***The Official MSI GT80 Titan Owner's Lounge***

What's the battery life like on this big boy while using the iGPU? I can't seem to find those run-times anywhere.

 

I've gotten similar battery life with iGPU and the sli gtxs. Obviously not gaming. Just browsing or doing office work. On average you get 2 hours with a max of 3 if you do very light tasks.

 

Guys I am confused by all of the latest posts. So does this laptop thermal throttle or not?

 

All computers will have throttling, its a safety feature built into the CPU/GPU once you hit a certain temp.

As long as you are not purposefully trying to make it throttle as the previous post have, then you should not have any issues with it under normal use.

 

It doesn't not experience thermal throttle and it is not plagued of thermal throttle issues, I have never seen thermal throttle show up in logs ever. But as said above all cpus have the potential for protection if they reach high temps.

 

Hi guys,
I am looking to add a stick of RAM for my MSI GT80, what would you guys recommend?

Thanks for your reply.

 

No thermal throttling whatsoever on these models as of yet. A lot of people keep confusing power throttle with thermal throttle, because of the design operation ofr 47W and how turbo actually works.

Neither the 009 nor the 001 variations have shown thermal throttle.

 

You should check the make/model of the preinstalled RAM and see if you can find a match. You can either check in the BIOS or by physically inspecting it.

 

You better add 2 sticks because you want to have the dual-channel benefits.
I would certainly go with 1866MHz DDR3L (1.35V) SO-DIMM.

 

So if I play BF4 with the 4980HQ for like 4 hours straight you are saying that it will not thermal throttle and hold its Turbo at 4GHZ the whole time?

 

Read my comment. You have to learn to distinguish thermal throttle and power throttle. The CPUs were designed for 47W in mind. You can confirm with ALL screenshots shown with burning tests and all, that intel 4980HQ initially consumes around 80W and then it regulates down to 47W just like every other intel CPU with TDP 47W. Only MX extreme editions can remove that limit, so you can consume how much power you want as long as temps are under control.

As I have seen from users with 009, intel 4980HQ seems to maintain around 3.8Ghz all around when tortured to the max, so you will most likely be playing between 3.8ghz and 4ghz for hours with temps on the 80s tops.

And as I provided other links before, it won't matter with BF4 nor will it give you any additional performance gain worth mentioning going over 3.6ghz of the 4720HQ at all.

 

So would you recommend just buying the cheaper Titan? What would be the reason to get the 4980HQ vs 4720HQ?

 

I think what you are asking about is power throttling and not thermal throttling. Normally if the hq cpu is maxed at load it will power throttle as its only allowed to consume so much power. So it will not hold its 4 core turbo setting for extended periods prob 45 sec at max.

But I do play BF4 from time to time and if the game is set with vsync for 60fps on ultra it will hold the clocks because it doesn't consume enough power and it will never thermal throttle either unless you have a ventilation issue.

If you are running 120fps it will power throttle on and off due to much more cpu power being needed to process the extra frames. Hopefully this answers your question and you have a better idea of each property.

Also at 4ghz you would be over clocking as stock is 3.8ghz at 4 cores, 3.9ghz on 3 or 2 coes and 4.0ghz on 1 core. BF4 runs at 3.8ghz as it makes use of 4 cpus most of the time.

 

Here are some benefits:
4980HQ vs. 4720HQ:
-32% higher on 8-threads "CPU Mark"[1] (10060/7639)
-27% higher on single thread "CPU Mark"[1] (2248/1771)
-115% higher on 3DMark 11 GPU[2] (1957/910.5)
-71% higher FPS on Unigine Heaven 2.1[2] (26.1/15.3)
-3 times more overclock on Intel XTU (600MHz/200MHz)
-Intel vPro Technology
-Intel Trusted Execution Technology

NOTES:
-Same TDP 47W
-$245 more expensive

[1] http://www.passmark.com/products/pt.htm
[2] http://www.notebookcheck.net/Mobile-Graphics-Cards-Benchmark-List.844.0.html

Note that all graphic benchmarks are of the IGP (Iris Pro 5200 vs. HD 4600) not the GPUs (970M vs. 980M).

 

So you are saying that for 245 dollars it's a no brainer to get the 4980HQ?

 

Thank you that helps a lot.

So let me clarify.

Will the Titan with the 4980HQ maintain 3.8 Hz the entire time playing BF4 for like 4 hours?

Also while the CPU is at 3.8 Hz are the 980M's being fully utilized to their maximum potential?

Because if this is the case I will probably get the Titan over the Sager NP9377-S simply because of the Sound and Super Raid technology.

