Ivy Bridge Mobile 3920XM TESTED!!!

The 6990M score can also be found in the article.

Intel's Fastest 22nm Mobile Ivy Bridge Chip "Core i7-3920XM" Tested

 

That's a GPU there.

 

Here is the 3DMark06 score for HD3000:

http://www.notebookcheck.net/Review-Intel-HD-Graphics-3000-graphics-solution.43710.0.html

 

Only 6995 for CPU score? I am not impressed, here is my CPU(2630qm) score in 3dmark06. So the fastest IB CPU(3920xm) outperformed the lowest SB quad(2630qm) by 36.9%. Guess this is what we can expect from a die shrink.

 

You forgot that he ran this chip at stock without oc whatsoever?

 

Are you comparing 6970M with the IGP (HD 4000) lidow?

EDIT: Or is it only using HD 3000?

 

That too! I was tempted to say but not so sure since i dont use 3dmark 06

 

Of course I am not comparing hd4000 with 6970M, I was just referring to the CPU and I was fully aware it wasn't overclocked. I was meaning that 6995 doesn't seem that impressive for the CPU score consider 3920xm being the fastest IB CPU.

Don't care the IB graphic performance personally, won't be using them for gaming or whatever.

Edit: the pic I posted was just to show the CPU score of a 2630qm in 3dmark06, and it's done with 6970M obviously, no HD3000 can ever score that stupidly high.

 

That`s what I thought. The IGP is very important for me though since I plan to buy a laptop with switchable graphics. And from the looks of it, it seems like a nice boost although this is not the final chip ready for retail

 

I'd rather see 3DMark Vantage and 11 results. More relevant than 3DMark06 IMHO. I'm aiming for an IB + possibly Radeon 7xxx GPU middle of this year. Unless AMD Trinity proves to be as powerful as early reports indicate, plus probably a lot cheaper.

 

Source:

Translation result for http://dell.benyouhui.it168.com/thread-1898781-1-1.html

 

Clock for clock, IB CPU is pretty much the same as SB. The range just gets a big boost in integrated graphics, modest clockspeed increases, and reduced power consumption.

Have to wait for Haswell, which will most likely be end of 2013, to get your 6 core mobile processors.

 

HD 3000 with 2860QM got 4781 points with 3DMark06. 2860QM is probably very close to 3920XM since it is slightly faster than 2920XM.
HD 4000 is 68% faster
Intel HD Graphics 3000 - Notebookcheck.net Tech

GT 540M with 2820QM got 8162 points with 3DMark06. HD 4000 with 3920XM is almost identical. How much the CPU does to the score I don`t know, but it would have been very interesting to take away the extra CPU score and be able to compare the GPU score of those two.
Review LG A520-T.AE31G (3D/2820QM) Notebook - Notebookcheck.net Reviews

 

Guys you trying to compare new CPUs and GPUs
But in reality these improvements are very little
For instance :
here are new ivy bridge laptop CPUs : it's XM QM and M
Intel's mobile Ivy Bridge CPU line-up revealed by VR-Zone.com

So how big difference is btw i7 2820QM and i7 3820QM ?
or 2920XM and 3920XM ?
Bear in mind they have still same wattage consumption only few hertz on top ..
And diff. btw Intel HD 3000 and 4000 isn't frightening .

Thing is it's useless to look on IT stuff every year (or maby every 6months these days) but every 3 or 4 years
THEN you might feel some significant boost
Buy your machine and keep it for 4 years then start thinking about change
But companies Intel AMD or nvidia they must go on all the time ..

 

That's nice if the GPU is comparable to the 540M. I hope we can see a similar jump in performance with Haswell, maybe something equivalent to the then modern "x"50M series, for example. If we do, I would potentially say I don't need a dedicated GPU.

 

its great that the integrated GPU in the 3920xm is much improved...however, any laptop sporting an extreme processor probably has a top of the line discrete GPu or a professional card like a quadro or firepro.

Regardless, i really hope the 3920xmis compatible with my mobo lol..

 

It won't be. its already been stated that Ivy bridge on the mobile platform will NOT be compatible with HM65 and HM67 chipsets due to a new mobile socket.

 

It's at least pin compatible. The guy in the original source actually tested it on his Alienware and said it wouldn't boot due to lack of BIOS support and also that Intel denies compatibility...let's see...

 

Well, isn't the HD4000 going to be common across all Ivy Bridge CPU's?

