Does a motherboard actually affect performance?

[polycroc]

Hi, long time reader, first post

I've recently bought an i5 3470 for cheap and got myself an Asrock Z77 extreme4 board..

I won't be overclocking, and apart from extra USB3 ports, and some features, does a board like this actually increase performance over a much cheaper budget board?

All the blurb on the board sounds impressive (5x speed on this, that and the other) but if I were to stick the chip on a budget board would effective windows and gaming performance be approx the same?

 

[Sheep221]

It doesn't, in case of overclocking, the premium boards only offer better VRMs and extra features and/or better design, but since the memory controllers are part of CPUs the motherboards are no longer a perfromance-based components.

If you notice the mobo ads and box features, everything that is mentioned are comfort and power saving features, maybe overclocking and increased lifetime features but that's all, there is nothing about computation performance at all.

 

[ShintaiDK]

The cheapest H61/H81 board performs exactly the same the most expensive Z77/Z87 board with the same CPU and using stock settings (no OC.).

You pick a board on features.

 

[aigomorla]

It doesn't, in case of overclocking, the premium boards only offer better VRMs and extra features and/or better design, but since the memory controllers are part of CPUs the motherboards are no longer a perfromance-based components.

If you notice the mobo ads and box features, everything that is mentioned are comfort and power saving features, maybe overclocking and increased lifetime features but that's all, there is nothing about computation performance at all.

this is very situational...

provided your doing moderate OC, then yes, your absolutely correct.
However if you go down to the extreme end.. as in sub ambient, the VRMS and cold choke play a lot into board selection.

Also, having more efficient VRMS allows less voltage required to achieve the same overclock.. sometimes even netting you a higher overclock.

Different vendors also use different bios's which can play to how stable your system will be at said overclock.

So... id say.. you can probably take everything u and i both said with a grain of salt.

 

[Smoblikat]

For a stock computer with current processors every single board performs the exact same.

 

[paul878]

Premium boards are just hypes, after the first 2 week you won't care what board is inside the case.

 

[alphascout]

Performance is going to be the same but VRM's and things like is what i am paying premium for and the pretty colors cant forget the color.

 

[Zap]

There used to be the possibility of a performance difference, back before the integrated memory controller. That got rid of probably 90% of any performance differences. Probably another couple percent can be accounted for by motherboard manufacturer "cheating" by having slightly higher base clocks (or these days automagically having higher Turbo speeds). I think if you normalized for clock speeds (and a single graphics card or integrated graphics), probably wouldn't have more than 1-2% performance difference between the cheapest $50 board and the most expensive $350 board.

tl;dr - Negligible

 

[Ketchup]

CPU performance will be the same (within a pretty tight margin of error). If you look at good reviews though (like the ones here on AT) you will notice some boards do better in SATA/Audio/LAN/USB performance than others. Usually cheap boards use cheap controllers, causing the difference, at least on the benches.

 

[Deders]

I've just read a very interesting article in PC Format that proves otherwise, for a while it did seem that all similar motherboards performed very closely, especially as more and more of the features were being put onto the CPU but there are some interesting results.

All the motherboards have the Z87 chipset:

In Cinebench multithreading tests some gave a 0.5 advantage, lowest being 8.03, highest being 8.50. That is a 5% difference.

In a shogun 2 benchmark the FPS range was even wider, although most managed between 38-39fps, the gigabyte Z87-D3HP only managed 34.5fps and the MSI Z87-GD43 managed to beat its bigger brother the GD65 by over 4.5fps with a score of 44.6fps, over 10fps higher than the gigabyte.

These anomalies were retested several times.

I remember the days when my Nforce 2 board could push enough performance out of my 2GHz 2400XP (with extremely well optimised timings and a very mild overclock) to beat a 3GHz Pentuim 4 in most tests.

 

[spat55]

If you have a non K chip then the Z77 board is wasted. The only thing really different motherboards do is give you more options, like more SATA III ports etc, other than that some have more features like power saving, overclocking etc but just out of the box the only difference is the physical features.

