[Test] Aorus X5

*UPDATE* After having repasted the cooling is much improved. See below original post please.

Original:
The Aorus X5 was launched in 2015 adding a 15-incher to their lineup and Gigabyte has been kind enough to send me a unit for testing.

My specification:
Aorus X5 - http://www.aorus.com/Product/Features/X5
i7-5700HQ 3.5 GHz
SLI GTX 965M 4 GB
15.6" 2880x1620 IPS G-SYNC
2x256 GB SSD Liteon + 1 TB 7200 HGST
16 GB DDR3


The X5 is very similar to the X7 both in appearance and size (apart from the screen real estate). The X5 comes in at 390 x 272 x 22.9 mm and 2.5 kg making it thin and light and very easy to handle.

To test the X5 in terms of gaming performance and cooling potential I did my usual test suite. Intel XTU to find lowest stable undervolt for the CPU, benchmarks and games were run at this undervolt and stock. All tests were run with the laptop elevated on a cooler and fans set to 100% in Command and Control.

Stock voltage for this 5700HQ is 1,110V. Lowest stable undervolt in Intel XTU was -80 mV landing at 1,032V. Real life voltage was later tuned to -70mV (1,042V) due to Battlefield 4.

Spoiler: Intel XTU stress test at -80 mV

Intel XTU stress test at -80 mV


Benchmarks used were 3DMark11 and Firestrike. Games were Battlefield 4, Crysis 3 and GTA V.
The behaviour was the same in all benchmarks and games: The CPU overheated and throttled, the GPUs ran really cool and gameplay ranged from decent to smooth. I didn´t experience any freezes or stutters related to drops in CPU speed. The only exception was Crysis 3 running at 3K High but that one is just too much to handle for this kind of hardware no matter the temps.

Spoiler: CPU and GPU temp and speed

CPU and GPU temp and speed


Generally gameplay was decent at 3K and smooth at 1080p.

Spoiler: Gaming performance

Gaming performance

Spoiler: Gameplay

BF4 1080p Ultra


BF4 3K High


Crysis 3 1080p Very High


Crysis 3 3K High


GTA V 1080p Very High


GTA V 3K High

In summary the CPU reached 95-97C in every benchmark and game, this was with the laptop elevated, max fans and undervolted CPU. A repaste would probably have helped, something I didn´t do. However gameplay didn´t seem to suffer from the throttling CPU and even GTA V at 3K High felt smooth. Considering the GPUs never exceeded 70C the cooling design should in my opinion be shifted a bit from the GPUs to the CPU.


UPDATE: After some discussion in this thread I decided to repaste the CPU (and GPUs) to see how much gain could be reached. Turns out quite alot. Max temps were now below 90C for the CPU except for BF4 which still creeped up to 96C. Though it would´t throttle as much in BF4 as with stock paste (3.09 GHz instead of 2.89 GHz).
XTU stress test, GTA V and Crysis 3 reached 82C, 82C and 89C respectively.

Of particular note is the 5700HQ wouldn´t power throttle either. In all tests save BF4 it held max clocks throughout. It could possibly be pushed to power throttle in these games as well with certain settings and in certain CPU-intensive levels. But I haven´t played through them to play anywhere and I´ve also held on to this review unit for far too long already and need to return it.

I also overclocked it to see if I could force the power throttle and it does indeed happen in XTU stress test.

Spoiler: 3.7 GHz, undervolted & repasted

]

3.7 GHz, undervolted & repasted

But I´m glad a thin laptop as an Aorus can be undervolted and repasted to adequately cool the CPU, if not in all applications and situations.

Here´s gameplay with undervolted and repasted CPU:

Spoiler: BF4 undervolted & repasted

BF4 undervolted & repasted

Spoiler: Crysis 3 undervolted & repasted

Crysis 3 undervolted & repasted

Spoiler: GTA V undervolted & repasted

GTA V undervolted and repasted

 

I have repasted mine and overclocked the cpu never gets into the 80s

You will understand when you repaste What is going on.

There are 2 heat synchs on each gpu.

2 arms from one and one arm from the other connect at the cpu. A bar connects them they are not welded.

So it is not even and you will see one gpu is over cooled

 

My cpu never throttles.

I even had pictures here of pictures I took of the repasting.

Where as many other laptops throttle the gpu that will never ever happen in the x5 it is cold.

The 965m at max overclock with this heat sync is cold.

 

I´m aware of the cooling design of the X5. That´s amazing having your CPU in the 70s even overclocked, what paste did you use?
Even more impressive is the neverthrottling 5700HQ. Do me a favor, run Intels XTUs stress test for 5 minutes and show me a screenshot of XTU and HWinfo. If it doesn´t powerthrottle at the 2-2.5 minutemark that´s probably a world first.

Since Maxwells release most laptops that throttle do so on the CPU. The GPUs were not the issue on this X5, they never exceeded 70C.

