Does a motherboard affect general Performance

[GreenD]

Hey

I started this thread to find out if the motherboard affects general stock performance of hardware? such as I am planning on upgrading a few parts on my old computer, which has a P8 H67-M PRO motherboard and i was wondering if the motherboard would affect the stock performance of a 2600k and a GTX 670, im not planning on overclocking them

Thanks

 

[brandonwh64]

You should be fine stock.

 

[TheoneandonlyMrK]

+1, ,,,,,should be fine even if you do want to OC them ,and i recomend a bit of a tweek, decent psu required though 600-700 watt corsair CX builders psu is ideal imho

 

[sneekypeet]

motherboards can have "tricks" that can effect the "stock" performance, sure.

 

[GreenD]

It has a 775 watt Thermaltake Tough Power PSU

 

[GreenD]

motherboards can have "tricks" that can effect the "stock" performance, sure.

Can these tricks decrease the performance?

 

[DarkOCean]

You are good to go.

 

[sneekypeet]

Can these tricks decrease the performance?

usually it is done to increase stock performance, behind the scenes.

 

[cadaveca]

Hey

I started this thread to find out if the motherboard affects general stock performance of hardware? such as I am planning on upgrading a few parts on my old computer, which has a P8 H67-M PRO motherboard and i was wondering if the motherboard would affect the stock performance of a 2600k and a GTX 670, im not planning on overclocking them

Thanks

Motherboard itself doesn't affect the performance, but there are two things that can...the BIOS, first and foremost, as well as the the parts used to build it(ie, USB 3.0 controller, LAN controller, etc).


These differences can have a profound effect...or none at all. For example, wPrime:

These tests were done with stock JEDEC memory(1333 MHz 9-9-9-24). As you can see, there are differences, but that's wPrime..how about SuperPi?

Differences aren't as much! Can it affect 3d as well?

Sure can!


That said, a simple BIOS update can change it all! Mostly this can be attributed to differences in subtimings for memory, timings for chipset communication, etc. Most of these options are not available to the end user, however, buy the right board, and you can tweak your way into higher performance than any other, too, although typically you pay more for boards that are capable.

 

[GreenD]

Well if I was forced to upgrade the board I would use a Asus Z77 Sabertooth at the moment, but this mainly seems to be affecting RAM, I Dont know what to compare the last chart with am I correct in saying this? and would this be a problem for non over-clocking?

 

[JrRacinFan]

The only thing I can think of, the lost potential with the K multiplier(overclock it, it's not that difficult) and impact of ram speed. Now the ram speed can be a factor dependent if you know what your ram specs are.

Say if it's specc'd to do 1866 cl9, well you get essentially the fastest dram throughput as there is a very minimal difference going above that.

EDIT:

Remember, H6x boards, from what I know 99.9999% can only take up to 1333. z68/z77/p67 has more ram dividers for higher throughput.

 

[GreenD]

The Main Reason im seriously seeing if I can get to go with the old motherboard is because it has Windows 7 OEM software on it, I Dont exactly want to go buy new OS as that is another chunk out of the budget

 

[CaptainFailcon]

any difference thats less then 10% is within the error margin so no motherboards do not have a effect on performance at all

 

[cadaveca]

any difference thats less then 10% is within the error margin so no motherboards do not have a effect on performance at all

Not with my testing. Max variation is 1-2%, if not less. 10% is a ludicrous number.


For example, I have one X79 board here that is signifigantly faster for multi-GPU 3D gaming than any other...tested with many OS installs, many apps, many different SSDs, drivers, etc.

I run my testing AT LEAST three times. usally 5-6 is the result vary, until I get two results, one after the other, completely identical.


Variations in test results are usually down to Windows managing itself, whether doing disk tasks, memory management, or other things. It took quite some time for me to identify these things, and then eliminate them, mostly by registry and services edits to disable the offending parts of windows.

If you get a 10% variation in testing anything, there is something WRONG.

 

[CaptainFailcon]

I don't trust benchmarks not on win32 anyway windows behaves differntly from run to run its near on impossible to get the exact same result 2/3 runs

 

[cadaveca]

I don't trust benchmarks not on win32 anyway windows behaves differntly from run to run its near on impossible to get the exact same result 2/3 runs

LuLz. WIn32? Don't blame ya there. I use WIndows 7 Ultimate SP1, 64-bit. No way I'd feel confident reviewing memory with a 32-bit OS when most kits are 8 GB or larger now.

Windows DOES NOT behave differently from boot to boot, at least not for me. But I have a very specific routine that prevents issues like that, so even if it did, becuase I can get identical results from boot to boot, it's of no consequence to me, either.

 

[Crap Daddy]

The Main Reason im seriously seeing if I can get to go with the old motherboard is because it has Windows 7 OEM software on it, I Dont exactly want to go buy new OS as that is another chunk out of the budget

Stick with your mobo. At stock CPU and GPU there's no real world difference with a better board.

 

[DisplayandReplay]

I think so. For example, some mother board fit you case well. It works well with your ventilation system. Sometimes the extra heat may hinder the performance of your mobo. Case, motherboard and cooling system really matter.

 

[phanbuey]

not really if you are talking about different brands around the same chipset/gens of CPU... if you are overclocking, a board that can sustain a high OC is key, but these can be found cheap from companies like Asrock and Gigabyte... Asus tends to be a bit overpriced.

The examples dave listed are comparing AMD to Intel boards (the A75) and also to different processors, which is a bit misleading. But as you can see in the the same chipset line the max difference in 3d is 5015-4709 = 305... which is about 6.4% - not terribly noticeable, but there.

 

[Outback Bronze]

The differences between motherboards of similar characteristics are MOST of the time negligible. Id say your sweet to go.