XPS 15 9560 owners thread.

Whether or not you think it makes sense pl1 is set by the cpu to 45w as explained here http://forum.notebookreview.com/index.php?posts/9855390
The laptop may have it's own ideas and trigger another pl1 via the bios but it's still related to wattage used.

And just to throw in the reason I don't increase the efficiency of the vram pads is to prevent the heat coming from the cpu being transferred into the vram via the heatsink.

 

Yes it will work fine, I have the 960 Evo 1tb and the 1st batch XPS 9560s had the 961

Sent from my SM-G920F using Tapatalk

 

Hello. New here, so would love your advice. I've been having the darned issue with intel's 530 drivers for a very long time now on my XPS 9550.. and I've been putting up with it... the warranty on my laptop was long gone. Anyways, after reading your Solution @GoNz0 , I managed to remove using DDU, but I have had to rollback all the way to 4352! Anyways just wanted to say thanks there, its all working now.

My question is, I just bought a 9560 - don't want to lose all my work etc and not really sure how to clone NVME SSD's... will it be okay if I just put my Samsung pm961 512GB SSD into the new XPS 9560 without rebooting windows on it, but as is with all the content? Not sure what SSD that laptop comes with, but I hear it's not as good anyway.

My concern is the drivers, chipset etc that could cause conflicts with the new system seeing as it's a different model.

Thanks in advance.

 

Hi all, I have a question.
Yesterday I was involved in a pretty bad car accident. My car is totally wasted. Luckily I was't injured!
I had my XPS with me in my laptop backpack and in its protective sleeve.
Unfortunately it got pretty damaged:







So as you can see it's damaged on the side, I also think the hdmi ports is broken. The screen is broken on the bottom and bottom-left side.

I purchased it at the end of June, luckily (I hope) with an "Accidental Damage insurance".

My question now is, will it be covered?
Read the terms & conditions and can't find anything that would suggest otherwise, but I just wanna be sure before calling to the insurance company.

Does anyone had any experience with this?

Thanks a lot!

J-J

 

If the CPU cores run at 1.5ghz they aren't drawing anything near max watts, so that should leave plenty of headroom for VRAM to draw some power. But let's assume this is the case, how much watts does the GPU VRAM pull would you say? I mean, if I cool them, there is no power throttling on the GPU whatsoever. The breaking point seems to sit around 55-60C measured on top of the heatsink above the VRAM (so actual VRAM temp could be anywhere from 60-75 I guess). Is the VRAM so power hungry when hot?

I'd agree, however CPU temps and even GPU temps are all well below thermal throttling limits. So suppose I'd want to try, what kind of thermal pads would be better? Any idea what the W/mK rating is for the Dell pads that come with it?

 

If you have Dells accidental coverage it will be covered, if you put your service tag in dells support site under the warranty link it will tell you there if you you have accidental coverage.

Otherwise yes call your car insurance and have them cover it. this will not be a cheap repair.
last - Glad you were not were not hurt.

Cheers
Dave

 

Easy way it to lock it into safe mode (google how to set the safe mode flag) transplant the SSD and after it boots you can take it out of safe mode. Stick the wifi drivers for the 9560 on the old laptop ready to install post swap and windows update will take care of the rest. If it all goes a bit tits up I would backup what you need (should have a backup anyway tbh) and format for a fresh start.

Many many other ways to do it better, you could simply connect both new and old to the home network, setup read privilages on the old one and copy of the network if oyu just want documents etc moving.

You had an accident and have accident insurance, go for it.
Hope the bruises heal quick!

Don't forget it has to trade off CPU and iGPU for that 45w, I can't explain why you are hitting it apart from the obvious but a PL1 is what it is and undervolting may keep you away from it but at the end of the day the chips a 65w chip with it's balls in a vice.

Dell I expect use around 6W/mK, ie cheap and cheerful, they do the job, anything above and you are passing heat the wrong way.

 

First, based on extensive reading of the posts here, I think the majority of 9560 cooling issues are with the VRM rather than the VRAM (the 9550 has a defective VRAM heatsink design which seems to be improved for the 9560).

Once the VRM sensor hits about 77-78*, the 9550 & 9560 generally will throttle. I think Dell did this because the VRM is underspecced & because the mosfets get very inefficient at higher temps (providing a lot less wattage) & because the XPS has very limited cooling capacity. There are other throttling schemes you might hit, including power related, but this VRM temp seems to be the most popular bottleneck.

You can see the VRM temps in HWiNFO64 as a (hot) ambient sensor. Also remember that any wattage reported by your system is a very poor estimate so is not especially useful.

I don't think the throttling reasons are always correctly categorized (or named). So that causes some confusion.

