Alienware 14 765m Overclock Quesion

Hello,

I'm only asking this question because my Lenovo y580 Was Capped at +135 MHZ core by Bios while using MSI AfterBurner.

Is this the same case for the Alienware 14?

Thank you!

 

yes
I've tried using msi afterburner and evga precision....the same

 

confirmed, it is the same

 

Ok thanks for the replies

What are some overclocks ppl have gotten out of the 765m? Can +135 be done?

 

I run mine @1Ghz GPU / 1325Mhz ram

 

so you underclocked the vram?
The stock setting seems to be 2GHz

 

Gddr5 is multiplied x4, our gddr ram runs @1000 (4000Ghz effective), mine are @5300 MHz, wich make up a little bit for the 128bit bus width restriction.

 

Ok thank you, what aare defaults? 850 mhz? You manged to pass the 135 Mhz barrier?

 

Just wanted to jump in and say that if you plan to overclock make sure you always monitor the temperatures of your machine. Here is how

 

Will overclocking void warranty if voltages remain untouched?

 

And i would like to point out that AW with this little machine did an amazing job as far as GPU cooling performances

 

sacrificing the thickness and we get the best in class cooling
seriously I can't think of any laptops with similar form factor and hardware can even match the alienware
msi ge40? razer blade 14? clevo w230st? gigabyte p35k/p34g?
from what I've read they all seems to have some thermal problems
perhaps the rog g46 can do better but there is no haswell and 700 series refresh.
our machines remain the best

 

There is a haswell with 700m series graphics cards out the g750, it was out even when you posted your comment (about the time i got it actually lol)

 

lol it's not really fair to compare the alienware 14 to a much larger 17 inch machine
I was comparing it with laptops of similar size

 

For comparison's sake, and to give you an idea what Oc'ing will get you.

Clevo W230ST 13" - Oc'ed +135 GPU (1032mhz max turbo), +500 mem (6ghz effective), CPU undervolted -89mV = 4905 on 3Dmark11.

NVIDIA GeForce GTX 765M video card benchmark result - Intel Core i7-4700MQ,Notebook W230ST

Temps top out at 70c on the GPU, 90c on the CPU. No throttling ever observed on the CPU, even in BF3 (currently the most strenuous CPU+GPU load I can find) after playing 2 hours.

Laptop was $1039 with 13.3" IPS screen (beautiful, btw). Running 2 mSATA 128gb SSD's in RAID O, with a 750gb HDD for storage. Price did not include the SSD's, which I got later.

So , Yes, the Alienware has some competition, though I suspect the CPU will run cooler in the AW14.

The question is: Does it matter, since the CPU never throttles anyway on the W230ST? I guess you get to say your CPU runs cooler, but it doesn't translate into much in the way of tangible gains.

 

Hi There ... I am overclocking using MSI afterburner ... core clock +135/Mem clock +200 ... Im runnning it on discrete mode ... stock clocks work absolutely fine but theres a huge improvement in FPS while playing with this OC on Arkham Origins ...my temps go to 71c .. I would like to know how to push this OC further. My guess is the FPS will drop if i increase the OC as im already hitting 71c. Anyway i can reduce the temps ?? U mentioned undervolting ... did u use Intel Extreme Tuning Utitlity? Could you tell me hw to go about doing this ? and does this help while OC ?? Also, does nvidia inspector work on drivers 331.65 ?? Thanks in advance

 

There are many factors that need to be compared as well, such as build quality, tech support, although the CPU can do until 90ish before throttle it can still be uncomfortable to your fingers. Even topping at 75ish (CPU running prime95) and 62 for GPU for me, I feel the heat on the keys and arm rest, which really isn't to my liking on my alienware but it's miles better than the other laptops I've had.

 

Hi Kondor999

I too recently bought a Alienware 14. You mentioned you added 2 mSATAs, how? I though there's only place for 1? My system came with a 750GB HDD and a 80GB mSATA.

Isaac

 

Edit: update values from current tests:

I thought I'd start with putting this here even though it's an old thread.
Any help will be much appreciated, would like to get more out of my gtx765m if possible.

I've been doing basic OCing for years but new to vbios modding and OCing at this level (voltage etc)...
So, I'm not sure if I'm doing something wrong, but I've read several posts from reputable looking people that suggest they're getting around +100core/+1000memory OCing gtx765m but I can't seem to get anywhere near the +1000MHz memory, I'm getting artifacts from ~+300...

Using NVidia Inspector, machine: Alienware 14 p39g, gtx765m, svl7's vbios, using heaven 4.0 to benchmark/test.

Best stable under heavy load seems to be +100 core (higher than most it seems?), +300 memory (Default: 902MHz Core, 4008MHz Memory, 1000mV)

Both core/memory seemingly hold maybe for another 5% on either value (or with rare artifacts) but artifacts start to appear and get progressively worse. Can push the core a lot further with only occasional artifacts but pushing memory higher gets bad fast and eventually driver crashes start happening occasionally somewhere in low-mid +400s. I've had voltage up to +100mV during testing but pulled it right back to +25mV as that's all that seemed to be required to stop artifacts. In this gtx765m overclock video the guy reportedly gets +125/+400 without increasing voltage at all?

I don't know enough about OCing to have any idea why my results are apparently so much lower, less than 135 limit before vbios flash (had to flash regardless as I was getting throttling). Did I just lose 'the silicon lottery' as I've seen some people say? Do I need to undervolt my CPU? I've seen that mentioned but not sure why they're doing it and I *assumed* +100mV used in testing on GPU would be doable without downvolting CPU? Sorry if this is just a lot of senseless waffle, with inexperience I'm unsure if I'm doing something wrong or if I simply found the particular OC limit for my machine.