 

i don't know if they will put in the older less powerful HD3000 in the cheaper ivy chips like i3's and i5's, but across the core i7's Hd4000 should be common. I was just stating that on this particular chip, an extreme processor, the integrated chip wouldn't be as important because these highend chips would inevitably be matched with a good discrete graphics solution

 

It's all HD4000, but my problem with Intel's implementation of the integrated graphics is that the HD4000 comes in at a variety of differing clock speeds. The highest rated chips, as in the ones most likely to get paired discreet graphics and thus have no need for much IGP performance, have the fastest GPU's, while the lower end models have slower ones. Of course with IB the difference is much smaller than the stupid difference with SB, but still, no reason for it.

 

Well I don`t know. I have never used Optimus before but if I can casual play older games without firing up the 75W GPU, I will be satisfied. Both for a cooler laptop and longer battery time when traveling and gaming at the airport or on the train etc etc. The more games with IGP the better imo. Bet Nvidia and AMD have made their NGC/Kepler GPUs better than previous generation at drawing less when idling

 

This is a good point. It's good to have options though. Especially for Ultrabooks and smaller laptops.

 

Actually, in the SB line-up, the HD 3000 chips don't vary too much in speed.

2960XM: 650MHz-1.3GHz
2310M: 650MHz-1.1GHz

Intel® Core? i3-2310M Processor (3M Cache, 2.10 GHz)

Intel® Core? i7-2960XM Processor Extreme Edition (8M Cache, 2.70 GHz)

 

I'm confused. The 3dMark06 score for the HD4000 iGPU is better than, say, an ATI 5650 HD that I have in my laptop. So what is it about Intel graphics that is so bad? Drivers?

On paper that iGPU should be amazing. I haven't used any Sandy Bridge CPUs so I don't know much about that HD3000 performs in games.

 

3dMark06 is not much of a GPU test these days. It is more of a system wide benchmark. GPU's have gotten much more powerful relative to CPU's since 2005 when it was released. 3dMark06 can in no way be used to determine relative performance of graphics cards if the same processor is not being used, and even then it is often bottlenecked by the processor.

 

I can't wait to test one in my notebook, with a bios microcode update/mod.

 

Why are integrated graphics useful if a laptop has Xfire or sli GPUs? Does the CPU work with the GPU to increase FPS at all?

 

It's for extending battery life. I believe the M17x and/or M18x have an integrated GPU that allows them to last 3.5 hours on battery. Without the integrated GPU they would last 1-1.5 hours.

 

Meh, nothing to get excited about, though.

Comparing the i7-2960xm to the i7-3920xm, they both have the same Level 3 cache at 8MB. Both are the same price at $1096 and the 3920xm is only 200 megahertz faster than the 2960xm at stock speeds. Other than that, I don't see much difference between Ivy Bridge and Sandy Bridge. Not impressed.

According to Intel's Tick-Tock schedule, Haswell will be pretty much the same thing as Ivy Bridge, just a different code name.

 

It's more fair to compare the 3920XM to the 2920XM. There will more than likely be an IB revision down the line just like there was for SB (2630QM -> 2670QM, 2920XM -> 2960XM, etc.)

No.....Haswell is a "tock" that will have architectural changes, thus improving clock-per-clock improvement just like Sandy Bridge versus Westmere. You can expect Haswell to be 10-20% faster than Ivy Bridge per clock. The "ticks", like Ivy Bridge, have basically the same architecture as the previous "tock" and thus don't bring much performance improvements other than increased clock speeds.

 

for me I will be excited that either the new chip give extra (unused) performance or the 2920XM and 2960XM will be cheaper :x

 

Now, that's a major disappointment! IGP aside identical performance to my 2960xm stock vs stock! Just because Intel doesn't feel threatened now, doesn't mean they have to stop moving I'm definitely skipping the IB...

 

IB's strong points (if you can call them that) will be in lower power consumption, slightly higher clock speed per core, and about 50% more powerful IGP.

So, yeah, ultimately, it's nothing to get overtly excited about.
The longer battery life gains will be minute when one takes into account that screens are largest power suckers of all components.
And if you are one of those who is using a discrete gpu, then most of IB advantages won't be really relevant.

I was disappointed that Haswell series will still be 22nm.
Lol...
They should have further reduced the manuf. process with that series if you ask me.

 

Agreed - we're not going to see much in terms of advancement with battery life until we get a new battery technology. We've been using li-ion for how many years now?

 

A couple of decades (at the very least).

 

And then there's this: Maximum PC | New Breakthrough Paves Way for 10X Better Battery Life, Charge Rate

“ We have found a way to extend a new lithium-ion battery’s charge life by 10 times,” said Professor Harold H. Kung. “Even after 150 charges, which would be one year or more of operation, the battery is still five times more effective than lithium-ion batteries on the market today.”

Kung’s research team has combined two techniques to combat both these problems. First, to stabilize the silicon in order to maintain maximum charge capacity, they sandwiched clusters of silicon between the graphene sheets. This allowed for a greater number of lithium atoms in the electrode while utilizing the flexibility of graphene sheets to accommodate the volume changes of silicon during use.

