MonotorBuggy :
JackNaylorPE #2 is confusing to me. Can you explain again?
Nvidia releases a reference card design, then its AIB partners have the option to improve on that card's design. For example, the 1070 and 1080 reference designs all throttle. The AIB cards don't. The most glaring instance of this is the RX 480 which exceeded rated power with the reference design, most of the AIB cards did not.
In past generations the difference in performance between reference designs was rather apparent. The better coolers provided much lower temps and lower noise. The better PC componentry also led to significant difference in performance, most noticeably with the VRMs as the VRM temp can limit OCs when the GPU isn't that hot.. This first became apparent with the 570 series where peeps with reference PCBs were blowing their VRMs left and right. On the 560 Ti, Asus, MSI and Gigabyte all had 6 or 7 phase VRMS while the reference cards (and EVGA) stuck with 4. The ones with the cheaper VRMs could not compete with those other cards having better quality VRMs w/ more phases.
This continued generation to generation but was really brought to bear in this 970 roundup by bit tech
http://www.bit-tech.net/hardware/graphics/2014/09/19/nvidia-geforce-gtx-970-review/1
If you go thru the review, you see that after tearing down all the cards, the reviewer took pains to see who improved various components and who "cheaped out". If you look at pages 2-4 of the article, you will note the various differences in cooling, componentry (chokes, VRM phases, fan control, etc) that you can pretty much predict the order of finish in the performance testing results before getting off the pages w/ component descriptions. They highlighted
Examining the PCB reveals a 4+2 phase power design – four phases near the rear I/O for the GPU, and two in the bottom right corner for the memory. This is a slight upgrade from the 4+1 stock specification but unlike MSI [6+2 phases] and ASUS [6 phases], EVGA does not use any specially crafted components.
The GPU MOSFETs are directly cooled by the main heatsink, which has a thermal strip on to draw heat up into the fin stack. On the other side of the GPU is a metal contact plate that partially cools two of the four memory chips on this side, leaving the other two exposed. It also cools the MOSFETs of the power phases serving the memory, but no thermal pads are used, so heat transfer is likely to be limited.
EVGA repeated this blunder with the 1070 / 1080 designs leading to performance issues, black screens and even fried PCBs
http://www.tweaktown.com/news/54774/evga-geforce-gtx-1080-ftw-catches-fire-video/index.html
To their credit, they addressed the issue and will send you a thermal pad kit which you can install to correct the problem in about an hour. To be fair, i must emphasize that these problems were usually limited to the SC line but this time the FTW line was also missing the thermal pads.
So in short, all the major manufacturers give you something extra with their AIB cards... this may include better coolers, more PCB based features / components like passive fan control, better chokes, VRMs, better cooling (thermal pads) and even hand picked GPUs. With the SC series, they give you a different cooler and everything else comes up a bit short of the competition...and usually has an impact on performance. Unfortunately Boost 3 kinda treats everybody the same and, to a large extent, nerfs the advantages in performance that these improvements could otherwise provide.
Now if I'm looking at two of anything in the same price range ... and one delivers "less", which one ya gonna take ?
MonotorBuggy :
I'm trying to keep cost down relative to electricity use, MSI has one 1050 that only uses the rails for power no PSU plugs, your opinion? Remember I'm not a gamer I and play them.
just want to download HD movies
Well in that case, I'll have to withdraw the 470 recommendation as it pulls 133 versus 77 for the 1050. But keep in mind, like a PSU, a card only draws what it needs. This is a quote from a 1050 Ti review
https://www.techpowerup.com/reviews/MSI/GTX_1050_Ti_Gaming_X/25.html
During gaming, power draw stays well below the 75 W limit of the PCI-Express slot. MSI's added 6-pin power connector seems to do very little here. Only in Furmark do we see the card reach 75 W, so with additional overclocking and power-limit adjustments, the card would go over the slot limit, which kinda justifies the inclusion of the 6-pin.
Now this is a "bigger card" than use and we see that it has an extra connector for those extremely rare instances where you might draw more than the slots rated power. But as you can see, during gaming the bigger card gets no where near 75 watts.
Here's the 2 GB 1050
https://www.techpowerup.com/reviews/MSI/GTX_1050_Gaming_X/25.html
So whether the connector / cable is there or not, the card is only going to draw 67 watts in gaming