Alienware 15 services woes

Most laptops dont. Its hidden by default. Needs an unlocked BIOS or an app like Throttlestop.

 

how can i check my timer resolution ?

 

http://www.mediafire.com/download/xzo9n84d8lze9nb/WinTimerTester_1.1.zip

There are other applications as well.

 

Nice to know.

Since the current AWs don't have the capability to use an unlocked BIOS then apps would be the only way.

 

Amen.

 

i am noob in this info XD i cant understand

 

Funnily enough, there's two people in this very thread who have spent quite a lot of energy stating that unlocked sBIOSes are unnecessary and not a loss of any concern for most users...

 

Who wants that extra bit of responsiveness. Not me, im fine with 30fps. /s

 

me ... i play FPS only and speed means live or death

 

Did Dell Encrypt the BIOS or something? I would actually like to try unlocking it if i can get the dumps.

 

Rapid aka use of memory to speed up your bench @tilleroftheearth can teach you about this Samsung Crap!!

@bloodhawk Dell lock their bios now like a virgin who lock absolutely everything!!
The previous socket machines had fully locked bios. Not much you could do. Only Slv7 managed to use a external bios chips with some success.

 

It's secure-flashed, and has been since the AW18.

If you can digitally sign your BIOS, or flash an unlocked BIOS to an empty chip and manually replace the BIOS chip, you can get a custom BIOS on it.

 

Ah thats a pain in the ass.
Im actually thinking about doing this with my system :


Will give me a quick switch with a BIOS chip dedicated for bench settings and one with daily use settings.

Source : http://blog.jacksoftlabs.com/tag/amibcp/

 

If bro @Mr. Fox still wanted to have his Aw18... I think he would soldered a new bios chip with unlocked bios like Slv7. The secured flash bios from DELLIENWARE crippled his machine. Max 330w wasn't his preferred Max load from a SLI machine. Dell crippled that model also like the new one. So nothing new from Dell!!!

 

Interesting. I didn't fully understand the diagram but the blog post gives me sufficient understanding. I'm not sure that would work for the AW models though, but you could give it a try if you had one.

Are you trying to use two Prema mods with different settings on your Clevo?

 

Yeap planning on. Need to dig up the schematics and check if this would work first.

One chip will hold high as hell settings and the other one daily use OC settings.

 

http://forum.notebookreview.com/threads/clevo-overclockers-lounge.788975/page-400#post-10293278

 

@judal57 use this
http://www.lucashale.com/timer-resolution/
it says that the free version won't do 0.5 ms resolution but it does... at least on mine
Use this to measure latency
DPC Latency Checker
PS: papusan is right ... rapid will just lock down a segment of your ram and dedicate it to caching data which u fetch from SSD more frequently, so it is just to increase benchmarks and is basically a very less flexible form of RAMDISK

 

I used set resolution on my old M11X R2, it helped a bit along with throtlestop. it doesn't affect much on my 6700HQ.. when are u getting ur liquild ultra ?? right now I don't want to change thermal paste, my thermals are pretty decent for now, and its not that hot these days in Bangalore. if ambient rises ill apply ICD7.. and hows the stress test going ?

 

Liquid ultra is available on amazon from booyahchicago reseller at arnd 2k..... and currently i downclocked to 3.1 ghz with -120 mv undervolt... temps at load while gaming are arnd 76-78... Intel stress test peaks for 2 hours at 81 degrees .... but if i revert to 3.3 ghz then temps blast to 90+ during gaming and stress test... This is all with my custom fan profile in hwinfo and thermal paste which was applied when i gave laptop to l3 team.
3.1 ghz means no current limit throttle in XTU but sadly performance drop in cinebench by a margin...
Can anyone post their cinebench results with i7 4710hq, I am only able to go to 580 with timer tweak(575 without timer tweak) and 7790 passmark

 

Yes I know its available, I went for ICD7 just to be on safe side.. I did a -100mv undervolt, and 34x on all cores, it was rock solid stable 3.4ghz with temps of 72c max XTU stress test.. I think you should stick with 3.1ghz, 81c is not bad with stock paste.. Early Fan Kickin gives around 3-4c difference on max temps..

 

Use Hwbot database http://hwbot.org/benchmark/cinebenc...Id=processor_3620&cores=4#start=0#interval=20

All bench http://hwbot.org/hardware/processor/core_i7_4710hq/

 

@Papusan thanks
@VICKYGAMEBOY there is a slight issue though..... temps "jump" around quite a lot
I just ran XTU benchmark nw(scored 637 i really don't know how the hell did downclocking increase my XTU score from 623 to this no clue how does this CPU work ).... the temps at 90 % load are arnd 71 but sometimes in the 20%-50% usage area temps suddenly from 71 degrees jump to 85-88 degrees. I don't get how can there be so much delta in temps
Cinebench score also varies.... i got 620 this time without timer tweak *_* at least closing the gap to a stock 3770

 

Wprime 1024M stress test is a fine real world test. Stressing a Hq processor +2.5 min
You can't get exactly same score each time. A small margin between bench scores is pretty normal.

 

580 to 620 is not a very small increase(around 8%), its a decent enough increase in cinebench....
and in XTU i never scored above 623 even if i ran it multiple times but maybe you are right it may just be random at least in case of XTU...
i ran XTU stress test for 2 hours, stays 90-100% load at all times with no clock drops nor any throttling VID didn't go above 0.9030, i think it was a good enough synthetic test to check system stability

I'll run wprime too and post results when i do

 

I had that issues as well on my 4720HQ, core spikes from no where, not sure if its the sensor fault or the stress testing software, but as far the temps are below 85c when stress testing its good.. Wprime Version 26.6 is better..