I estimated that Dell gives the CPU and GPU each about 45w max available in ideal cool situations. The CPU & GPU are powered separately with dedicated mosfets & chokes. If the CPU consumes less power, the GPU is still capped at about 45w (but the system benefits as there is more thermal headroom).

It is worth noting that the laptop 1050 GPU can be configured to run at much higher than 45w. I have seen some people claim up to 65w or 75w but have not found a relaible max spec. The XPS struggles at ~45w.

 

Thank you, as always!

 

Hi to all. I have found a couple of Patriot DDR4 16GB PC4-17000 (2133 Hz) CL15 SODIMM.
Are good for the Dell and, about the price, they cost about 150$ (are new)

 

for the 9560 you should look for DDr2400, 2133 will work at the slower speed. Not much in it if you benchmark them as slower ram has tighter timings to make up for it.

 

Thanks for reply. So you think that aren't good?

 

New Bios 1.50 outside
1. Fixed Microsoft PBA (BitLocker) recovery key screen prompt issue when plug in/out WD15 and TB16 dock.
2. Fine tune thermal setting for improve CPU/DGPU performance.
3. Fix thunderbolt hard drive issue in legacy mode
4. Enhance WLAN performance in 2.4GHz Ch6.
5. Enhanced DCPM warning message.
Version
Version 1.5.0, 1.5.0

 

I am using this RAM 2666

 

Cant wait to try before and after

 

I have you my answer m8
Up to you.

Sent from my SM-G920F using Tapatalk

 

I'm curious what is the Userbenchmark ram score.

 

http://www.userbenchmark.com/UserRun/5016178

 

After upgrade I've got some random WiFi drop which never happened before. It can be a coincidence and a problem with my router/provider so if anyone has similar problems AFTER upgrade to 1.5 - let me know, please.

 

Tnx. Best score I've spotted there was with Kingston HyperX DDR4 2666 C15 2x16GB, 85%. With the 9550 there were around 90% with certain G.SKILL 2666.

 

I'm 1.4.0, its also found Wifi is not stable. I believe is KillerWifi problem...

 

The ram that I have found is in the image.
https://imgur.com/a/bxgSw
I have see the 2400 version and the price is really high. Around the double.
I will have the proportional increment?
So, if i spend 300 $ instead 160 I have so much speed?
With the difference I can buy a samsung 960 evo.

 

That still does not explain why the throttling does not occur when I cool the VRAM with a blow dryer to keep it around 30-40C max. When I'm talking about GPU power limit throttling I'm talking about GPU-Z "perf cap pwr"/HWiNFO64 "Performance Limit - Power" for the GTX1050. All the while the VRM cooling is entirely ignored other than the excess air, and I'm deliberately pointing the cool air AWAY from the VRM and only on the VRAM under the QR sticker.

During Pwr throttling, GPU temps are in the 40-50C area, CPU is around 40 as well. When I cool the VRAM, GPU climbs to 68C at 1.7ghz and there is no throttling whatsoever.

Maybe I have a bad VRAM chip?

I ordered some 11W/mk pads to test as well as some K5 Pro. I understand what you're saying, but if the GPU temp is 15-20C below the VRAM temp, I think it's prudent to try

Right. I had that on the CPU with prime95, but there's zero throttling now with a -150mv undervolt. I added some thermal pads to the mosfets as per the iUnlock guide, but honestly i don't think they're needed after the undervolt. But this is not the throttling I'm talking about. The GPU power throttles even if I force the CPU cores to 1.6ghz max.

I don't really mind if the 1050 runs steadily at 1-1.3ghz with an undervolt and VRAM downclocked to -500mhz. The problem is that I can't even undervolt it when it goes into power throttle at 585mhz. The VDC goes to 0.750 or below and stays there; MSI Afterburner doesn't go that low so I'm entirely unable to even try an underclock or undervolt :|

 

Generally, "performance" RAM isn't worth so much in real life, despite the advantage in specialised benchmarks and specific tasks. 9560 comes with 2400 from the factory (9550 2100). But also the latency (CL) matters.

 

This is the part where I duck out as it has all been answered in the temps thread apart from the bit with the PL1 however you look at it. I have no more answers for you, a PL1 is aPL1 and means it hit it's max wattage assigned by BIOS or CPU

I doubt you have a bad vram chip but feel free to broach the subject with Dell, they may swap out a refurbished *cough* motherboard and it may fix it if you get an engineer capable of doing the job without trashing the laptop, this isn't a gaming laptop but a few people are hellbent on making it cool like one so head to the thermals thread and read it all as there is no need to ask the same questions again

Should have read the thermals thread before buying 11W/mK pads as they are hard and you need soft as it explained in the thermals thread. So good luck making them fit as most people know they won't, or they will cover the vram and not the CPU as a few people found out before you wonder why the CPU is 90 degrees C at 800mhz.

