Five $160 To $240 990FX-Based Socket AM3+ Motherboards

[Crashman]

[citation][nom]psiboy[/nom]"990FXA-UD7 includes a lighted CLR_CMOS button" shouldn't "lighted" be "lit" or "backlit" or something? Other than that cheers for a great article Thomas![/citation]Back lit works maybe?

 

[SessouXFX]

[citation][nom]f-gomes[/nom]200$ for a BD motherboard??? Isn't this like buying a toilet seat? I mean, what you will drop there will be shit, anyway...PS: I am a huge AMD fan, I really am.[/citation]
I'm doubting that "AMD fan" claim very seriously.
Because if you were an AMD fan, you'd know that any board utilizing AMD chips is backwards and forwards compatible with the next iteration. 8150 fx won't be the one and only chip to use these boards and there's still time to improve performance in the near future, before the Vishera cores come out later next year...

 

[alcalde]

I enjoyed the review, even though I'm still sold on the Gigabyte board (having no desire to throw 4 NVidia cards in it). However, there are some areas that might have shown much more difference between the boards that weren't considered.

What about data throughput? There were no storage tests done. This is particularly problematic given that the boards are using different add-on controllers in some instances for extra SATA ports or e-SATA, etc. The review was more useful than most in actually pointing out the models used, which makes the omission of any tests more confusing.

There is also a difference between ethernet chips and firewire chips, but again, no throughput tests. In my next-to-last motherboard purchase, throughput was the deciding purchase factor, as a certain other review site was the only one of half a dozen reviews I read to check and find a significant USB throughput difference in the board I was otherwise going to purchase. Picking up on the article's opening discussion of PCIe lanes, this review was the only one to explain how the PCIe lanes were divided not only among the slots, but among the internal components like SATA, USB, etc. as well (again showing where the board other reviews had left me favoring had certain devices sharing a PCIe lane while the board I ultimately went with used separate lanes for each).

It's possible for these third-party chips to suffer from hardware, BIOS or driver issues, and could really be a nasty surprise for a purchaser. I think it would be a great idea to include these types of throughput tests in future motherboard round-ups.

 

[Crashman]

[citation][nom]alcalde[/nom]I enjoyed the review, even though I'm still sold on the Gigabyte board (having no desire to throw 4 NVidia cards in it). However, there are some areas that might have shown much more difference between the boards that weren't considered.What about data throughput? There were no storage tests done. This is particularly problematic given that the boards are using different add-on controllers in some instances for extra SATA ports or e-SATA, etc. The review was more useful than most in actually pointing out the models used, which makes the omission of any tests more confusing. There is also a difference between ethernet chips and firewire chips, but again, no throughput tests. In my next-to-last motherboard purchase, throughput was the deciding purchase factor, as a certain other review site was the only one of half a dozen reviews I read to check and find a significant USB throughput difference in the board I was otherwise going to purchase. Picking up on the article's opening discussion of PCIe lanes, this review was the only one to explain how the PCIe lanes were divided not only among the slots, but among the internal components like SATA, USB, etc. as well (again showing where the board other reviews had left me favoring had certain devices sharing a PCIe lane while the board I ultimately went with used separate lanes for each).It's possible for these third-party chips to suffer from hardware, BIOS or driver issues, and could really be a nasty surprise for a purchaser. I think it would be a great idea to include these types of throughput tests in future motherboard round-ups.[/citation]At one time a team member pushed for interface testing, but it dragged test time out by three additional days per board. You wouldn't see many comparisons that way.

Tom's Hardware has a chip tester though and I'm hoping that some of his work will help buyers understand the differences between interface controllers.

 

[Guest]

I have the Sabertooth, it's a brilliant board.

 

[kooltime]

WoW, listing sabertooth in same class vs crosshair V non even testing that vs these others as a performance board ??

Sabertooth is intended for a different market segment, not a performance board as the corsshair V is specifically designed for.

The Crosshair V is the same cost range as all these others. Best board for AMD is the Crosshair V currently hands down. Anything else is for folks that dont need performance over cost and probably wont be doing any overclocks. Not including a board that is the current world record holder is silly claiming others are even close in performance levels, especially since they are all around same cost. If you tested that also here and shown some flaws it may have then a valid review would be had, but since you do not, its the best option.

Buy a cheapo board doesnt matter save some bucks or go with Cross hair V, there really is no other choices for AMD right now. The cost of that is same as all these others are same bracket range.

 

[alcalde]

At one time a team member pushed for interface testing, but it dragged test time out by three additional days per board. You wouldn't see many comparisons that way.