 

Although your hardware don't have stability problems in one test doesn't mean all is great.
Other bench can show you the total opposite. Some can have great stability with full load and stability problems in lower load. And different results depending on bench.
There isn't many test sites who use Xtu. If you want to check your hardware function correctly vs similar... You should use same bench tests like reviewers.

 

You might be unlucky and windows update was doing his thing. Caused lowered performance when running.

 

XTU Benchmark is very VERY reliant on RAM. People with 4700MQs have top score on HWBot with like 1600MHz cl6 and 1333MHz cl5 RAM. The more data your RAM can push in X amount of time the better that benchmark works. I would not use XTU's benchmark for any accurate representation of your CPU's performance being stable.

TSBench? Sure.
WPrime? Sure.
Cinebench? Sure.
3DMark's Physics test? Maybe.
Catzilla? Maybe.
Super PI? Maybe.
XTU Bench? Definitely not.

 

I ran fritz chess mark and wprime64 1024 ..... no downclocking/throttling of any form in both the tests and temps stayed below 82 .... only in XTU benchmark i see those sudden jumps in temperatures
@D2 Ultima @Papusan How is CL6 possible with 1600 mhz :O ??? i checked almost all RAMS from hyperx fury to vengeance non offer anything less than cl9..... And if its altering RAM timings then how can we do that in the echo models and test it so that the RAM doesn't heat up or stability isn't hurt. No option in bios to alter ram timings in echo models btw.
XTU has an option to alter ram timings but I don't know about a good ram test and generally never had any experience in altering ram timings or overclocking ram. Just i only know this much that the lower the cl timings the lesser clocking is possible and if u increase cl timings more then you can overclock ram more easily

 

One thing a lot of people miss is that the best stability test for Skylake is Video Encoding , specially H.264.

 

For everything, really. I have crashed more than one person's "benching stable" overclock by setting up a livestream to abuse their PC.

 

Oooh, never thought of using that. Thanks for the tip.
True but my Skylake Over clocks crash way more with encoding than my San Bridge-E ones did.

 

I was actually referring to Haswell-E overclocks specifically with the encoding. But yes, Sandy-E definitely is a stable overclocker once you get things set. Ivy-E as well.

 

How do i alter ram latency and clocks using a windows application and how do i check for stability of memory

 

Cant do it in Windows.

 

sad

 

I haven't even seen a desktop board that lets you do it while in the OS. So forget about it in Laptops, just not possible.

 

intel xtu let you ... but your pc is not going to boot anymore hahahaha i had to remove the bios battery coin and put it again to reset the settings

 

Woot? XTU lets you change memory timings within Windows?
But if you have to reboot to apply whats the point O_O

 

yes... i had to reboot and my pc never boot again XD, i had to reset the bios removing the coin cell

 

drat

 

So finally tomorrow i am going to update the OP with the final resolution from alienware regarding throttling ....... You need to customize it via downclocking, undervolting(no different than fiddling with a clevo with barebones bios just more restricted) to reduce throttling. In layman's terms what you are told is definitely not what you get with these people(mostly underperforming systems will be delivered) ..... Well played Dell

 

That's not a resolution. But it's exactly what people with these systems have to deal with.

However, don't fret! Remember what @Caladdon said! ALL YOU NEED IS BIOS A06 RIGHT? Didn't the Alienware techs tell you this?

Unless... that isn't a solution to the problem?

 

Wow - Someone just can't let it rest can they? Must absolutely have the last say, even after a previous thread was closed by moderators because it devolved into mindless trolling.

The BIOS update does work. I know because and I literally cannot make this any clearer for you ..... I AM USING IT!

Your opinion will not change this. Period.

My Alienware 15 has experienced NO throttling issues whatsoever in real time scenarios and I am enjoying it immensely.
And what really is the cherry on top is it's allegedly the "worst" offender too with the i7 4720HQ, GTX 980m and 180W power adapter.

I have zero and I mean ZERO interest in running damaging stress tests or anything that will put my system in a operational situation that simply WOULD NOT happen under real world usage.
I'm waiting for the stupidity of these tests to reach the stage where "users" complain that it can't work under water......

You are simply wasting your "Keyboard warrior" skills on me.
I'm satisfied with my system, like the incredible majority of gamers that purchased them. Most just don't invest the time arguing with narrow minded individuals online because they are too busy enjoying an awesome gaming experience.

However as expected, you take the issues that have been mentioned by the unfortunate few on this site and blow them completely out of proportion in a vain effort to amplify your opinion that these new systems suck.
Love to break this to you - they don't. They perform well, sold well, received enormous amounts of praise from various and credible reviewers.

If you're going to claim that Alienware have stated these systems suffered from issues then provide the documented evidence to do so and not falsified quotes. Do that or remain silent.
And even if you can provide that - It's not changing the reality of the situation for me and many others. Our amazing Alienwares work as intended.

In the meantime I both will and can provide useful first hand experience information on just how well it's performs, and let me tell you - It is amazing.

 

Yes, because your machine is everyone else's.

 

The machine being talked about in this topic is also not everyone's machine. BUt some users including yourself do act like that.

 

i managed to get 3,1Ghz stable, more than that (throttling)

 

Why is XTU saying 2 Cores active? Is it still ****ty that it cant even detect active cores properly?

 

i dont know .... they change things like throttling indicator ... now is "yes"no" before was in percentage