Asus G73 Benchmark Thread

Scook, So I should assume that everything is all right with my system right? My 3dMark06 and PCmark05 scores also came out very very slightly lower. I compare all my scores to notebookreview.com's g73jh review. And one last question. Is it still possible for me to have a bad graphic card eventhough the machine got almost same 3dmark, PCmark scores as the review? There are no symptoms of having bad graphic card but i didnt play any games yet either, so should i assume that my whole system is good because wprime, 3dmark, and pcmark test scores are almost as good as the review??? Or some parts still may be corrupted??? Thank you for your time and patience.

 

Only a few consistant GSOD's will let you know you have a bad card.

 

The scores are not identical between comparable systems every time.....your system is fine

 

Hey Guys,

I was thinking is it worthwhile to do a little Overclocking to get better results while playing games? If so, what program should i get or is it to much of a risk to do it.

Rob

 

Could I by chance bother somebody on here to download and run the new FFXIV benchmark and post scores for this laptop? I am considering buying one but want to know how well it might run this game in particular. Thank you very much.

PS~ My currently laptop with 9800GS only scores 1798!! Yuck on the low 1280 x 720 resoltion.

 

Hey All,

Does anyone know if the tigerdirect A1 is still coming with the 1333 ram?

 

1) it doesn't really matter performance wise.
2)Probably, all of the newer G73JHs have come with 1333 ram.

 

Are these benchmarks represent true performance of the G73?
Just checking to see if the GTX480m is worth it with all its crazy watt use and heat.

 

Don't forget that the 480M was paired with an Intel Core i7 940 2.93GHz...4x more cpu points in 3DMark Vantage compaired to the 720QM. Bamm.

 

Futuremark will be off because the D900 uses a desktop processor. This will also affect gaming benchmarks just not as much. Also gaming numbers only show which is better/worse for that particular game. I'm sure they can find two games where the 5870 outscores the 480M. Take that into consideration with the fact that the 480M will cost more, produce more heat and use more power, it will be hard to justify getting the card.

 

Besides, that other review site states the 480M will be on par with a desktop 5770. The HD5870M with some overclocking can also come very close to the 5770, if not match it with some lucky units. With the 100W draw from 480W, OC'ing will be voltage limited since MXM max is 100W.

 

+18% performance
+100% heat and power draw (100W vs 50W)
And it's going to be way more expensive...

edit: THEY STUCK A STOCK G73 AGAINST A DESKTOP X58 CHIPSET HIGH END i7 SYSTEM FOR THE 480M.

HOW IS THAT AN EVEN REMOTELY FAIR COMPARISON?

edit2: 100W is the MXM limit so you're not overclocking the 480M at ALL

 

They could of at least given the poor G73 a 800/1100 boost

 

Hey guys,

For gaming with Bad Company 2 or FarCry2 Crysis etc. Is overclocking my BB G73 i7 720 worth it? Or overclocking my 5870 worth it?

I will consider it if it increases the performance, so could someone please give me some info and perhaps a link to the program that does it best?

Thanks!

RJ

 

Wow ok, how did you get almost 3ghz on the i720QM

The only overclock method I am familiar with right now is setfsb, I can get a BLK speed of 165 for about 1.8ghz but yours is still at the stock of 145? or is it just turbo boost at work with 1 core on that screenshot.

 

it's turbo boost. I can only get the BLK speed to 149 for 1.756 ghz.

 

My ASUS G73JH- RBBX0 Benchmarks: How does it look? Hey!... I beat a few of the top 3dm06 scores.

I have not used setfsb, I have use ASUS OC tool once and thought it froze and then uninstalled it months ago. I think the settings stayed for some reason and I say that because my CPU says 1.60GHz, but my RAM says 710MHz and 3dm06 reads the CPU @ 1.708ghz. I guess I've got the benefits of Direct Console without having a process running.

I am using Driver 10.6 and my OC is 800/1100. I pretty much crash at anything higher and sometimes at even that OC. I didn't get a good OC'er.








STALKER: CoP - Extreme setting 0AA/Low SSAO [I need to lighten up on the Sunshafts obviously]


[Here's a plug for Tuneup Utilities 2010] I re-ran the STALKER: CoP with Tuneup Utilities 2010 and it's turbo mode on. It was the only bench I did here with it off, it was on with all these other benchmarks. May I add that I keep my processes low to begin with, I was running 33 processes with it off and 35 with it on in Windows 7 x64. [FYI: I got my TU 2010 at Target for $20]






More Games Benchmarks Here.

 

More huge images.

 

well this is the 1st time i've pushed my system in the 2 months of owning it, afraid of pushing it further...

 

Ziddy's vBios 3, Catalyst 10.6 and stock clocks.





