Msi or Asus Gtx 760

Airforce11

Honorable
May 24, 2013
7
0
10,510
I planning to build a gaming rig(Im not an expert in Building Rigs) My Specs is FX 8320, Mobo - Asus 970X, (Mobo Manufacture is Mobo) my budget for Mobo is Inr 8000 maximum Inr 9000. Which Manufacture GPU shall i go Asus GTX 760 Direct CU or MSI Twin Frozr GTX 760. Is there is any motherboard for INR 9000 with USB 3.0 & Other Latest Updates.
 
Solution

AwesomeFX

Honorable
Oct 19, 2013
199
0
10,710




Oh please Chris, stop the fanboy train. ASUS and MSI are cutting costs by fooling their customers in various ways.

I'm gonna do some research for you, to get you out of your fanboy bubble. Please watch this video and see how ASUS and MSI are fooling us:

http://www.youtube.com/watch?v=zDxFbAhu4Bo


You to OP.
 
Sep 22, 2013
482
0
10,810


Fanboy? I have a Gigabyte card.

I had 2 7870s recently, a Sapphire and Asus. The Asus with the DCII ran a full 10C cooler. I don't know what your f&$@&ng problem is, and I don't care but speak from experience and do some reading,

If the OP is choosing between only these 2 cards, I'd suggest the Asus specifically because of the DCII. It's quiet and effective.

If he's open to other options, my "fanboy" goes out to Gigabyte. The Windforce is the quietest, most effective GPU cooler I've used.

Note that there's plent of info right here on Tom's in reviews to substantiate my opinions.

Attack me on the forums again with a flame post and I'll report you're 12yr old comments to a mod.
 
Sep 22, 2013
482
0
10,810


Please also note that your video has absolutely nothing to do with the OPs question. I am an EE and I can also tell you that using a different form of voltage regulation doesn't mean they're "fooling" us.

You get what you pay for. If you actually research the voltage regulation used on higher end board you'll see the more expensive boards tend to use more power phases. This is effectively a built-in power conditioner (which you can, and should use externally anyway).

Small, low-voltage electronics don't usually have extremely tight voltage regulations, but once you get into overclocking you need a predictable voltage variance, ideally as small as possible. Multiple phases allow the voltage variance to be smoothed and re-smoothed through each additional stage.

However, to build a board with more power phases costs more money, both in R&D to design around the added components and in manufacturing itself.

In addition, all electrical components are rated for a certain level of tolerance, whether its what they can handle in voltage variance, or what they will allow in voltage (essentially how accurate their marked rated voltage is). These parts become more expensive the more accurate they are. The normal rating levels are stated as +/-10, 5, 1, and 0.5%, with high end parts sometimes getting as small as 0.10% increments (very expensive).

So there is no "fooling" going on. You're a fool if you think you'll get +/-1% tolerance on a board with a +/-5% tolerance price.

You want great voltage regulation? Buy a great board.
 

AwesomeFX

Honorable
Oct 19, 2013
199
0
10,710


That's the essence of my post, the boards that claims to have 8 but really only has 4 are sold at the same price as Gigabyte boards with 8 true. (Usually 150-250$ range)
 
Sep 22, 2013
482
0
10,810


Voltage regulators are not the only component to consider in this price range, but yes, you're right that there are some with a various level of "true" phases that are available in that price range.

However, the guy in the video doesn't consider a couple of key factors:

1) The combined cost vs. performance ratio of the parts needed to create a true 8 stage vs. a true 4 or 6 stage can be significantly different.

2) Just using parts to create a 4/6/8 stage voltage regulation stage doesn't mean that you can use all of the other same parts and be just fine. ICs are built with specific Vrr, Vpp, etc., specific current ratings, etc. which in turn affects the other components in your design. Change the mosfet IC and you may effectively change the entire design.

3) There are other ways of regulating how clean the voltage signal is. A significant amount of AC noise can be eliminated using a) A decent power conditioner before your PSU and b) a quality PSU. You can also build a 4/6 phase voltage regulator that is maybe 50-75% the cost of an 8 phase where you've added or configured the circuit to adapt to AC noise and result in a lower cost with the same, or nearly the same performance.

4) CPUs have a fairly low tolerance for AC noise. If the 4/6 stages that were using what he calls a doubler (which is really probably a NOR or NAND gate combined w/a voltage amplifier rather than a dedicated mosfet) were so horribly noisy the unfiltered harmonics would result in serious issues and wouldn't be suitable for a CPU, but instead we simply see that better boards tend to have better stability for overclocking.

5) Your window for AC noise, and thus voltage variance (aka tolerance) gets smaller the higher you raise CPU speed due to the fact that CPUs rated speeds are based on a given set of wider tolerances than your OC'd speed. In other words, they're guaranteed stable at X% of tolerance taking into account the poor voltage regulation on OEM (Dell, etc.) boards and inconsistency across the industry. The higher you raise the frequency from the rated frequency, the smaller the tolerance becomes.

6) Cost is a factor, but part availability is a major factor, too. Let's say a manufacturer designs a new board and is about to launch it. They've invested time in R&D and marketing and have the PCBs being pressed. Suddenly the IC they used is no longer available. It is EXTREMELY expensive in some cases to replace this part. You can't just swap it for an equivalent in many cases, because the results will vary. A good example is Ibanez swapping out the JCM386 for a different IC in their classic Tube Screamer TS-808 pedals. People went so far as to mod the TS-9 back to a JCM386 because while the new IC was on paper the same thing, it didn't perform the same and didn't sound the same.

Gigabyte may indeed have more true power phases, but there's a lot to be considered with how other manufacturers *utilize* the less expensive methods. If the manufacturer is CLAIMING 8 *true* phases, but doesn't actually provide this, that's another story.

Even 4 true phases w/8 simulated phases is cleaner than 4 phases alone, so a 24 phase voltage regulator with only 8 true stages will still have less AC noise that an 8 true-phase solution, if it's implemented correctly.
 

AwesomeFX

Honorable
Oct 19, 2013
199
0
10,710


Smawesomeness!

Thanks for the incredibly detailed reply, I'm studying electronics and find your post very interesting :)
 
Sep 22, 2013
482
0
10,810


No problem. And you're right to be skeptical as manufacturers can definitely be misleading with their spec sheets but if you know what you're looking for and do a bit of research, you'll find there are quite a few boards that have only 4/6 phases that still do a very good job of voltage regulation, but there are certainly some who don't.

I was very on the fence about buying my ASRock Z77 Ext 4 until I saw some detailed reviews checking the Vdroop and other voltages and found ASRock actually does a very good job on an otherwise inexpensive board. Gigabyte is very reliable, as well, which may be why you see so many hardware sites using one of their boards as a go-to for testing (like Tom's) ;).

In fact, my next board will probably be a Gigabyte, but so far I love my ASRock.
 
Solution