -- only on reboot, never on boot from power off. The fact that they appeared all of a sudden and did not disappear when I reverted to factory default with Alien Respawn is a little concerning.

I eventually ran Memtest 4.3.7 and saw errors almost immediately -- swapped in replacement RAM and saw errors again (in the exact same memory address) so I am thinking something is off with my CPU/Motherboard.

Matrix Leader6, was anyone else in these forums seeing a similar error? Did you ever run memtest86? It seems like there would be more people on these forums with the problem if it was something common to all Alienware laptops and Windows 7 (unless most people use Windows 8.1 and don't see the error)...

 

Never ran memtest since it cannot be run from Windows and I wasn't bothered to create an ISO and run it.

Here is a little update, the error randomly appears sometimes, when it does, go to your BIOS, restore optimal defaults, and it fixes it again. Try that

 

Interesting...

I am trying to get word from other Alienware 14 users on Windows 7 to see if they see this error as well. It doesn't appear to be very wide spread as your post is the only other one I have found with the WHEA 18 error (if anyone else has seen this WHEA 18 problem in Windows 7 -- please let us know). The bottom line is these fatal hardware errors shouldn't appear, at least not on a regular basis. It is also concerning to me that I didn't see these errors right away. I had the system for several months before they started to appear.

Did you see your WHEA errors from the day you got your system?

I am concerned that there is either an incomplete memory map in the BIOS (seems unlikely) or a potential hardware issue with the memory path (controller, bad slot, etc). Either one can eventually cause stability issues and errors. I am set to send my system back to Dell for a Motherboard/CPU replacement, something I was reluctant to do, but something isn't quite right here.

I'd be interested to see what a Memtest86 test shows on your system, both from a cold boot and from a restart, if you see the same errors in the same address space it would suggest Dell has problems with the memory map in the BIOS. I am using http://www.memtest86.com/ to test. If you have an old USB laying around it will create a bootable drive for you. I would run it in the standard non UEFI mode (4.3.7), that is where I am seeing errors fairly quickly.

Unless several others can confirm they are seeing the same issue -- indicating a very common error with newer Alienware laptops and Windows 7, I am still leaning toward something being off with our systems -- the WHEA error shouldn't be happening so regularly. Alienware support had no record of this issue in their support database and had to resort to Google to troubleshoot, suggesting this isn't terribly widespread (or that people aren't noticing it).

Again, if anyone else has seen this WHEA 18 error on their Alienware system -- please let us know!

 

@Matrix Leader6; Is it so that this event did not come up more when you switched to a extreme processor?

 

With the extreme CPU, it only came once when I used Windows 7 for 24 hours. Much less than with the 4900MQ which happens upon every reboot only if connected to AC power. If you reboot with battery power it never shows up. I have a feeling it is somehow related to the heat, maybe when the CPU exceeds a certain thermal threshold the error starts showing up.

 

I don't think so. A Intel processor have a throttling point AT 100 degrees. Why should it be a certain thermal threshold with a cpu ? A extreme processor can normally be hotter than a i7-4xxx mq processor anyway... Why should it be related to heat then?