Its about 255 points better than my stock run when I received it. CPU was 100 points lower though.

 

I keep looking at the modded vbioses, but what do the offer?

I know one has lower voltage, which is nice, but I'm running full stock on 206 and it's never GSODed (unless I go overkill on OC)

 

http://forum.notebookreview.com/asus/492562-asus-vbios-v3-everything-but-perfect-powerplay.html

http://forum.notebookreview.com/asus/482491-gsod-fix-g73-owners-vbios-update-guide.html

 

Yeah I'd read those but it appears that the only thing I would gain would be less heat/power draw, and the laptop is working perfectly for now.

I guess I'll wait a bit longer to see if we can set automatic downclocking when not in use (2D/3D modes)

 

If you feel you don't need it, then don't.

 

Hello, i have a new record with my G73 in 3dmark06!!

 

........... That does not have me convinced in the slightest

 

You left out some stuff....like CPU-z and GPU-z......do those and people will believe you

 

wow!
looks like LOD is working over time in that benchmark.
around 3.0 to be more to the point.
2.0 looks a bit too low against a 12k 3.0 score....
not saying anything though....

 

wrong resolution anyway

 

damn, nice catch moo

 

Still a nice score for that res. Use an external monitor and run at the default 06 res.

 

Hi guys,
I have an A2 with 720QM. I am interested in overclocking my CPU as I use my G73 to run CFD simulations which utilise all 4 cores, hence any CPU speed gain would be benificial. Would SetFSB give me a noticeable speed gain?
Based on your experience, what is highest safe & stable FSB speed setting which is suitable for prolonged period (a simulation could run for a few days and i don't want to cook my beloved G73).
I'm a total noob at this so your help would be much appreciated.
Out of curiosity, how would you work out the CPU speed % gain based on the higher FSB speed?

 

160mhz fsb is what I leave mine at

 

my ASUS g73jh-a1

Gpu Clock 877MHz Memory 1200MHz

http://img36.imageshack.us/f/p9572.jpg/

temps are high because i don't use any laptop cooler and her in DR is more and less 30c

 

Hey mew, when you run a full load (try Cinebench 11.5 if you haven't) on the 160 fsb oc, what i7 turbo multiplier does it lock on, x12 or x13? Thanks.

I know that when you apply the Asus TT which raises the BCLK (aka FSB) to 142MHz instead of the normal 133MHz, you will loose a multiplier from the i7 and actually run at a lower speed than without. It also runs 5C hotter.
As you see, I have asked mew how the Set FSB software affects this multiplier.
You work out the CPU speed by multiplying the i7 turbo multiplier by your BCLK.
Asus TT Off is: 133MHz x 13 = 1.729GHz
Asus TT On is: 142MHz x 12 = 1.704GHz

If this SetFSB software does not cost you your x13 multiplier, then you won't be taking 1 step back to go 2 steps forward. What you should expect is a lot of heat for running full loads with an OC BCLK. You will be cooking it and possibly get no benefit, you will have to run tests.

Q

 

Thanks Q. I think i need to take a step back and get my head around the basics before I jump into the deep end.
Below is my understanding, please correct me where i'm wrong.

1- The 720QM is a quad core processor @ 1.6 GHz, which means each core is 1.6GHz thus total 'power' is 4x 1.6GHz = 6.4 GHz (in comparison to a single core processor)

2- Intel boost is a dynamic overclocking which OCs the active cores. With Intel boost CPU speed is 1600MHz + 1/1/6/9 * 133MHz for (4/3/2/1 active cores respectively). Thus:
with one core running, it operates as single core @ 1600+133*9 = 2.79GHz
with two cores running, it operates as a dual-core @ 1600+133*9 = 2.79GHz
with three cores running, it operates as a tri-core @ 1600+133*6 = 2.4GHz
with four cores running, it operates as a quad-core @ 1600+133*1 = 1.73GHz
so at full load, cpu speed is 1.73GHz which is (1600 + 133) or (133*12 +133) or (133*13) i.e. 133MHz FSB x13 multiplier.

3- ASUS TT is an ASUS software that overclocks CPU & RAM.
It increase FSB (or BCLK) to 142MHz but multiplier is 12 hence speed is 142*12=1.704GHz.
It seems that with ASUS TT your multiplier decreases from x13 to x12 because you lose the Intel boost. Assuming that's correct, would there be a a way to turn on ASUS TT and keep Intel boost working?

It will be interesting too see what mew reports back re multiplier when using SetFSB.
Mew, can you also check what Intel boost is doing when you use SetFSB? maybe look at the intel boost desktop gadget?