Tom's Hardware has a chip tester though and I'm hoping that some of his work will help buyers understand the differences between interface controllers.

Thank you for your reply, Crashman, but there's something I don't understand. Why would running a benchmark or two on the hard drive or USB performance add three entire days per board? Hasn't Toms gone so far as to acquire a 1000fps video camera and count frames just to get an idea of image lag for the best monitor reviews anywhere? Why is that deemed vital, but running an extra interface benchmark too much work? What if one video game FPS benchmark were omitted?

I did some checking, and legitreviews ran some Crystalmark and HD Tune tests on the Gigabyte board for their review. I don't know enough about their testing methodology to completely trust the results, but just as I feared the Asus Crosshair V board had significantly higher USB 3 read performance, by *18.5%*! The Gigabyte also had lower SATA throughput scores, and these results are enough once again to make me reconsider my initial preference in motherboard upgrades.

I've been reading (and trusting) Toms since the days of Mr. Pabst, and defending his initial claim that NVidia's Riva TNT was going to be superior to 3dfx's Voodoo 🙂. I'd rather a motherboard round-up that took a few extra days than one that failed to provide very crucial information.

 

[Crashman]

[citation][nom]alcalde[/nom]Thank you for your reply, Crashman, but there's something I don't understand. Why would running a benchmark or two on the hard drive or USB performance add three entire days per board? [/citation]Because it's not just a test or two on a controller or two, its a battery of tests for every controller. The HDD suite at that time took about 1.5 hours to run per controller, some boards had three (chipset, internal SATA, eSATA) plus USB 2.0 and 3.0, firewire, network controllers, audio and a couple other tests. They guy who wanted "everything tested" meant everything.

And the guys above HIM were insisting on four articles a month. So, I looked at what could be well-tested in separate articles, saw guys doing controllers, and decided to leave the controller tests to them.

 

[g-unit1111]

[citation][nom]buzznut[/nom]Take a look at the UD5[/citation]

Yeah I've been heavily considering that one as well. I really want to get a board with UEFI for OC'ing purposes.

 

[Guest]

oled tv thats the future i think

 

[youssef 2010]

[citation][nom]timbo1130[/nom]How is this relevant to enthusiast? Bulldozer is out classed by Sandy Bridge I don't care if there are a few less sata ports. If you need to upgrade your better off going with Sandy bridge and z68 or p67 or wait for SB-E and X79.[/citation]

The problem that faces SB is the connectivity limit. AMD never faced this problem and yet its products remain cheaper than Intel. Intel follows a simpler design that offers great performance at very low power requirements. The reason why they have such bottlenecked connectivity is a mystery to me

Too bad that Gigabyte's board had problems with OC settings. They seem to be running into many issues lately

 

[damric]

One of the better reviews by Tom's.

 

[SessouXFX]

SB-E doesn't look too good. If you go Intel, you'd be much better off with SB than SB-E....

 

[benage]

Watch out for ASUS stock firmware. I just got duped as the stock bios doesnt support the FX cpus, argh! The crosshair V boards support ROG usb flash which works without a cpu, wish i'd bought that now!

 

[Crashman]

[citation][nom]benage[/nom]Watch out for ASUS stock firmware. I just got duped as the stock bios doesnt support the FX cpus, argh! The crosshair V boards support ROG usb flash which works without a cpu, wish i'd bought that now![/citation]Asus is supposed to be shipping boards with the new firmware, BUT you should be wary of the possibility that someone might have a new-old-stock board.

 

[Crashman]

[citation][nom]hongday[/nom]If you ignore x58 because SB offers better performance, you ignore anything AMD has because a SB setup offers better performance. If you want 36 or less lanes, x58 still offers better processors than you can hope to get from AMD. Bizarre logic.[/citation]If you're including CPU, chipset and prices, you're comparing AMD Bulldozer to what, an X58 with Core i7-950?

I think the i7-950 is a valid target for Bulldozer, thanks for bringing that up!

 

[Paul Hurst]

I have a 990GXA -UD7 MB with an FX-8150 Processor. I am looking for a graphics card for this MB. Can I use just one card or would it be better using 2 or more? Could you also give me the name and model of the card(s). Thank You.

 

[junkeymonkey]

maby i'm wrong but all i see with the AM3+ is a dead end they just took an phemon2 board and converted it to work with AM3+ and never put out a true AM3+ board . i biught when it first came out and the thing was there was to be a new board with the 1090fx chipset well that was droped and now all you got is an old board thats been revised over and over with a long list of bios updates pre revision.. and nothing new board wise. buying AM3+ is like buying an old car with a new motor.