This post is more than 5 years old
2 Posts
5
231020
Dell XPS 9550 CPU Multiplier Stuck at 8
Hi,
I recently noticed my new xps 15 9550 is really slow. So I downloaded CPU-Z to take a deeper look. I found the multiplier of CPU is stuck at 8, no matter how much stress I put on the CPU by running stress test like prime64.
Then I tried to disabled SpeedStep in BIOS and the problem still remains.
Now I have to use ThrottleStop 8.00 to force the multiplier to 26 in order to get get the normal performance. In ThrottleStop I had to uncheck BD PROCHOT option in order to force the multiplier. According to ThrottleStop document, BD PROCHOT stands for Bi-Directional Processor Hot. This allows other items to signal the CPU to tell it that the CPU is hot and that it should throttle even when it isn't. It might be the GPU or the power regulator when it gets too hot or starts putting out too much power that signals the CPU to throttle and conserve power and cool down the overall system temperature.
Is this a known issue? Is there a good solution instead of relying on ThrottleStop?
=====Attached CPU-Z report.
Processor 1 ID = 0
Number of cores 4 (max 8)
Number of threads 8 (max 16)
Name Intel Core i7 6700HQ
Codename Skylake
Specification Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz
Package (platform ID) Socket 1440 FCBGA (0x5)
CPUID 6.E.3
Extended CPUID 6.5E
Core Stepping R0
Technology 14 nm
TDP Limit 45.0 Watts
Tjmax 100.0 °C
Core Speed 798.0 MHz
Multiplier x Bus Speed 8.0 x 99.8 MHz
Stock frequency 2600 MHz
Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T, VT-x, AES, AVX, AVX2, FMA3, TSX
L1 Data cache 4 x 32 KBytes, 8-way set associative, 64-byte line size
L1 Instruction cache 4 x 32 KBytes, 8-way set associative, 64-byte line size
L2 cache 4 x 256 KBytes, 4-way set associative, 64-byte line size
L3 cache 6 MBytes, 12-way set associative, 64-byte line size
FID/VID Control yes
Turbo Mode supported, enabled
Max non-turbo ratio 26x
Max turbo ratio 35x
Max efficiency ratio 8x
O/C bins none
Ratio 1 core 35x
Ratio 2 cores 33x
Ratio 3 cores 32x
Ratio 4 cores 31x
TSC 2592.0 MHz
TSC_24 24.0 MHz
APERF 798.2 MHz
MPERF 2594.0 MHz
IA Voltage Mode PCU adaptive
IA Voltage Offset 0 mV
GT Voltage Mode PCU adaptive
GT Voltage Offset 0 mV
LLC/Ring Voltage Mode PCU adaptive
LLC/Ring Voltage Offset 0 mV
Agent Voltage Mode PCU adaptive
Agent Voltage Offset 0 mV
Temperature 0 41°C (105°F) [0x3B] (Core #0)
Temperature 1 43°C (109°F) [0x39] (Package)
Power 0 4.12 W (Package)
Power 1 1.90 W (IA Cores)
Power 2 0.06 W (GT)
Power 3 2.15 W (Uncore)
Power 4 1.28 W (DRAM)
Voltage 0 0.80 Volts (VID)
Voltage 1 +0.00 Volts (IA Offset)
Voltage 2 +0.00 Volts (GT Offset)
Voltage 3 +0.00 Volts (LLC/Ring Offset)
Voltage 4 +0.00 Volts (System Agent Offset)
gmanny
5 Posts
16
January 8th, 2016 15:00
For you I've got a workaround.
You need to shutdown the PC, then open the bottom lid up (unscrew 10 torx t5 screws and 2 philips ph1 under the XPS plate) and disconnect the battery for 10-15 seconds, then reconnect it back.
Other people in this thread discovered that you also need to hold the power button for 5-30 seconds while the battery is unplugged for the fix to work.
On my notebook it returned the speedstep back.
thom.a.vandijk
1 Message
3
August 25th, 2016 09:00
I had the exact same problem but I didn't want to open the back of my laptop.
So what I did was letting the battery fully empty, then when I wanted to turn it on it would automatically shut down. I did this for a few times until I was certain that the battery was completely empty. (I know that this is not very good for the battery)
Now everything is working perfectly fine! And speedstep has returned back! Running at 3.0 Ghz!
gmanny
5 Posts
2
January 8th, 2016 15:00
I'm seeing this exact issue too.