 

[piasabird]

I would say that maybe you can get better overclocking on some motherboards. It depends really how hard you push the components. Whether it is overclocking or something like encoding video or compiling graphics nonstop overnight having a more durable and stable power circuitry might make a big difference. Most people dont push a computer very hard doing e-mail and watching youtube.

 

[piasabird]

I use to say the same thing about Asus. What Asus would do is they would make motherboards and they would have a standard version and a PRO version of the motherboard Usually on the Pro they may include Bluetooth or a better network controller or maybe more SATA connections or a RAID chip. If you want one or two harddrives and just the averag i-5 processor, or an i-3 processor all the extra connections are worthless. In fact extra connections might be a drag on power resources. If you want more SATA connectors then maybe the z version of a chipset might help with that.

 

[Zap]

I've just read a very interesting article in PC Format that proves otherwise
...
In Cinebench multithreading tests some gave a 0.5 advantage, lowest being 8.03, highest being 8.50. That is a 5% difference.

As I mentioned above, the motherboard manufacturer could be "cheating" with higher clocks.

MultiCore Enhancement

Also, some motherboards ship with the BCLK a tiny bit higher, like 100.5MHz instead of 100.0MHz.

 

[Deders]

As I mentioned above, the motherboard manufacturer could be "cheating" with higher clocks.

MultiCore Enhancement

Also, some motherboards ship with the BCLK a tiny bit higher, like 100.5MHz instead of 100.0MHz.

It's possible but over 10FPS difference in a highly threaded game? The roundup did conclude that the gigabyte board was definitely one to avoid for much more than just this reason.

As for the higher clocks, most boards I've seen in the last 10 years have a feature that adds 5% or so to the base clock.

 

[jkauff]

Better Asus boards come with some nice performance-enhancing software, like AI Charger + and USB 3.0 Boost. The latter makes a big difference in my external drive transfer speeds.

 

[Deders]

Does it? I thought you'd be limited by the actual drive performance.

 

[Zap]

It's possible but over 10FPS difference in a highly threaded game?

"Multicore Enhancement" benefits highly threaded stuff even more. Normally the highest CPU turbo speeds is only if a single core is loaded. Multicore Enhancement makes the CPU run at the highest turbo speeds even when all cores are loaded. Some manufacturers boost that even farther. For instance the Gigabyte G1.Sniper 3 board that AnandTech tested ran their test 3770K at 4.0GHz with all cores loaded. Normally with all cores loaded, that CPU would be only 3.7GHz.

The AnandTech article also mentioned boards they've tested in the past with "stock" FSB/BCLK speeds up to 102.1MHz, from default 100.0MHz.

Manufacturers do this tomfoolery because they know we (you, me, everyone else in these forums) will look at benchmarks and go, "oh wow, that motherboards is TEH FASTEST! MUST BUY!"

Better Asus boards come with some nice performance-enhancing software, like AI Charger + and USB 3.0 Boost. The latter makes a big difference in my external drive transfer speeds.

Does it? I thought you'd be limited by the actual drive performance.

I think it does some voodoo with reducing USB driver overhead or something like that by enabling UASP (USB Attached SCSI Protocol).

 

[2is]

Better Asus boards come with some nice performance-enhancing software, like AI Charger + and USB 3.0 Boost. The latter makes a big difference in my external drive transfer speeds.

Unless your external drive is an SSD, it won't make a bit of difference. USB 3 is plenty faster than the fastest mechanical drive.

 

[jkauff]

Unless your external drive is an SSD, it won't make a bit of difference. USB 3 is plenty faster than the fastest mechanical drive.

I can only speak from experience. I have five WD My Book USB 3.0 external drives, and file transfers are as much as 25% faster using the Asus utility. This is with multi-gigabit size files--maybe I'd see less of a difference with a lot of small files.