 

I have never logged for 4 hours that's a pretty difficult question you are looking to be answered. Regardless BF4 at 60fps doesn't need but about 3ghz and only at times to run so either way both cpus are going to give the same performance.

Both cpus max out the 980 gpus, testing between the two really the only difference comes in with cpu needs such as physics and shading. If you are gaming both cpus are overkill (in your question), I have used both the 001 and 009 and you just can't go wrong here unless you already know exactly what you need.

 

That's a very good point thank you but I meant if you turned off vsync and ran it for 120fps will it maintain 3.8hz for a long time?

 

Re-read post 813 above, it will run into power throttle (not thermal throttle) due to the amount of frames being served and clocks will bounce averaging 3.5ghz or so most likely.

 

I see. So does this bother you?

I mean if you built a desktop you could probably overclocked CPU at 4ghz turbo continually with proper cooling and run BF4 at 120 FPS without any power limit throttling.

Is this of no concern with laptops today because this is the best we can get today and the best way they can be designed in 2015 so we have to deal with the best we can get?

 

I don't play any game at 120fps and I depend upon the built in monitor so it is of no concern to me. I purchased the Titan for the complete powerful package.

Sure a desktop could hold what ever you throw at it with the right cpu but I am just not a desktop person, I enjoy my mobile freedom.

 

I just want to comment on this, these are incorrect. I have posted my scores on passmark a couple pages back, and Intel 4720HQ scores 9400+ on the CPU, and single thread rating around 2k if I remember. I will dig my scores on my laptop when I get back.

The rest are useless metrics because power limit prevents extra OC, and 3dmark/uniengine on integrated graphics are a useless metric for a high end gaming machine like this.

The real difference between CPUs is 10% or less performance.

And my passmark link is here and you can confirm it is mine because of the system MS 1812 is GT80.

 

Again.... you will hardly increase FPS by an extra 100-200mhz over BF4. You can well run the game without issues as it is, and SLI works wonderfully. The desktop sure is more capable, but it is bigger, consumes more power and is quite a lot less portable for anything.

 

The values were correct. It's just outdated.
CPU Mark is a moving average of all reported results. That means it changes over time with every new results submission.
The values I've presented are from last month.
The current values are:

CPUMark-4980x4720_20150228 by Cormogram posted Feb 27, 2015 at 6:22 PM

That gives (as of today) an increase of 28% on 8-threads and 20% on single thread "CPU Mark"[1,2].
Again, this is a comparison of CPUs. The metrics presented here are usefull to help us decide how different CPUs perform and compare to each other.
Who needs a faster CPU it's not for me to tell.
Considering the GT80 CPU is not upgradable and that the 4720HQ IGP is a lot slower than the Iris Pro 5200 I think the price difference is well worth the performance gain.
Please define "real difference". I don 't know that metric.

[1] http://www.cpubenchmark.net/cpu.php?cpu=Intel Core i7-4720HQ @ 2.60GHz&id=2448
[2] http://www.cpubenchmark.net/cpu.php?cpu=Intel Core i7-4980HQ @ 2.80GHz&id=2327

 

That's the thing, laptops are more personal, so really it comes down to each person if that IGP performance is worth it or that extra potential MHz.

 

If by "no brainer" you mean worth it yes I do.
The Iris Pro 5200 (the 4980HQ IGP) is good enough to run many games with decent quality, such as Dota 2 on ultra @ 1080p with 40.7fps[1].
Note the prices that I've mentioned are the RCP by Intel[2].

[1] http://www.notebookcheck.net/Intel-Iris-Pro-Graphics-5200.90965.0.html
[2] http://ark.intel.com/compare/78934,83503

 

Agreed, many do not switch to IGP unless mobile on battery, most do not game under these conditions, and truthfully (though capable), this notebook (and most power gaming notebooks) is not the best choice for mobile application. Also, when plugged-in to the PA (where battery life is not so much a consideration), most will use the dedicated GPU's.

For the budget minded, the base processor is still a choice, and the notebook will still game awesomely. Under this circumstance, I'd recommend folks save a few pennies, and enjoy the 4720HQ. It's one thing to promote the top processor, and another what folks can afford. Personally, I do not see the IGP alone, as reason to opt the 4980HQ.

 

This is why MSI are not just offering the 4980HQ, the 4720HQ is not a bad chip.

If you clock it to 3.6ghz it would stand toe to toe with the top 3xxx generation chips or the 4800/4810MQ chips at stock.

 

What is the point of comparing the CPUs of lower end or other notebooks if we are using here the GT80s? Both 4980HQ and 4720HQ score higher than those old averages.