Kung’s team also used a chemical oxidation process to create miniscule holes (10 to 20 nanometers) in the graphene sheets -- termed “in-plane defects” -- so the lithium ions would have a “shortcut” into the anode and be stored there by reaction with silicon. This reduced the time it takes the battery to recharge by up to 10 times."

--------

But no word on manufacturability and timing if it will even be a reality.

 

Research success does not necessarily mean commercial viability, unfortunately.

 

I am happy I won't need to upgrade from the 2920XM.

Sandy Bridge is pretty boss anyway, Intel did a great job there.

 

Thing is we will only be able to charge the battery at the rate of the power adaptor lol.

Maybe notebooks would come with a quick charge station that can quick charge the battery so you dont have to carry around a 500W brick lol.

 

That bit of research is at least 5 years old.
Graphene is a relatively old concept and it's superiority over commonly used materials in computers was known for over a decade.

While I will agree that research and market availability are 2 different things... one has to keep in mind that our capability of using available research in practice within a year or 2 (at most) from the time of when it first was available, is viable.
We have the industrial capacity, the tech and resources to do it, but there's no 'rush' to bring it to the market.
It's better to gauge people some more for more and more money in the meantime.
Quite frankly... that research should have been a practical reality over a decade ago.
And personally, I'm not a fan of fuel-based batteries concept either.
It's virtually akin to cars and gasoline - effectively the fuel will be an extra expense in the long run (and for what... only 10 days of battery life? - lol, I can't wait to see just how much cash people will be spending to refuel their batteries every 10 days if that thing becomes a reality).

There's no 'conspiracy' at play here though... merely common interest (and it's quite frankly getting sick/old).

 

yeah even overclocking wise 2920XM and 2960XM is in the same...limit? (dont know what word to use). Ivy Bridge only have smaller die and better IGP (which I have absolutely no interest in).

Well I just got a 2920XM OEM for $550, not sure if it is a right decision :-s

 

that's not true.

if your power adapter can output say 2 amps for duration of 1 hour, when unplugged your battery will not be able to provide same 2 amps for 1 hour. There are losses when you charge the battery due to the chemistry and/or heat involved. So indeed you could charge your new chemical type battery with the same adapter for less time while still being able to get the same result from the battery itself after the adapter is unplugged.

 

What I mean is say you improve both the capacity and charge rate of a battery by 10x.

So my battery would go from 74Wh to 740Wh, and the charge rate would go from 50W to 500W.

My power brick cant supply 500W, only 150-180 (depending on the brick) so that would mean with the machine off I could only charge 150Watts per hour. Extending charging times. To get round this you either include a 500W brick (and reinforce the charging circuitry by a huge amount) or a 500W charging station.

 

well in this case if you use same 150W adapter then for one hour you could charge your old battery up to 50Wh due to the charging limit of the battery, while you could charge your new battery up to 150Wh due to the adapter supply limit.

end result is you get 3 times more juice in your new battery for the same charging time and with the same adapter used.

now if you want to maximize the potential of your new battery then you buy yourself a 500W adapter, but you still get improvement with the adapter that you have.

just saying ..

P.S. consider laptop being off at the charging time, for clarity of the example used.

EDIT: sorry for the somewhat distant discussion from the original topic

 

Yes but if you want the full charge to take full advantage you are looking at a longer time.

Thats why I talk about a charging dock, if you intend to put these batteries in smaller machines, 150W is uncommon, its usually more like 65-90W.

 

The 2920XM and the 2960XM are the exact same chip, the only difference is the stock multipliers.

It's gonna be interesting to see how far you can push the beast, the only limit is temperatures, as usual.

IB will probably run a few degrees cooler at most, and from the benchmarks we have seen (albeit barely any ).. same clock for clock speeds.

 

I used to write news stories about that stuff but stopped some time ago. Those rumors rarely ever come to anything. "Oh we can make batteries last 200x times as long!" Sorry, not interested unless you can get it to market. Otherwise it's worth as much as me making a blind wish.

The only battery advancement of the last few years that came to fruition is the 3-yr batteries that last more charge cycles. And those have the compromise of less capacity vs. the same-size battery.
Fuel cells were demoed by Toshiba in 2006 but never took off.

 

Where does everyone find these awesome OEM XM prices?!

I went with a 2920XM ES, cause that was the the only sub $600 one I could find. I would snap any 29x0XM OEM for $550 up in a second!

 

Upgrading microcodes will not be enough be enough. Also what will you do about the lack of suppport of the HD4000 in the old BIOS? That will be difficult to upgrade I guess.