Prime 95 doesn't use the iGPU and isn't that punishing on a modern chip.

 

Thanks for the reply. Well I am not a hardware technician and so
I am not able with the test and so how can I understand if I buy something useful or not?

 

As I said, I have never had problems with WiFi (maybe only on first day, with default drivers, since then I'm using clean drivers from Killer website, currently ver 12.0.0.318).
The problems stopped so I think it was something not related with the BIOS upgrade.

EDIT:
Link to the clean drivers:
http://www.killernetworking.com/driver-downloads/category/other-downloads
and download Killer INF package, then install from Win10/11AC folder with Device Manager

 

So riddle me this then:



Nowhere does it indicate a PL1 limit reached, all ambient temps are well below the reported VRM temperature issues. Note this is just seconds into a Unigine Valley benchmark. If the wattage numbers are believed CPU is not even close to 15W. I haven't opened the case for this test but in this scenario, IR gun reads 60C+ on the VRAM below QR sticker.

Or does this PL1 limit you talk about not show up anywhere here?

Still gets the CPU as hot as the next thing which is all I care about, and iGPU isn't taxed by anything. But do you know a better alternative?

 

There is a possibility that some of your VRAM thermal pads are not making proper contact with your heatsink, your heatsink is bent or defective. Not a common problem on the 9560 as that was reengineered. Also I don't think blowing a little air over the VRAM should have much impact if there are fitment problems. Check those carefully first.

The heatsink made poor contact with 2 vram chips on my 9550 (popular problem). I purchased the Fujipoly 17 thermal pads in 0.5, 1.0 and 1.5 thicknesses and spent a lot of time making sure the each of the 4 pads had just barely enough pressure. It took a long time but thermal performance is superb now. There are some good engineering tips for getting the right thickness in the thermals thread.

With soft, fluffy, low rated pads it is easy because you can use slightly thicker pads than necessary and the fluffiness covers with a big margin of error. The fluffy pads also cover for imperfect alignment of the chipface & heatsink interface. So the lower rated pads perform much better than you might expect.

With the higher performance pads (read-dense) you have very little margin of error. If the pad is too thin, no contact to heatsink. Do you have 100% contact with the chip and heatsink interface? If the pad is too thick, then you push the heatsink away from the other VRAM chips AND the GPU; it takes very little pressure and the thermals thread shows several people tripped up here killing performance. GoNz0 also notes that higher-performance pads risk transferring heat tothe VRAM chips you are trying to cool.

Note the VRM needs very litte additional cooling to perform well. Just removing the case bottom may be enough. If not, the "excess air" you are blowing in the region should be. So your issue may still be the popular VRM problem. Just tough to guess.

I recall reading about one XPS with failed VRAM (it was a 9550). Possible but not a popular failure mode.

I think Dell's engineers sent a significant amout of heat from the VRM mosfets through the motherboard, through the VRAM chips, to the heatsink (no air cooling as stagnant so the motherboard is the only major way to transfer heat). That makes thermal diagnosis maddening. But still provides you some imperfect information to consider.

 

@infernix PL1 will stick for a prolonged period of time to let components recover, no riddle. Maybe you should have a look at pressings reply and see if you have already caused a problem messing about inside or it came with a bent heatsink, and take this to the relevant thermals thread at the same time

Or you might just have a **** board, who knows, I already suggested you should call Dell.

 

I won't edit my last post as there are some tips others can use for diagnosis. And I still think it can be relevant for your investigation.

But based on your pictures, it seems you don't have "recognized" thermal issues as temps are so low. I really don't understand how running cool air resolves your problem. At these levels power and thermal flags should not be influenced by a bit of air.

Could be a bad chip or contact or cold soldier joint but there is no way to diagnosis that without schematic.

Did you try uninstalling the GPU modding software? Those can be a bit buggy.

 

duplicate

 

My first unit was by far the worst premium computer I have ever bought. Occasionally, it froze, but I decided I could live with that. It stopped booting into Windows. We reset it, reinstalled Windows and flashed a BIOS update. All efforts failed. They replaced the SSD, which failed. They sent two techs to replace the motherboard and RAM in the same day, which meant both sticks ended up being replaced. That failed. I sent it in, and they sent it back to me saying it had been rigorously tested. The screen occasionally went black or became taken over entirely by ugly glitches, which cost me my work. They had me update screen and video card drivers. That failed to solve the problem. They declared it a failure and sent me a new unit. So far it is working well. My fingers are crossed.

 

I checked my invoice and was sure I had the "accidental coverage".
Once I had my (in that time) €750 iPhone 5s insured for damage, but when I needed it they refused it, because it was jailbroken, although you could just restore the phone to stock iOS with iTunes. Couldn't restore myself because all the screen and the home button were broken.
That's why I was worried...

