Sign in with
Sign up | Sign in
Your question

Motherboard for new AMD build.

Last response: in Systems
Share
January 29, 2012 1:10:38 PM

Hello, I am planning on building a new computer if the very near future, with these specs:

AMD FX 6100
Sapphire HD 6870 1GB
8GB Corsair Vengance 1600 DDR3
HDD from older computer
Corsair Builders Series 600W PSU
CM HAF 912 Plus Case

I also have plans to buy a SSD later on.

However, I have a dilemma whether to buy the Asus M5A97 PRO or the Asus Asus M5A99X EVO. Does the EVO provide any greater performace over the PRO ?

Thank you for the help ! :) 



More about : motherboard amd build

January 29, 2012 4:58:19 PM

For the most part, motherboards aren't a performance sort of thing. Their purpose for the most part involves linking together the things that deliver the performance.

I am a proponent of going with only as much motherboard as someone is going to use and not more.

The two big things that are deal breakers with motherboard choice are 1) Necessary RAM, and 2) Necessary Video Cards.

Any micro board you can get for $50 can hold 2x 4GBs of RAM (8 total) which is more than enough for almost anybody's personal use. One video card is usually good enough for the average person's personal use as well.

If you need to be able to push up to 4x 4GBs (16) or 2x video cards, that would push you up to needing a regular atx motherboard for ~$125.

That is $75 that most people pay as insurance in case they ever need a future capability.

Most people would be better served by putting some of that $75 towards Crucial RAM and XFX PSUs instead of whatever else they are using and saving the rest, or spending some of the rest to move to the next higher video card level which helps to make up for no potential for 2 video cards somewhat.

Out of the motherboard types you chose, the EVO allows for better 2x video card performance (at x8/x8 instead of x4/x4) if you want to get 2 video cards. You would be paying $50 more for that added capability and little to nothing else.

In any event, I think you would be fine with the PRO and you might want to consider using some of that $50 savings to make the RAM and PSU changes I mentioned above.

Also, you are kinda getting towards the i5-2400 processor range. You may want to consider going with the Intel setup instead of the AMD FX setup.

There are two reasons why
1) A lot of times the FX motherboards need a BIOS update in order to be able to work with an FX processor. If you don't have another AMD processor laying around that you can use while you update the BIOS then your computer might never turn on.
2) The FX processors were a flop in the marketplace for a reason, the i5-2400 is about as good as an FX-8150 (if it is unOCd) and with much less heat and no configuration problems (BIOS update necessary before it works).

The 8150 costs a lot more than the 2400 just to perform worse. If OCing is on the table then the 2500k unOCd is about equal to the 8150 OCd and the 2500k OCd just makes the 8150 OCd look sad, still for less than the 8150 costs.

I am surprised nobody has chimed in with this yet, but usually a lot of people here sniff out people that are interested in spending more than $125 on an AMD processor and spam them with requests that they switch to Intel.

You will probably have at least 4 or 5 by the time the thread closes.
a b Ĉ ASUS
January 29, 2012 5:09:33 PM

Quote:
I am surprised nobody has chimed in with this yet, but usually a lot of people here sniff out people that are interested in spending more than $125 on an AMD processor and spam them with requests that they switch to Intel.


I have an AMD CPU and I really like the Phenom II but I really hesitate to recommend FX at this point. If you're getting AMD go for the X4-960T or the X6-1100T. This article explains a lot of reasons why I'm against it: http://www.tomshardware.com/reviews/windows-7-hotfix-bu...

It's not just the BIOS issues that everyone is having problems with - there's a lot of software issues as well.

As an AMD user I will not say to totally shy away from AMD, but I will say to stay away from the FX because there's so many problems associated with it. AMD's done much better than that, and I will agree that it makes more sense to get an SB-based system but you can't go wrong either way.

This explains a lot about FX's architecture and why it wasn't up to the standards that AMD had promised: http://www.tomshardware.com/reviews/fx-8150-zambezi-bul...

Quote:
However, I have a dilemma whether to buy the Asus M5A97 PRO or the Asus Asus M5A99X EVO. Does the EVO provide any greater performace over the PRO ?