 

Sure I will

Not undervolted or anything over here and I rrepasted months ago it has come up to the 80s since then

I can run the 3dmark etc with no throttling at all

I used gellid ultra

 

I have to say I had my doubts regarding your claims and I repasted my X5 with Gelid. The results do seem much better. It still creeps up to 92-93C in GTA V and Crysis 3 but it no longer throttles.

I´m gonna try to get some more tests done before I return it.

 

Mine is repasted with Coollaboratory Liquid Ultra and stays below 70°C during a CPU-only stress test. No thermal throttling during gaming either. Did you guys really get completely rid of the TDP throttling observed during an Intel XTU stress test though?

 

You have your doubts why? You think I'm making something up?

Sorry bud one of us has absolutely no idea what they are talking about and one of us has tested his laptop hundreds of times since last summer

I have my doubts about anything you say i have my doubts you even know what repasting is. I have my doubts we need to waste any more time with you at all.

 

after 2 mins 30 seconds of the intel xtu stress test
my package temperature is 72c no throttling

do you have any pictures of repasting your x5? No? Really?

 

and at the end of the 5 minute stress test the temperature was 79c

after which because the fans were on max it fell to 49c almost within 10 seconds as you can see

 

everything you say is totally incorrect. since maxwells release, the 980m throttles on every laptop....

Im not sure why someone would send you a laptop to test what youre practicing to do or what. especially one that came out last summer. We tested everything.

 

i can overclock too to maximum and the cpu will not throttle in the stress test for ...

forever. hours. years.

I dont know what is wrong with you, you may have changed a setting in the cpu that messed it up.

But most importantly nothing you say about current hardware is correct. the 5700 almost never throttles. it throttles way less than the one before it or the one after it.

I think the other gigabytes with it may throttle more than the aorus they have way less cooling and are smaller too. normalltyt if yuo mean the p35x it also has a red hot 980 m

 

When I first repasted it it was in the 60s for a 3dmark test. The hottest it gets is during the cpu part of firestrike I can show you that in a minute. Ill reproduce this newbies tests I guess.

I can get the temperature up to 90 during firestrike

 

In my XTU stress test the CPU stayed at 3.49 GHz for the full 5 minutes (after repaste). No thermal, power or current throttling. Which is surprising considering it´s an HQ.

Wow you´re coming of pretty strong, offensive borderline douchebag. Especially since I just posted I saw the same results you claimed. Well not quite, I can´t keep my CPU under 80C while overclocked.
I had doubts because these thin laptops have a history of overheating the CPU, which mine also did, and I´ve seen your fanboi posts in the X5-thread and think you are a wee bit biased towards your own machine.
You don´t have to believe me. I have tested enough Clevos and Gigabytes to know what I´m doing.

Haha just stop This isn´t even true on the slim laptops. Please hop on over to the Clevo section with that claim, I´d like to see the responses.
I´ll give you a single point for the last sentence, I was surprised too when Gigabyte asked me because of its age but I didn´t ask.

 

Hey everything you said was wrong and obviously ignorant. You came off as pretty douchey to me, and everyone can review what you said and see its totally incorrect.
you contradicted the actual owners of the machine which is pretty arrogant and couldnt have an outcome other than being wrong imo
Theres some remark about how my 5700 broadwell cpu must be a record or whatever, when all tests show the same thing the broadwell doesnt throttle much in most of the laptops its in.... which are gigabytes and msi mostly

not clevos. So if you did test.... gigabytes.... right? It does run hotter and throttle in their p35 p34 might be a model with it i didnt look it up.... But not in aorus laptops which are going to be just the x7 and x5
I guess I can try to see where that comes from if you tested a p35. Ive seen like a dozen tests on the aorus x5 from the aorus forum and they all pretty much are the same. They could need repasting and they could not. Mine didnt throttle even when it was new but it was a lot hotter
it may be in other laptops but not clevos that i can recall

 

Do you have a picture of repasting the x5 you dont own? with gellid ultra, coincidentally the brand i used and you just had laying around.

RIIIIGGGHHHTTTT douchey

which part of what were looking through is a fact which is entirely BS?

you asked me to take time out of my day to display for you what I know for a fact which isnt friendly its rude. So I took that time for you or for whoever needs facts. I guess if someone was looking to buy the x5 when theres a new one coming out, ignore this guys temperatures theres something wrong with his tests.