 

Wow Wal, I see someone has been reading Wikipedia

What I can tell you from experience is with a 133MHz bus (normal), the i7 720 will run:

1.73GHz on 4 core loads (presumably 3 core loads also)
*2.4GHz on 2 core loads
2.8GHz on a single load

* it was difficult to find a dual core load, honestly I don't remember what it was.

These results are in line with this: Intel Turbo Boost - Wikipedia, the free encyclopedia

The desktop gadget has never gone to 2.8GHz for me, it seems to top out at 2.67GHz for some reason. If you want to see what your turbo multipliers are doing in real time work, download the "Argus Monitor", great tool.

I also found this article to be informative when I researched the i7 mobile way back when and theres always Intels website. Intel Core i7 Mobile CPU (Clarksfield) Review - HotHardware

I don't think there is a workaround to get your x13 back with an increased BLCK, it's why I asked mew or anyone whom OCs the BCLK as to what their full load multiplier is. I bet it will be x12 and therefore one step back before however many steps forward.

 

Good old Wikipedia

So on full load, ASUS TT slows you down and runs hotter. That's no good.
How about on light load i.e. one or two active cores only?

FYI i use a program called MPICH to manage multi-core processing of my CFD runs. MPICH allows me to create any specific number of threads. so if i break up my calculation into 3 threads to run over 3 cores (assuming each thread runs on a core, not sure if that's what happens in reality?) , that would be the most efficient given that with intel boost 3 cores @ 2.4GHz = 7.2GHz as opposed to 4 cores @ 1.73GHz= 6.92GHz ???

 

Not correct, and it's not nearly that simple. There's no "7.2GHz" or "6.92GHz". You may want to do some research on how CPUs work.


How many threads to choose in your MPICH app can vary widely depending on what the code in the threads does.

Assuming all the threads are running the same code:

If the threads never have to wait for some resource another thread is using, then running one thread for each core will always get the most done regardless of Turbo Boost. If the code uses resources that can only be accessed by one thread at a time (like disk drives, network adapters, etc.) then it's possible performance will be degraded using a thread on each core so using less threads may perform better. So it's something that can only be determined by testing/benchmarking.


These Turbo Boost discussions are always way over-simplified. There's so much more going on. For example, there's always hundreds of threads running in Windows at any given time (even with no apps launched) - yes, most threads are idle most of the time but the OS thread scheduler is still running. So a Windows app is never going to get exclusive use of the cores (which is why it's real hard (or impossible) to get a single thread to push the Intel Turbo Boost to the max).

 

You can also set things like affinity and priority to influence the thread scheduler some

This could have interesting effects on Turbo Boost I think

 

Yes! We as programmers know that, but that means nothing for someone just configuring a 3rd-party app.

My main point was that the software has a lot to do with it.

 

With laptops, since overclocking them is so simple compared to a desktop, manipulating and tweaking software is about 80% of the battle haha

 

try shrinking the picture or hosting it somewhere and giving a link

or adding it as an attachment so we see a thumbnail

Good score but no one likes scrolling (I have a WUXGA screen and it is still too big)

 

Yes to heat and sorta to speed, I have examples of single and full loads using Cinebench 11.5 Part 1 & Part 2 with real results of a full load video rendering.
I don't know what all that 1/1/6/9 or those, i guess we'll call them effective GHz numbers you come up with are supposed to mean, but according to the examples I linked, a 3 core load would be the same as a 4 core load and run at 1.73GHz.

Q

 

So i decided to fork out the $5 'donation' for SetFSB.
the highest i could increase BCLK to was 144MHz, above that i was getting freezes or BSODs.

Interestingly, i did get x13 multiplier and my CPU speed was 1873MHz.
My score was 3.11 points in Cinebench 11.5, compared to 2.92 points at normal speed.
Average temperature increase was ~1C.

With ASUS TT on it got same score of 3.11 but temperature was 4C higher than at normal speed.

Btw, what is the best way to insert pics in a post?

 

You can upload them directly by using the attach feature, or use a hosting sight like photobucket or tinypic

 

Hi,
I like this thread. Three months ago I was bought G73jh too, and this thread was help me to decide between other notebooks. I am lucky because I don't have problem with BSOD, GSOD and other X-SOD only if I overclock CPU or GPU too high. My max values is 160Mhz on CPU with TT off (1924Mhz real frequency not the one from CPUID) and 840/1100 on GPU with v3 vbios. I wish to set higher clocks, but I've got the GSOD after a trial. I am overclocking with setFSB and MSI Afterburner.

There is no x13 multi on 720QM max is x12 no matter what (on four cores):
Intel® Core? i7-720QM Processor (6M Cache, 1.60 GHz) with SPEC Code(s) SLBLY

Vantage after OC
P9544
GPU - 9032
CPU - 11499

 

That is a pretty good vantage score, congrats!

 

Then everyone is in on the conspiracy except you

Q