The 990X is the most current AMD platform available. The 970 is really good as well. In fact there's a couple of boards I'd recommend as well:

- http://www.newegg.com/Product/Product.aspx?Item=N82E168... (same board I have and I really like it - Gigabyte's BIOS is really easy to work with)
- http://www.newegg.com/Product/Product.aspx?Item=N82E168...
Related resources
a b Ĉ ASUS
January 29, 2012 5:13:44 PM

He is correct, you will have to do a bios update for the new boards. I am running the same CUP. I had a 955 before this and it took me a month to realize what needed to be done. I had to get another copy of windows and install it and then i went it to ASUS AI Suite 2 and flashed my bios right in the OS then when i got back into windows i tested and made sure everything was still gonna work. I Said my goodbyes to my 955 and shut my computer down and changed to the FX 6100. once u get past the bios windows installs the drivers for you (Windows 7) and to go in a bit more depth, once u have updated the bios u must go back and set primary boot in the UEFI Bios. but i have heard that people will go and borrow a Athlon or somthing from Tiger direct for a few days just so they can update their bios. Is not as much trouble as it sounds believe me. I was frustrated but after i did it. i was like wow this was.....easy. But if you use a older drive from a different computer your gonna need to re-install windows. Trust me i tried the same exact plan but when when windows came up...well take my word for it you won't be able to use that copy of windows anymore...(Unless u use the original parts from that computer)
January 29, 2012 5:16:29 PM

It depends on whether overclocking is something you'd be interested in. A micro atx board is fine, but it sounds like you are interested in something more robust.

These two boards look very similar. The main difference I see just at a glance is the 4X USB 3.0 support for the EVO. They both look good from an overclocking and multi GPU support perspective.
Even with the latest graphics cards, an 8X/8X pci express configuration is not much different from the 16X/16X lanes provided by the 990FX chipset. So dual GPUs shouldn't be a problem.
I think you'll be happy with either of these boards, just ask yourself how many USB 3.0 devices do you need?


Ah..correction. The 990X will be better for FX CPUs.
January 29, 2012 5:34:48 PM

The FX has mixed reviews. Ive dug pretty deep in regards to these processors and have come to the conclusion... its not totally the processors themselves, but the software they encounter. I take benchmarks with a grain of salt... you'll find most people here worship and treat benchmarks as Gods which is truly disturbing to me. I've followed the lines of benchmarks before only to be let down in a "real" world situation. Linux users seem to be having great success with FX processors. Ive seen more praise than failures.

In fact, there is an amazing benchmark of an old 2005 HP Pavilion a1230n I uploaded to Passmarks Performance Test that I should be shot for uploading, but I did it out of sheer amusement because I know benchmarks arent what they are cracked up to be... its benchmark shouldn't even be in the spot its in, in comparison to whats around it and its running stock motherboard and processor.


Either way... with that said... I agree with Raiddinn about the motherboard.
January 29, 2012 6:13:19 PM

It actually isn't the 990FX chipset that allows x16/x16 it is one chip in particular called an NF200.

When you are in x8/x8, the processor can directly work with the graphics card slots.

When you want x16/x16, the NF200 has to sit in as a middleman in everything that goes to both cards.

This is why x8/x8 usually outperforms x16/x16 by a few FPS. The cards aren't using more bandwidth available than what you have in an x8/x8 setup AND having to go through the NF200 slows down the data moving back and forth.

IMHO, nobody should try hard to get x16/x16, because of how hard it is to use more than x8 bandwidth (this isn't usually the bottleneck).

AFAIK, though, none of this really matters because I know the PRO doesn't have an NF200 and I am pretty sure the EVO doesn't as well.

If you managed to get 2x video cards in the PRO setup it would be x4/x4 and if you managed to get 2x video cards in the EVO it would stay on the regular paths in x8/x8.

a b Ĉ ASUS
January 29, 2012 6:23:27 PM

biscuitasylum said:
The FX has mixed reviews. Ive dug pretty deep in regards to these processors and have come to the conclusion... its not totally the processors themselves, but the software they encounter. I take benchmarks with a grain of salt... you'll find most people here worship and treat benchmarks as Gods which is truly disturbing to me. I've followed the lines of benchmarks before only to be let down in a "real" world situation. Linux users seem to be having great success with FX processors. Ive seen more praise than failures.

In fact, there is an amazing benchmark of an old 2005 HP Pavilion a1230n I uploaded to Passmarks Performance Test that I should be shot for uploading, but I did it out of sheer amusement because I know benchmarks arent what they are cracked up to be... its benchmark shouldn't even be in the spot its in, in comparison to whats around it and its running stock motherboard and processor.


Either way... with that said... I agree with Raiddinn about the motherboard.


