FutureMark: 675mx equivalent?

Hello,
im planning to buy a notebook and probably will be gaming on it, i know notebooks are not best for gaming but still 675mx is not a weak card, just below 680m in terms of performance.
I have been looking at futuremark scores, and according to 3dmark 11 score 675mx scores around 4100 and Desktop gtx 560 scores 4050. so is 675mx really equivalent to desktop 560 in terms of performance, thats rather weak eh..?
is it?

 

I'm not sure about benchmarks and synthetic tests. What I can attest to, however, is the real life performance of the card. Not a single game in my library plays at less than full HD settings. Only the most demanding games cause me to turn down AA, etc just one notch. Most of the games in my library are at least 1-2 years old though. The most demanding game I have right now is Metro 2033, everything on full hd and high, with some of the settings turned down one notch, and i get a solid 30fps. Stock clocks as well.

 

Can any mod shift this topic to other forum, probably Graphics and gaming one, couldn't figure out where this would go, i would like more comments

 

no idea on the 675 but the 680 scores P6196

where did you get the above scores from?

 

Best Graphics Cards March - 2013

 

thanks for link. interesting page from futuremark.

check this one out Mobile Graphics Cards - Benchmark List - NotebookCheck.net Tech

 

thanks for the link but notebookcheck only provides benchmarks for notebook cards, so i cant really compare 675mx to desktop 560, so my question still remains unanswered is 675mx really a equivalent of desktop 560 as per 3dmark 11 benchmark

 

tick the box at the top that says desktop and it will show both

 

Don't fully trust Futuremark 11. Its a very Shader bound test that strongly favors chips with massive shader arrays without much consideration for other factors. That being said, it is bloody good at measuring exactly that. The true performance of a Graphics card is the Shader power + Memory bandwidth + ROP/Texture power. I find Futuremark Vantage is a better all round representation of true performance as it is very sensitive to the Memory Bandwidth of the card.
Case example, my GT 650m GDDR3 version beats the GDDR5 version in 3Dmark11 because the Shaders are higher clocked therefore have more shading power. However, it is much slower than the GDDR5 version in most games by 30% due to the extremely low memory bandwidth. 3Dmark Vantage very accurately depicts this difference.

The GTX 675mx has the following specs:
1152 gigaflops of shading power
48.0 Mtexels of texturing power
19.2 Gpixels of ROP output
115 GB/s of memory bandwidth

The GTX 560
1088-1276 gigaflops of shading power
45.4-49.8 Mtexels of texture power
25.9 Gpixels of ROP output
128 GB/s of memory output

We can see here that the GT 675mx is indeed very similar to the GTX 560 but it is much weaker with ROP output and memory bandwidth which tells us it will take a bigger hit from MSAA than the GTX 560.

The GTX 680m has almost double the Shading and Texture power vs the 675mx with about the same ROP output as the GTX 560. However, it has the same memory bandwidth so it is actually quite bandwidth starved thus the performance is nowhere near twice of the 675mx. Compared to desktops, the performance will be about the GTX 560Ti at the worst and about the GTX 580 at the best.

 

thanks for some nice technical specs there, so then basically 675mx is equivalent to desktop 560, looking at what you said i can conclude overclocking the memory will help more in increasing the performance of the cards.

 

Sorry I didn't have a basic 560 on hand to test, but I did have a 560 TI OC please find desktop 560ti 3dmark11 GPU score at 4320 and Notebook 675MX at 4088. I would say the mobile 675mx is pretty much a standard desktop 560 Ti

 

Yes OCing the memory will help hugely with performance! Stick it up to 1250mhz and enjoy your new card