New Apple MacBook Pros outbench my 940xm

So I just read on Engadget that they tested the new MacBook Pro that was released today using Geekbench and it scored 9647 points. My M17x with a 940XM using throttlestop only managed 8300. Are the SandyBridge processors really that much better?!

 

Therese so little detail here to base any judgement on anything. Can you post a link? Points in what? What CPU? What benchmark? What hardware(GPU)?

 

MacBook Pro (early 2011) with ThunderBolt hands-on -- Engadget

I can only assume it is the 2720QM as Apple does not spell it out for us.

 

Meh, I don't know anything about geek bench, but I doubt a 2720QM will outbench a 940XM. Maybe at stock sure, but not overclocked as SB isn't TDP unlocked, that I know of. At least not yet. Plus I don't trust ch engadget says. They just repeat stuff on thier site that they read on other forums, and put in little effort in doing thier own research.

 

I don't know much about Geekbench either, but it appears to be designed to bench the processor and memory which is why I was a little surprised that it outscored mine. I have my 940xm @ 3% overclock plus the TDP set at 86w in throttlestop.

UPDATE: Just for fun, set my TDP to 90w and hit 8794.

 

That geekbench means nothing for comparing a gaming PC. I assure you your M17X R2 will get higher FPS on all games then that Macbook pro.

 

Oh, I don't doubt that for a second and I am also not saying I am selling my R2 for a new MacBook either, but I am surprised to see the Sandy Bridge parts bench so high. Don't get me wrong, I know they are supposed to be improved over the Arrandale processors, but I didn't realize the jump would be quite that much.

 

Macbooks will never come close to alienware in gaming..
Rest assured

 

Well the 2820QM seems to score the same as a 940XM at stock clocks if not higher on most CPU benches. Yeah the improvements are great.

 

As mentioned above - Geekbench is CPU benchmark and yes - the 2720qm / 2820qm's should be more powerful (at stock) than 940xm.

The 940xm may be more powerful when OC'ed (haven't read anything though) but remember, these are both new architecture (typically faster clock-for-clock) and a shrink (940xm = 774m transistors // 2720qm = ~1bn transistors)

So in short - yes it should outbench your CPU, however your laptop should have faster graphics. 1GB GDDR5 6750 sounds great, but would expect it to fall near or under 5830 level

Sandy bridge mobile quads are benchmarking great and can beat desktop quads to a point, here's to seeing what Ivy Bridge brings!

 

I fully expected them to be faster, but it was a bit shocking that a $300 CPU was 15% faster than a $1000 CPU in a single generation.

 

That's why everyone is going crazy over Sandybridge architecture

 

The 6750 mobility is about as powerful as a 5770 mobility. The 6750 mobility doesn't even make the chart compared to CF 5870's or 4870's or even a single 5870 or 4870 for that matter.

 

How does that CPU score next to a nicely overclocked 940XM? And he's right, in a gaming situation the R2 would walk all over it.

 

He probably meant "Is Sandy Bridge really that fast" but not "Is Macbook Pro better than Alienware" LOL

 

how did they manage to power and cool a 45w processor with the same 85w adapter and cooling they used for the 25w dualcore chips????????

the dualcore macbook pros were not exactly cool. this thing must run mid 90c or more during prime.

 

That is exactly what I meant. I think I was just floored by the benchmark score that a MacBook Pro put out as they are typically not speed demons and yet it just smacked by new R2 all around. I still wouldn't buy one

 

Define nicely overclocked. I cranked the TDP up in Throttlestop and had all 4 cores cranking at 3.5ghz with HT running at 100% and still only scored 8760 in the same benchmark.

 

There must be some cooling changes somewhere. The motherboard is new and I am sure the cooling solution is as well.

 

The biggest difference is the power consumption. If you look at the desktop SB numbers here, the 3.3GHZ quad not only beats the six-core 980X at times, but it does it using less power than the 2.66 i5-750.

That all means that the new chips are probably using much more turbo boost at 45TDP than the previous gen was. You get zero boost running 8 threads on a 720QM, but I bet the 2720QM still gives atleast a couple bins above the base clock.

 

Are you monitoring temps? Chances are your hitting 95* and the CPU is throttling back.