GT80 holds the highest score for CPU on both accounts, with 4980HQ scoring 11000 and 10900, and 4720HQ scoring 9400. The more powerful CPU represents a 17% higher score. this is the reality, and other scores are meaningless because we are not running on acers, lenovos or hps with limited cooling. You need to know how a particular CPU performs on the notebook you want. Several machines use these CPUs that might not have our adequate cooling, and thus throttle. This is the reason why on average the 4720hq has lower score than even 4710hq, despite being the same CPU except higher clocks for us hahaha.

It is fine if you want to game on IGP. Personally, it is a waste of battery and power, but to each their own. In that regard, iris pro might be a worthwhile upgrade. For most, it will be the extra performance it brings via increased core clocks.

But my point was gamingwise, you will see very little difference in performance when using the dedicated GPU, between both processor offerings. The 4980hq will always give better performance because of its increased turbo power though.

It will be interesting to try gaming with the dedicated gpus, while on battery. Maxwell is very good at limiting performance when on battery, and maybe they will give a playable game without issues.

 

First let's take a look of what really is been measured by CPU Mark:

As always many factors may interfere with the benchmark. CPUs are designed to be interrupted and in Windows it happens a lot. So the following recommendations apply:

Particular to the CPU tests there are environmental variables (such as temperature) and internal settings (such as clock frequency and voltage tunings) that may affect the results.

Since so many factors may not produce the best results and certain discipline is required to achieve them chances are there are many not-best results submitted to the PerformanceTest database.

Also, we must consider the influence of a not-best result to the average since for the subject CPUs (4720HQ and 4980HQ) there are only 32 and 16 samples in the database.

Worth to mention that PassMark does not provide the standard deviation of the measurements, limiting any possible statistical analysis of such results.

That been said the metric "CPU Mark" is not the perfect benchmark.

I think everybody in the IT business knows that.

Although, it's a pretty good and wide available metric for CPU performance comparison.

Off course, having a direct benchmark made in the system of interest is better than the average result. Thanks for those numbers BTW.

Note though that the variation of your results to the CPU Marks of the 4720HQ[1] and 4980HQ[2] are of 18% and 8%.

That's a big difference specially to the 4720HQ.

Can you run the CPU test with 10 iterations again?

You may set it on "Edit > Preferences... > Test iterations".

[1] http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-4720HQ+@+2.60GHz&id=2448
[2] http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-4980HQ+@+2.80GHz&id=2327

 

So what does all this technical jargon translate into layman's terms?

The 4980HQ is 10% better and worth the 300 bucks?

 

It's more than 10% faster and less than $300 more expensive, but yes it is to me.

 

I wonder when everyone is going to get the GT80 Titan 009 back in stock?

 

It is not worth to get gt80 with 4980hq. If you play games you want see any difference if you play benches you will see more points in 3d mark howerver it produce more heat and as you know it is soldered to mainboard. I am 10000% sure that 4980hq will be more defective than 4720.

 

You are saying that it runs hotter and will cause more problems and that it is better to get the other model the 4720HQ?

 

all GT80 models comes as a backpack to carry it?
I saw that the 3 GT80 resellers only , Gentech has no option for a spare power source.

 

@grinna
Nop mine was without backpack. I am searching it .
@GACollegeStudent
I prefere 4720qm. If it is socket version I would use i7 4940mx for sure . It is soldered to mainboard so I think there will be more problems with 4980 than with 4720 it is 100000% sure.

 

Was notified on Friday that my 009 is still not at the build shop and is still has no defined arrival date, I am now to the point of watching other members of this lounge and build a list of settings/tweaks/benchmarks to run once I have mine here.

Thank you for keeping the conversations intriguing. To that end what about this post?
http://www.tomshardware.com/news/microsoft-directx12-amd-nvidia,28606.html
Specifically the following 2 points:

1. "Part of this new feature set that aids multi-GPU configurations is that the frame buffers (GPU memory) won't necessarily need to be mirrored anymore. In older APIs, in order to benefit from multiple GPUs, you'd have the two work together, each one rendering an alternate frame (AFR). This required both to have all of the texture and geometry data in their frame buffers, meaning that despite having two cards with 4 GB of memory, you'd still only have a 4 GB frame buffer.
DirectX 12 will remove the 4 + 4 = 4 idea and will work with a new frame rendering method called SFR, which stands for Split Frame Rendering. Developers will be able to manually, or automatically, divide the texture and geometry data between the GPUs, and all of the GPUs can then work together to work on each frame. Each GPU will then work on a specific portion of the screen, with the number of portions being equivalent to the number of GPUs installed."

