Clevo P151EM with 670M and 3920XM tested

I hope these have not been posted before

All credits for these benchmarks and screenshots goes to 慕容蛛蛛

 

Great) XM CPU + a little faster card the previous gen, a nice touch to the upcoming P151EM refresh!

 

For reference (assuming the 670M is legit):

2920XM w/ 570M gives P2690
NVIDIA GeForce GTX 570M video card benchmark result - Intel Core i7-2920XM Processor,Micro-Star International Co., Ltd. MS-1761 score: P2690 3DMarks

2920XM w/ 580M
Low: P3096 NVIDIA GeForce GTX 580M video card benchmark result - Intel Core i7-2920XM Processor,CLEVO P170HMx score: P3096 3DMarks
High: P4895 NVIDIA GeForce GTX 580M video card benchmark result - Intel Core i7-2920XM Processor,Alienware M17xR3 score: P4895 3DMarks

 

not bad but nothing amazing. the 06 score is not impressive since its mostly all CPU. a cpu priced at >$900.......

the real beast will be the 675. im hoping for gtx 570 desktop-like performance .

 

Real beast will be the 680.... not the 675. 675 is simply a 580m..

 

yea but when is the date for release? i don't consider it within reach and not worth judging yet. thats like saying 7990 is the real beast. no date for it as far as i know. if 675 > 7970 amd is screwed.

 

Ryan is simply saying that the 675M is literally a GTX 580M, just flashed to show a different name. There's no point in hoping more more performance than what you already see.

AMD would have to do something pretty stupid, to keep the 7970M from demolishing the 675M by a vast margin.

 

What is also interesting is the CPU score of the Ivy bridge 3920XM with 3dmark06. According to notebookcheck the 2920XM scores in average 6079 points. 3920XM scored 7060. So that is a 16% increase.

 

WHich translates into nothing worth getting excited about when it comes to how real world applications will perform. There will be no real noticeable differences.

 

To be fair, 3920xm is 400mhz increase in frequency, so.............. it's not an architecture win for clock cycle performance.

 

Okay, so the 3920XM runs at:

3800 MHz (1 core)
3700 MHz (2 cores)
3600 MHz (3 or 4 cores)

The 2960XM (using this as an example because it has the highest mobile sandy bridge cpu stock clocks) runs at:

3700 MHz (1 core)
3600 MHz (2 cores)
3400 MHz (3 or 4 cores)

And at these stock clocks it gets 6820 on 3DMark06 whilst being 200Mhz less on 4 cores.... clock for clock performance is the same?

 

The clock for clock performance difference should be around 5%. Atleast thats what they found out when they tested out the desktop Ivy

 

I'll try running my XM at the same clocks as the 3920XM, but even before doing that I can guarantee it's not 7% for the mobile chips

 

Yeah do that. Looking forward to see the result

 

Here you go ^^

This is a 2920XM running the same clocks/multiplier as a 3920XM:



Very close, Ivy Bridge is almost 1.5% faster? SWEEET! (sarcasm)

The CPU score is 6955 and that score probably beats all QM Ivy Bridge Chips...

 

lol 1.5% difference didn`t see that one coming.

I am pretty shure we will see bigger clock for clock difference than that since the recent tests with Ivy showed 8% better frames in games and different software also saw around the same improvement with the same clocks with Sandy vs Ivy.
Anandtech saw 5-15% better performance with 3770k compared to 2600k, with 3770k clocking 100MHz faster.

Plus not to mention that notebook CPUs get around 10% higher clocks on our Ivys

 

Are the higher-end 600 series parts going to be Kepler or further optimized Fermi cores?

 

Up to GT 635M - Fermi
GT 640M/GT 650M - Kepler
GTX 660M - Most likely Kepler
GTX 670M - Fermi
GTX 675M - Fermi
GTX 680M - Kepler

 

ZZZZzzzzzzzzzzzzzz, probably see 8% better framerates because of the HD4000 integrated graphics core and vs. the old HD3000. But again WHO CARES? Power users arent going to see jack for differences, nor are they going to be worried about the performance of a cheap on board video processor. I mean 8% is worthless. Hell 15 or 20% isnt anything to write home about either. :SLEEP: :SLEEP: If you have a Sandy bridge i7 this very mild upgrade isnt worth keeping tabs on(unless your just a benchmark junkie).........its going to be a few years time before something comes out thats worthwhile upgrading to which is capable of giving in the 50-100% range improvement at the very least.

 

No not more FPS with the IGP, but with a discrete GPU. This is almost clock for clock (look at 2600k vs 3770k) and you already get 8-10FPS more in average. 400 MHz difference like the mobile CPUs are gonna be even bigger.
AnandTech - The Ivy Bridge Preview: Core i7 3770K Tested


If you have SECOND revision of Sandy Bridge, I totally agree that this upgrade probably isn`t worth it, unless you plan on using Optimus since its a big improvement with the IGP.

However if you have the first SB CPUs, 2720QM, 2820QM, 2920XM, 2630QM etc, its going to be worth it in my opinion since that is where we have the 400MHz difference. That is my guess, but we will have to leave it up to the reviews to show if I am right or wrong.
Who knows, mobile CPUs may scale different than desktop CPUs

 

you don't think that since the new boards are backawards compatible with SB that we'd be able to use optimus with a SB iGPU?

 

HD 4000 is better, don't think it matters much though. Looks like the only thing worth looking forwards is GTX+optimus?

 

Yup.. And backlit keys if you're going to a P150/P170 model