 

First of all I bet you are running the 32bit version and they are running the 64bit
Second, if you perform a retention mod to your CPU and apply the thermal paste correctly, you should be able to run TS (at least) with TDP 94 and a multiplier of 26x and the temperatures to max out at 79 (considering you are about 27 C ambient) and score at least 9113 (that's my score with the 32bit version without pushing my cpu to the limits)


I won't pay though $20 just for the 64bit version to find out how much better the i7 940XM can do vs the new mac... (if you are, PM me the email with the serial number necessary for registration).

btw the 940XM with retention mod and TS kills even the 2920XM...
wprime v1.55 1024M
stamatisx's 3min 49sec 164ms wPrime 1024m run with Core i7 940XM @ 3579MHz

superpi 1M
940XM vs 2920XM
9.935 vs 11

http://forum.notebookreview.com/har...grades/545462-2720qm-vs-2820qm-vs-2920xm.html

http://hwbot.org/community/submission/1056672_stamatisx_superpi_core_i7_940xm_9sec_953ms

 

FYI, even a 2630QM beats the crap out of 940XM in ALL benchmark! I was shock but that's that.

http://forum.notebookreview.com/hardware-components-aftermarket-upgrades/556663-really.html

 

Do you mean benching 940XM maxed out, next to stock 2920XM? Just stating the obvious its not exactly fair! multiplier on 2920XM is unlocked, beside using smaller 32nm tech delivers both higher OC'ing potential, and less heat.

940XM is still pretty sweet, though the industry do move fast.

 

Are you happy that your OCed 940xm can beat a stock 2920xm?

 

What do you mean by fair? We don't know for sure that the 2920XM will be able to take advantage of ThrottleStop and the OCing capabilities will depend on the OEM (remember the 5% BIOS overclock of R2? what makes you think that they will give you more with the 2920XM???) and remain to be seen, which means that for the time being, the 2920XM, performance wise is not worth the $1000+ because you can do better with almost half the money...
That a CPU has its multipliers unlocked means nothing, you will need the appropriate power to support them, the appropriate cooling and find a way to adjust the TDP (ThrottleStop).
I think the scores that I provided are good enough to support my sayings.
So, when we'll see the extend of its OCing capabilities then we will have a new conversation...

But I agree the industry does move fast indeed

not just happy, very happy and very satisfied as well, cause so far -which is more than a year- I still haven't seen better performance from a mobile CPU (overclocked or stock)

FYI TS does miracles with the 940XM. (just sayin...)
Btw, I am not advising you to check at Hwbot.org because the shock will be even bigger

 

The fact is:
The new SB quads outbench your 940xm

However the statement
"The New Apple MacBook Pros outbench my Alienware M17x R2" is debatable.

CPU wise MBP wins. GPU wise M17x R2 wins.

 

Show me your numbers and I will show you mine

and you based that on geeckbench... ok...

 

and to close this thread
M17x R2
940XM OCed (not external cooling or anything else, which means that the score goes higher...)
score:10582 (cause sometimes people are blind)
Alienware M17x : Geekbench Result Browser

when the 2920XM will hit this number PM...

the number was a courtesy of ThrottleStop

 

Nice results stamatisx. I thought your 940XM running 64 bit Geekbench code might show an improvement in performance.

The 2920XM has the potential to be the future king but at the moment, I haven't seen a single review where it is compared to a fully overclocked 940XM. It is easy for the guys writing the bios to decide the maximum performance level of a 2920XM. If the 2920XM can not be fully unlocked, the 940XM might remain king until the next generation of CPU is released.

 

I think everyone is still a bit taken aback by the SB jump in performance, but it is unusual for such a leap in the industry, this just all shows how the business is catering more to mobility than before, I doubt the next generation will be as big a leap.

Sure, I'd love to get anything close to SB levels with my M17x-R2, but without any foreknowledge of the real power of SB I don't think anyone should regard their selection as now "weak"; these systems still do pretty darn well for virtually everything despite weaker benchmarks. Heck, the R1's are still very decent rigs for gaming purposes and most everything else even.

 

SB was a big leap in the areas that matter most to mobility - performance per watt. On the desktop side, the improvements weren't as profound - at least not until the high end line comes out for SB - but it's a huge deal for mobility, where every watt of TDP is important.

 

Hey unclewebb , don't make ThrottleStop for 2920XM , then 940XM will be the king , but I know probably someone out there will....

 

Good idea but it's too late for that. I've already added Sandy Bridge support to ThrottleStop.

The problem is that Intel has built these new CPUs so laptop manufacturers have the final say when it comes to how much you will be able to overclock them. If the bios sets and locks the maximum turbo multiplier or maximum turbo TDP/TDC settings, there is nothing that ThrottleStop or any other software is going to be able to do about that.

We will have to wait and see if any laptop manufacturer has the guts to release a bios that lets a 2920XM run at full speed the way any 920XM or 940XM can run.

 

Given the need to prevent people from using thier warranty for frying their CPU, I don't see that happening. It's all about profit, not performance for these guys. They want to have top numbers at stock, but thats about. I think they are going to put the lock down on the 2920XM. My opinion, but I think they will to reduce the warranty pay out.

 

Yup. This guy knows what he's talking about!