I do too, because every system is different and what's tested will not always reflect how your system performs. And I never cite the reviews and benchmarks on Anandtech as credible because they show you what hardware they tested but time and time again they do not, under any circumstances tell you what other hardware they test with. The thing is a lot of these benchmarks that are published are fabricated to give them the results they want.
January 29, 2012 6:27:17 PM

g-unit1111 said:
I do too, because every system is different and what's tested will not always reflect how your system performs. And I never cite the reviews and benchmarks on Anandtech as credible because they show you what hardware they tested but time and time again they do not, under any circumstances tell you what other hardware they test with. The thing is a lot of these benchmarks that are published are fabricated to give them the results they want.



Finally, Im not alone. :pt1cable:  :pt1cable:  :pt1cable: 
January 29, 2012 6:40:16 PM

They probably just don't test with multiple hardware setups. Most likely the thing they list as their test platform is the only one they used.

Just sayin.

If anything, I would complain about Anandtech having quite lofty system specs in their testing platforms that don't really reflect the average setup very well.

From what I can tell they use a 2500k for all their testing and while that is a commonly suggested processor around here it is still something that only a tiny minority of the whole world uses.

I think I might rather see them do at least some benchmarks with an i3-2120 or something instead so people will know if their video card would be bottlenecked by the dual core CPU or whatever.

If they are only going to do one set of tests, I guess I would rather see the 2500k to get a better idea of the max capabilities of the card, but the 2120 would still be really nice.

That being said, it can take a pretty long time to do these sorts of benchmarks, especially if they have to do more than one in a week if that is the case.
a b Ĉ ASUS
a b À AMD
January 29, 2012 7:39:05 PM

I like the M5A97 as one of the better single graphics card boards .

I dont think there are any issues with FX processors except higher than power usage than we hoped for . As you can see from the link g-unit posted above the FX 8150 matches or exceeds the performance of the intel 2500K almost all the time .

But I dont think the FX 6100 is a good buy . It doesnt have the encoding muscle of the 8 core and it doesnt game as well as the FX 4100 . Games dont use 6 cores so you are better off with 4 cores clocked higher .


EDIT : Anandtech benches are nonsense . They use impossible hardware set ups that attempt to isolate one component ....like using the same $300+ graphics card with all processors whether they cost $100 or $300 . It builds a misleading picture of performance .
And so does their use of lower resolution screens for benchmarking .
a b Ĉ ASUS
January 29, 2012 8:20:13 PM

Raiddinn said:
They probably just don't test with multiple hardware setups. Most likely the thing they list as their test platform is the only one they used.

Just sayin.

If anything, I would complain about Anandtech having quite lofty system specs in their testing platforms that don't really reflect the average setup very well.

From what I can tell they use a 2500k for all their testing and while that is a commonly suggested processor around here it is still something that only a tiny minority of the whole world uses.

I think I might rather see them do at least some benchmarks with an i3-2120 or something instead so people will know if their video card would be bottlenecked by the dual core CPU or whatever.

If they are only going to do one set of tests, I guess I would rather see the 2500k to get a better idea of the max capabilities of the card, but the 2120 would still be really nice.

That being said, it can take a pretty long time to do these sorts of benchmarks, especially if they have to do more than one in a week if that is the case.


I understand that they don't have the time and resources to test with every part and component out there (nobody does :lol:  ) but the thing is when they do a benchmark they should at least take the time to explain how they arrived at those results. Any science major will tell you that you need to do that for any controlled experiment.
January 29, 2012 8:32:49 PM

Sincerely? I don't think for games the FX processors are THAT nice comparing to sandy bridge.. I a friend that plays skyrim at 1080p, and changed the FX for a sandy-bridge, he got a HUGE upgrade.. even at that resolution that shifts the weight to GPU (accordly to him about 30% more fps)
a b Ĉ ASUS
a b À AMD
January 29, 2012 8:49:18 PM

vitornob said:
Sincerely? I don't think for games the FX processors are THAT nice comparing to sandy bridge.. I a friend that plays skyrim at 1080p, and changed the FX for a sandy-bridge, he got a HUGE upgrade.. even at that resolution that shifts the weight to GPU (accordly to him about 30% more fps)


Tomshardware testing of Skyrim had it running on dual cores as well as quadcores .
Its one game that will benefit from the stronger individual cores of a sandy bridge , but its the exception and not the general experience .
January 29, 2012 9:53:34 PM

I just troubleshooted a problem with a FX processor over the last week that was fixed after the OP acquired an older AMD processor and used it to update the (iirc) 990 FX BIOS.

This problem is still seen "in the wild" apparently, and it shouldn't be completely ignored.

If you are pushing i5-2400 pricing territory you may want to just get that instead.
!