Earlier this week I posted a short preview of the newly announced NVIDIA GeForce 200M mobile GPUs.  At the time (and in reality still today) we don’t have a lot of information about the GPUs themselves, the technology on them, changes in architecture, etc.  We have been able to gather some more information in the last few days though about the power consumption and performance these parts will offer and how appealing they may or may not before system builders.

NVIDIA GeForce 200M GPU Update - Power numbers and efficiency - Mobile  5

One the driving points behind NVIDIA’s marketing push during this announcement was the key phrase “Twice the performance at half the battery life.”   Wow, who can argue with that right?  They are lofty claims though and marketeer speak doesn’t always live up to the claims.  By simply referencing some old data on our own site and looking through some previous documentation on past mobility GPU offerings, I decide to put together a few visual aids to help demonstrate NVIDIA’s GF 200M chips.

NVIDIA GeForce 200M GPU Update - Power numbers and efficiency - Mobile  6
* The TDP result for the GTS 150M is an estimate

Here is a graph showing us the raw computing power of the current generation of GeForce 200M mobility GPUs and the previous/current GeForce 100-series of mobility GPUs.  Clearly the GTX 280M and 260M offer the most power you can get in a mobile platform from NVIDIA and that the newly announced GTS 260M and 250M GPUs outpace even the top end 100-series part, the GTS 160M. 

NVIDIA GeForce 200M GPU Update - Power numbers and efficiency - Mobile  7
* The TDP result for the GTS 150M is an estimate

This slide shows us the maximum TDP – thermal design power – for each of those NVIDIA mobility GPUs as well.  Now you can see how it is those GTX parts can show such incredible performance – they eat a LOT of power.  But notice the very steep drop off in power consumption to the GTS 260M; it uses just about half of the power of the GTX 260M and only 57% of the power of the GTS 160M.  I bet you can guess what graph comes next…?

NVIDIA GeForce 200M GPU Update - Power numbers and efficiency - Mobile  8
* The TDP result for the GTS 150M is an estimate

The performance per watt is probably the most important feature of a mobility GPU and the new GF 200M series definitely improves on that point.  Notice that the sweet spot seems to be the GTS 250M as it achieves 12.86 GFLOPs per watt of power used compared to the 7.49 GFLOPs per watt of the GTX 280M.  That essentially makes the GTS 250M 72% more efficient than the GTX 280M – that doesn’t mean the GTS 250M is a better PERFORMER, after all the GTX 280M reaches 56% higher GFLOPS ratings, but in terms of power efficiency for mobile gaming, the GTS 250M wins that battle. 

The rest of the new 200M series is impresses as the GTS 260M bests all the other GPUs in this comparison.  Even the GT 250M and GT 230M are proving to be more efficient than the entire GT 100-series lineup and basically show-up the GTX 200M line as well. 

Now, of course, all of this data is specifications provided by NVIDIA itself through various sources, so we have to take the results with a grain of salt (we have lots of salt by now in this office) until we get our hands on finished product.  If NVIDIA’s PR team is being truthful about the “100 design wins” for the new GF 200M parts then the products more than likely live up to the hype as OEMs need proof before betting the farm on any mobility product.