 

wait so were not even asking for thermal throttling. you have undervolted the cpu an you are power throttling now? my test shows that it doesnt power throttle either but thats a bizzarre switch.

for those who dont know haswell cpus are terrible with power throttling but overall broadwell is very difficult to get to power throttle in any case but its apples and oranges. Not overclocked, your broadwell will not power throttle ever..... ever....totally different problem

so my xtu tests at 5 minutes show 16c below yours without overclocking.

but if you overclock the cpu you can power throttle it. I can do a test for you that may show it.

it will be nothing like haswell it will be barely detectable. i will show a 5 minute stress test. I doubt it will show up in that an xtu stress test is mild, i would be able to see it in the phyysics test of 3dmark 11
absoutely there is power limit throtttling when overclocked it got to 67 percent. repasting shouldnt change this in any way. you can play with undervolt etc I did months ago but i just dont overclock it now.

even overclocked it doesnt thermal throttle. its not even close actually but a 5700 will power throttle overclocked like a haswell

 

there it is i guess the never throttling 5700 at 2 minutes 30 seconds. its there at 5 minutes 10 minutes..... its there at 30 minutes.

I guess its possible to power throttle it but mine never does neither does any of the other owners. Mine never throttles before I repasted it for a few months but then eventually it would thermal throttle after long gaming sessions.

just a rude, incorrect statement you have there based on ignorance and i was glad to point it out for you.
you insinuated i was biased or whatever and i was glad to point out youre full of it.

 

In the bottom left of your posts there´s an Edit-button. Please use it.

This shows a temperature of 91C. Thats´s more than "not 80s".

I presented my findings, nothing more. I later revised them after a repaste. In my original post this contradicted your experience, is this arrogant? The 5700HQ in my X5 thermal throttled like crazy, that´s not arrogant, that´s fact.

Let´s move on before you embarrass yourself more.

Thank you for taking time out of your day to test this. After all, that´s what NBR is for, testing laptops and contribute to the community.

Where do you even get this from? Not once have I stated I power throttled. After I got rid of the thermal throttle I expected a power throttle but it doesn´t happen. The HQs usually do but this one doesn´t. This part is very interesting and we need to test more.

Saying soldered Broadwells wont power throttle is just wrong.
After repaste I got a peak temp of 83C in XTU, just slightly above your 77C.

So your 5700HQ will actually power throttle., as oppossed to never throttle, ever

The absence of power throttling is indeed interesting. But it doesn´t change that:

1 Most slim laptops overheat the CPU
2 Most HQ-CPUs can´t hold their clocks for more than 2-2.5 minutes, this one can.
3 You are obviously very attached to your X5. Even if you´ve tested yours 100 times it´s still just a single laptop. Please check the other subforums for info regarding throttling CPUs and general behaviour.

 

This result differs from your previous claim:

Also, why did you choose to have the graph drawn over the course of one hour? By doing so, you make it harder for us to verify that you saw zero frequency fluctuation from power throttling.

Could you please start the XTU stress test, then open HWInfo immediately after, and take a screenshot right before the test ends? If the reported minimum value is still 3492 MHz across all cores upon completion, I'll tip my hat to you. I also get a flath graph for the frequency, but HWInfo will pick up on frequency drops that occur for brief moments. I end up with at least one core not having sustained the turbo clock for the full 5 minutes. No core ever drops below 3GHz. The tiny bit of throttling I'm witnessing usually kicks in past the 2 minute mark, and I have plenty of thermal headroom.

 

At stock voltage it starts to Power throttle at the 2-minute mark, temps are good though:

Spoiler: Stock voltage

Undervolted -70 mV to 1.041 V it holds the full test, 3.491,7 on all cores. Temps are actually a little higher but I didn´t let it cool after the stock test so not completely fair comparison:

Spoiler: Undervolted

 

10 minutes straight
75c no throttling no change.

 

Hmm... Once again, I saw slight throttling on one core while undervolted and more severe throttling at stock voltages.

1st run, undervolted:

Spoiler

2nd run, stock voltages:

Spoiler

@stamar: Thanks for the latest upload. It's nice to see a second flat graph for the core frequency. You weren't able to replicate CPU overclocking without throttling though, were you? Not that it's a big deal. Seeing X5s maintining their turbo boost throughout stress testing is good enough.

 

I have yes but i am not interested in that project at the moment. I did last summer undervolt and overclock

 

I see your voltage is higher than mine. My stock is 1.110V and yours is around 1.144V. And I can undervolt to 1.041V while your lowest seem to be 1.072V. That´s probably why you still see slight power throttling. Now if you lost the silicon lottery or I and stamar won it I don´t know.

I did some game testing as well. GTA V and Crysis 3 now peaks at 82C and 89C respectively and are holding max clocks. BF4 still creeps up to 96C and thermal throttles though not as much.
I have updated the OP with this info.

Edit: I overclocked it to try and force a power throttle and mine does in XTU:

Spoiler: 3.7 GHz, undervolted & repasted

3.7 GHz, undervolted & repasted

 

My x5was actually stolen.

And as much as I did like it I'm going with alienware so good luck.

The biggest thing for me on a laptop is the screen. Alienware has a better screen for less at my time of purchase so I'm going for the 17r3.


A new x5 Would be so expensive .