Unfortunately my car insurance doesn't cover it. Only myself, passengers, the car and the counter party.

I called them today. The guy on the phone was very friendly. He booked everything in the system and Monday a courier is coming to pickup the laptop.
Hopefully I'll get it back within 2 weeks.

 

Are both USB 3.0 ports supposed to have the Powershare feature? On my laptop, Powershare only works for the USB port on the right side but I'm pretty sure according to the manual both ports should have this feature.

 

Just 1 afaik.

 

Well sadly the MSI GS63VR is going back because of the fan noise. It may be just a bad/off balance fan, but I don't think I want to deal with it. The laptop is really nice over all but the build quality is not as good as the XPS.
One thing that I am sad about:

 

First time poster here, getting a bit desperate with my XPS15's screen. It's the FHD model, and the screen's been replaced by Dell twice now. The original screen had some noticeable discoloration along the borders, especially so on lighter colors. Like a slight fade from dark to light. The replacement screen had a enormously green hue, fixed with turning the green 30% down, and also the same discoloration along the borders. The third screen, just installed yesterday, also has the same green hue and same border discoloration if you look at it from a slight angle. Angle doesn't need to be big, sitting right in front of the middle of the screen, and the sides appear darker. Move towards the darker sides, and they appear lighter. My external monitor is perfectly fine, so can't be other hardware or my eyes. Since it's three screens in a row, is this normal for this screen? Or should I just keep pushing them until I receive something better?

Besides that, the latest screen also has a larger discoloration to ever so slightly purplish on the right. Plus it's misaligned, with about a millimeter of screen is missing on the left side, and a black bar visible on the right, so definitely has to be replaced again.

 

Indeed, tell them you want a new build if this is 3 failed repairs as they have had several chances.

 

Whats up with Dells warranty upgrade methods? I am looking into an extended and maybe accidental warranty for my 9560.

I went online to Dells site expecting a nice little option page to pick a choose the plans and pay online. With warranty's for various devices and electronics other than Dell I have always been able to visually compare the yearly costs along with the warranty types, see the prices, fine print etc.. visually for each and or combined warranty's . Dell wanted no parts of that and insisted I picked a plan (over the phone) before they explained it and provided me a price.I would have been on the phone forever taking warranty notes and comparing packages and would probably not get all the details until purchased. keep in my mind I am not familiar with Dell so I went into this blind because their site warranty site is very vague. I was confused during the conversation so I just hung up on the guy. Once again I like my warranty choices in print before I purchase. Anyone else feel Dells warranty shopping experience is lacking and borderline shady?

Any recommendations for another warranty service or should i go through the hassle and time with the Dell phone calls. Any links to what their warranty's cover, benefits, repair time etc? I am still baffled by the unprofessional experience!!!!

 

I can get the left side to work by first plugging the device I want to charge into the right side USB port. Doing that somehow activates the Powershare function on the left side USB port.

 

^ as far as I know, powershare is only available from the right side. (at least on xps 13 that is)

 

I know, I know, the same question but I need help to make a decision.
Crucial 16GB DDR4 16GB DDR4 2400MHz - CT16G4SFD824A (250 euro two pieces)
or
Crucial 16GB DDR4 16GB 2133Mhz - CT16G4SFD8213 (150 euro two pieces)
Those are the best price I have found. All new.
And, most important, what do you think about?
Are good?

 

Why would you pay 67% more for a speed difference that you will never notice?

 

Thanks for the reply. Is what I think about but because I am not expertise in hardware
I am searching for some suggestions from who is more expert than me

And... It's sure that the first are compatible with 9560?

 

Ok, I figured it out. Powershare does work on both USB ports but because I have a Logitech nanoreceiver plugged into the right side, it thinks that that is being charged when my laptop is in sleep mode. Assuming that only one device can be charged at a time using Powershare, that is why the left side won't work when the nanoreceiver is plugged into the right side. If I remove the nanoreceiver, then Powershare works on either the right or left side.

Now my question is, is there any way to make the nanoreceiver not make use of the Powershare function? That way I won't have to remove it everytime I want to charge a device on the left side.

 

A Bluetooth mouse? Of course twice as expensive as the normal wireless one

 

No it's not sure but why don't you check the Crucial website:

http://www.crucial.com/usa/en/compatible-upgrade-for/Dell/xps-15-(9560)

 

How does the new BIOS 1.5.0 work for you guys? How about the idle frequencies and fan behaviour? I'm a little worried to update to that one because I've had some Dell laptops become louder after such "tuned thermal settings" that Dell mentions...

 

I haven't experienced any negative issues when going from 1.4 to 1.5.