As an example by contrast, CptXabaras reports his Alienware 14 gtx765m gets:

TEMPS: stable around 65-67 degrees. I'm comfortable up to around low 80s so plenty of room there. Not really knowing what I'm doing with voltage but I didn't want to go higher than +100mV (1100mV) and above +50mV seemed to make no difference regardless. With room to move on temp and voltage, seemingly it's just the artifacting that's stopped me?

------------------------------------------

A thanks and a mention in case it helps others, stevie's below post lead me to the necessary vbios guides. Using NVidia inspector to set OC values and MSI afterburner for on screen display, in game minor tweaks and reporting.

 

Bumping this, found I was still getting occasional artifacts on above settings so now have brought it back to +90 core/ +270 memory. Any thoughts?

 

The Alienware 14 sucks for overclocking the GPU, it is a locked down piece of rubbish. The CPU on the other hand can be overclocked well.

What I hate most is Dell has imposed a thermal limit on the machine around the GPU, although it can be overclocked +135MHz and I know this as it was able to do it on one of the earlier BIOS, but quickly throttled really hard over 63c bring the clock speed right down to 400-600MHz.

In a later BIOS, it was decided behind closed doors to fix the issue, as was a problem without overclock also. What we got was a BIOS that allow normal speed to be fine without no throttling in and around 63c, but moment you attempted to overclock even a just few MHz above stock (10MHz was enough for this to happen), then the machine would kick down to 400-600MHz.

There is certainly something in the BIOS tell the machine to jam on the brakes if the speed is overclocked anything above stock. Even though I know the machine can handle it. Hell even a rubbish HP machine that isn't even a real gaming laptop platform does not do this kind of stuff.

Dell like to try and say you are going over your power limitation of the motherboard, but this is something they have actually doctored into the BIOS for their own piece of mind, more likely in order to lessen the few unit that will come back for warranty due to heat issue. It would be a small percent at that, but its all about profits once the machine has been sold, less than come back to Dell means more money lining their pockets.

 

well it's not happening on mine.
Mine is on A09 bios. It can sustain +135 with no thermal issue or throttling whatsoever
The gpu can go up to 70C and stay there. Don't think the 63C thermal limit exists anymore

and you are able to overclock your cpu? is it a extreme edition chip?
don't think my locked 4700MQ can overclock at all LOL

 

All Haswell i7 quad core can be overclocked, they all have a certain amount of multiplier bins open above there turbo

4700MQ = x2 bins unlocked

4800MQ = x4 bins unlocked

I am able to push the single core performance from 3.7GHz stock to 4.1GHz, same goes for the quad cores but obviously lower.

Download Intel XTU and start tweaking an extra 200MHz


As for the GPU:

Have you checked GPU-z to on the monitoring tab, that will sure enough tell you if the clock speed is being reduced, you might not be noticing it.
Pretty sure it is ok at 63c or higher with stock clocks which are fine. After overclocking that is another story, basically goes back to it old tricks pretty much after it goes over 63c.

 

Nope dude. I am pretty sure about it.
But for fairness, I did a furmark run, I overclocked with nvidia inspector, +135 MHz
temp. goes up to 70C. and the whole system is stable as hell
no drops in clockspeed whatsoever

 

No problem here as well, I can overclock it without ease and above 63 degrees as well. Though I'm running at stock clocks at the moment, since for me it is fast enough at those.

 

That pretty cool, there is not way I can run mine with furmark let alone just a normal game above stock GPU clocks, just will not allow me to do it.

I had nothing but bad service trying to source a HDD caddy from Dell, and not being able to tweak out the GPU for extra just plain sucks.

I have got rid of my Alienware 14, now I don't have to deal with this overweight and thick POS.

 

That problem just happened to me, while I was trying to do valley benchmark
The clock jumps to 405MHz and unable to recover unless reboot
Later after many "diagnosis", including tweaking nvidia settings, removing the battery....
It seemed to have something to do with multiple monitors setup, as I'm running off an external monitor thru hdmi

Using it's own monitor, the 405MHz issue doesn't seem to happen anymore, instead the driver just crashes
Yes it's a sad thing this machine cannot overclock

 

Very strange, valley benchmark is one of the best to test with as it does have a core clock and temperature present on screen. I know that it slowing down, as when the frames drop there is a counter, that aside I can see it with my own eyes. Wasn't just that, could see it else where in other places e.g. games. I never connected any other monitors,so rules that one out, just right on the unit plugged in. The Alienware 14 is one sweet well built machine, and it can obviously handle the heat with its excellent heat pipe array and decent (but loud) blower style cooler fan, shame about not being able to tweak much. For me I couldn't get a hold of a optical bay HDD caddy even though Dell has the parts, they just denied me from buying any of them, and having a GPU that cant be tweaked in such a spec machine is absurd.
What let the Alienware 14 down is its heavy weight and thickness. Once on a table is ok, but then end up a bit far away from the display unless you hunch down to it. Can be used on lap with a lap desk, but then are dealing with 3KG of weight, and the need to keep the underside clear. Most other manufactures are moving towards slim and light especially nearing the end of 2014.
I am happy to keep my Alienware M17x R3, that just does what its mean to do, a desktop replacement in an area where I cannot fit a full desktop. That thing never goes on my lap lol.