2."Laptops with dual-graphics solutions, or systems running an APU and a GPU will be able to benefit too. DirectX 12's aggregation will allow GPUs to work together that today would be completely mismatched,"

The first may be a chance for the SLI setup on the GT80 line to gain a good boost from the 8+8 gig format if we can optimize SFR(Split Frame Rendering)

The second may allow( for benefit or drawback) access to the IGP, therefore potentially making the decision of processor more important of a choice.

 

Like what kind of problems, can you please elaborate.

Because tons of people have the 3299.00 Titan in stock and I can pick that one up at any time.

The 3699.00 is always out of stock and seems more popular.

 

Hello all, I've purchased the gt80 titan sli 009 from HIDevolution with some extra upgrades, like adding more rams to have 32GB or ram and to have 4 ssd m.2 raid0 or as msi call it raid3 512GB - I worked and tried gaming on this laptop for about 4 Days and died without any warnings I was playing BTF4 and it was running smoothly especially that I was turning on the fan boost from time to another. the temperature was low about 70 for the gpu and 65 for the cpu. I was playing games just for test because I bought this this laptop for my multimedia work basically. Since I'm a former animator and video editor. I sent it back after paying the shipping and the customs extra .. they promised me to give a new unit. But now I'm afraid if this problem would happen again. I was reading your reviews and every buyer review. I think no one has the same problem I had. any advice?

 

the fans were running and the keyboard led light were on but nothing on the screen. they authorized me to reset the bios cimos but it didn't work.

 

I turned on the laptop, loaded nothing, only started the 10 iterations. My score was actually higher haha, overall it was now 5K. Since this score is withing 5% of my current score, or so the passmark website said, it could not be uploaded and since I have the trial version, it doesn't let me do anything with it.

My CPU score went up to 9605.
Single thread 2230

We need the results of other GT80s users, since those are the relevant scores.

The 4980HQ runs faster, specially in short bursts where it will have even more advantage. Such advantage will be more noticeable in software that is very sensible to mhz difference. For general gaming, the performance different is very minor, compared to things that require every mhz there is. Even then, it is not THAT large of a difference. This isn't core i3 to core i7. If you want the best performance, 4980hq might be the way to go. But 4720hq is more than capable and should be good on its own for anything.

Yeah I wonder why. I do know there were more 001 units available than 009, so either they produced very few units, or it is more popular. But at the price it comes, I am more inclined to believe they simply produced less 009s. I was considering, initially, the 4980HQ myself, but between the price, TDP and availability, I went with 001.

It could be possible, but I am not sure. I mean, I have old machines with CPUs that run hot, even to this day, with over 90 degrees temps, that after 4 or 5 years, still run fine. So I don't think they will die any time soon or degrade that much. Most of the time, the GT80 runs much cooler than what these torture tests make you believe. My average temps on a day are around 70s haha. I am sure 4980hq is a similar story

I know GentechPC and many other resellers here do come with backpack. Mine did

As gaming improves, the burden will lie more on the GPU, so even if the CPUs are soldered here, they will perform admirably for years to come

Sorry to hear that, I am not sure what happened there. I hope they sort you out soon and get your unit back fully working, or get you a new one that performs as you need The reality is that every model laptop in the world will always have a lemon or two, i have had bad laptops that died on the first day as well. It sucks but it happens.

Let us know how things go!

 

I am thinking about using this laptop professionally by DJing, do you think the extra power will help with these types of applications?

 

When THIS situation gets resolved.

 

Yep, it's a pretty bad situation.

 

There are many retailers with the 4980HQ in Germany and Austria.
I've talked to some of them and they had it in stock last week.
Are you in the US? I think some of them can ship there.

 

Not sure if you use Pro Tools but that kind of application will certainly benefit from the 4980HQ.
I think the GT80 it's one of the few laptops that can run Pro Tools comfortably[1].

[1] http://avid.force.com/pkb/articles/...ro-Tools-HD-11-System-Requirements&popup=true

 

That makes the variation even greater: 21% (9605/7951). The 4720HQ have a much bigger dispersion than the other CPUs.
Looks like it's because of the few samples collected so far -- probably most of them are not-best results, or perhaps there's a problem with the data.
Anyway, considering the numbers you provided (more reliable) the profit is 15% (11000/9605).

 

The 4980HQ should actually have even higher short turbo boost performance because you can use intel xtu to add up to 4.4ghz quad so at least for 30 seconds it would add even more performance.

I bet a tweaked cpu could reach interesting results.

As for the 4720hq results yeah you can see the submissions and some are lenovo that probably were using balanced performance or something.

 

It would be better if they separate the results by the clock frequency the test was made. That would help a lot!
Are there any BIOS mods for the GT80 that can release the kraken?