Dell, please don't tell me that I wasted $2300 on a laptop that can only do 800 MHz, that's too cruel!
you4ever
2 Posts
0
January 8th, 2016 16:00
Good to know there is a workaround besides the ThrottleStop. Wish I could verify it independently. However, I've returned the laptop for refund before the return period expires.
Wonder how did you figure this workaround out? I exhausted (at least I thought so) all online documents/topics and dell technical support (which is almost nothing).
Thanks
gmanny
5 Posts
0
January 9th, 2016 03:00
Well my return period ran out and in Russia you need a warranty technician to say that the issue is unfixable before you can get a refund, so I was preparing the laptop for warranty by removing dust filters I put over air intakes on the bottom, when I got the idea to also unplug the battery for a short time. And it worked.
That of course is a very terrible workaround, you need to basically lug a torx screwdriver with you everywhere in case it happens to you on a trip, so I would very much like a fix for it from Dell.
schmoo2k_3772ce
2 Posts
1
March 14th, 2016 12:00
I had a similar issue and this "fix" worked for me.
I am none the wiser as to the cause of the issue (while researching the issue, it was suggested that either a "bad" power supply or a "thermal" event may trigger this "safe" mode, but I have to evidence to support that).
gmanny
5 Posts
0
August 25th, 2016 14:00
One of my colleagues also had a similar issue with his XPS 13 9350 (in February 2016) and the support person said that Microsoft released a faulty update then that led to a multiplier being stuck and suggested Windows reinstall. Which helped, interestingly enough.
So, just for people who may find any other suggestions not working for them, reinstalling Windows may also be a valid option.
schmoo2k_3772ce
2 Posts
0
August 25th, 2016 14:00
FWIW I had tried letting the battery drain to 0 as well, but it didn't help - I suspect I wasn't patient enough to let it get all the way to empty as it kept shutting down to protect itself....
Anonymous User
1 Message
0
September 20th, 2016 11:00
For gmanny - many thanks - that worked a charm. As you noted I unscrewed the back, disconnected the battery (also held Power Button for 5 seconds as per the Service Manual), reassembled and I'm back to the normal high speed in Linux. Prior to this I was stuck at 800MHz in Linux (Mint 18 with 4.7.0 kernel), Linux via USB boot (to Mint 18 with 4.4 kernel) and Windows 10 (on the other partition), for a couple of weeks, with no idea how I got into this state. Cheers!
nsolovan
12 Posts
0
September 29th, 2016 17:00
Hi all!
My laptop is a Precision 3510 with i7 6700hq.
I did a fresh install of W10Pro, I updated the BIOS with the latest version available for download but no success. The only way I can get 2.59 Ghz, is when I launch Windows in Safe Mode. I tried with a live version of Fedora. The CPU will stay in 799 Mhz.
I will try to drain the battery, as I don't want to lose my warranty. Anyway, I'm not happy with this situation.
If you guys have any other suggestion, all of them are welcome!
Thank you!
shuyg
2 Posts
0
October 9th, 2016 03:00
Removing the battery was the solution for me too.
I did connect a 65 Watt power supply and had some Windows 10 updates. I don't know which triggered the CPU to get locked at 800MHz.
shreeneetrathi
4 Posts
0
November 14th, 2016 00:00
Did this solve your problem?
taylor.kline
1 Message
3
February 16th, 2017 11:00
gmanny I never would have believed this was the fix but you got it!
I wrote here about discovering your solution:
www.reddit.com/.../
Lilster
1 Message
1
October 13th, 2017 07:00
Thank you so much for taking the time to document this fix. I had an out of warranty XPS 15 9550 and it was running very slow. Geekbench4 was showing (single/multi) 1227/3713 compared to the community scores uploaded for the processor being 3729/11640.
Quickmark also showed the Mem bandwidth as 4mb where I would expect to see 12MB +.
Called Dell who said it must be software if you have updated everything and if the diags pass. I could not convince them this was a hardware issue.
Only suggested options were:
- £39 to send away for 10-12 days for diagnosis (cost to fix to be agreed later)
- reinstall windows and that will probably fix it
I did as suggested, took back off, disconnect battery, pressed the power button for 30 seconds and then reconnected and rebooted.
Like a different machine but also I have the actual benchmarks to prove it.
Geekbench 4 now getting 3530/12014
Quickmark mem bandwith up to 15GB/s
I just hope anyone else who has this find this post. Thanks again