Discussion in 'Reviews' started by cadaveca, Nov 12, 2012.
To read this review go to: http://www.techpowerup.com/reviews/ASRock/FM2A85X_Extreme6/
Thank you for the review.
ps.: Just curious: why is the post from Nov. 2012 ? Is it just the date of the template post?
With very good performance and a lot of features, this board seems like a steal at $90
Yea, that's odd.
It is little things like this that ASRock is great at doing. They really do pay attention to the little details.
It is also nice to see the FM2 processors getting some high end board attention. This platform might be a budget minded platform, but that doesn't mean that people won't be building higher end rigs with it. And as an enthusiast I love to tinker with everything, I actually prefer tinkering with budget minded platforms more so than high-end stuff. But it is the worst when your ability to tinker is limited by the motherboard. I'm glad we have ASRock to provide a true high end board for the FM2 processors!
Just a little typo I noticed, I think you meant the FM2 ASRock board.
My guess is that is when he started writing up the review. But with the PSU issues, and needing to ask ASRock for another board and send everything off to ASRock, it probably took a little longer to finish.
That was when I began working on this review. As mentioned in the review, due to a PSU issue, I requested a second board when I encountered problems, and the second board came direct from Taiwan(as do most of my samples, since I typically get them before the product launches). It has the same problem, so I then sent both boards, my APU, and my PSU to ASRock in Taiwan for testing, and then had to wait for it to be tested and then returned before I could do my review testing. Of course, all of this took time.
This PSU issue, and a few other, possibly related hardware issues, are why I haven't been posting many reviews lately. It's not a shortage of samples...I've got a tonne sitting here. Diagnosing problems properly takes time. I contacted ASRock on this issue about the PSU issue, they wanted to verify to be sure, so I sent them the stuff. They confirmed the issue, of course, which also prompted me to do other testing with other products for research into what exactly went wrong.
I'll talk more about other issues I found with other products in the coming weeks, as the opportunity presents itself.
Heh, that's a lot of time, it's nice that you finished it after all. Looking forward to hear more about the PSU issue. Thanks again.
Sometimes reviews take many months to complete. Sometimes due to myself, sometimes the product maker would rather the product wait. I do aim for completing a review every week, if not more, but at the same time, I don't rush stuff, and make sure I'm confident on what I am reporting.
I often get many products at once, saves the company on shipping at least. The Gigabyte Z77N-WiFi, they sent to me in September, and I got like 5 boards from Gigabyte within a week. All of those are done now.
I have four other products that have been waiting for months, a few memory kits, and a HTPC. HTPC was been waiting for other relevant hardware, and will be done very soon, two of those memory kits were bad, and at the same time, the PSU out of memory testing rig, that Antec, obviously had issues, and my memory testing platform is now dead and needs replacing...which means I got to re-bench a whole bunch of memory kits to get numbers for that one review.
Doing reviews isn't all fun and games, unfortunately. When you get products before they hit retail, you tend to run into bugs here and there. Boards are one of the most problematic, evidenced by BIOS updates for board being put out many many months after a launch.
I have three more ASRock boards here, one is doing a 72-hour Burn-in testing right now, another will get started Friday, and in the meantime, I'm editing pictures and making graphs for next week's review. I usually have three or four reviews on the go at once. Oh, and I'm doing memory testing as well, well, at least, I will be working more on that later today.
EDIT: Oh yeah, one of my test rig cards died too, just got the email that it has arrived at the vendor...for the second time in two months. I've had a lot of funky hardware lately, and perhaps all due to this PSU...thankfully Seasonic stepped up to the plate and gave me a new one!
Yes, I know very well that testing computer hardware is nothing like a simple "walk in the park", it never was.
Thanks for all the hard work well done
Hard to beat for the price. Bustin' nuts review as always man.
wow an almost perfect 10, Ever since reading reviews of Asrock past 2010 they been kicking ass and then taking names down and putting notches in their gun belts.
I've only played with a few ASRock motherboards but I've noticed one thing they do that ASUS and Gigabyte don't do. If you set the FSB to 100 in the BIOS then it's 100. With ASUS and Gigabyte it's usually higher than 100, hence why there's a little variance when testing motherboards.
It would seem this board is 99.8 when set to 100...
All board, including every ASRock I've ever used, has a slight variance in the FSB from where it is set.
I understand that they rarely ever set to the BIOS FSB speed. What I was saying is ASUS and Gigabyte tend to run higher (example 100.5-103). The couple ASRock boards I've installed were much closer to the set FSB in the BIOS.
This is what I'm saying might account for the differences in performance.
Great review, just the board I was interested in for a build...
I only would complain about the 3D tests, they're pretty poor in variance, I'd like to see a RAM speed round-up on APUs as well as more games and configurations under those tests.
I'm hoping they're already preping up a A88X succesor to this, with VirtuMVP and Realtek 898 all over again, as I'm thinking of might as well wait for Richland now, since my "customer" doesn't have the cash at hand right now anyway.
Yep, exactly, glad you caught that.
Thanks for your feedback. Unfortunately, no new test additions are planned at this time. The tests I do use are specifically chosen as they show differences between boards, on average. Each test is run manually, and is run 5 times to cover test variances. Each test is sensitive to different parts of a system, like F1 2010..it's very sensitive to changes in memory performance. PCMark and 3DMark are sensitive to other things(desktop and general 3D performance).
SuperPi alone takes of 20 minutes. multiply that by 5, and you have nearly 2 hours spent on a single test. I also do burn-in tests of all boards, for 72 hours(something I picked up from working in a PC store).
If I told you how much doing reviews actually pays, You'd understand why time is so important to me. Every review, I have to carefully budget my time. Board reviews are the most complex, as they incorporate more components than any other part, and as such, there is a huge amount of work involved in putting out a board review like I do. The fact of the matter is, I put out more test results than many other sites, and I actually use each product for at least a week so that I might uncover any oddities that can happen due to memory or parts incompatibility.
Doing construction, I make 2.5 times what I get for a single review, that might take a week or more, but in construction, I earn that 2.5x in just one hour. The fact I've been doing reviews now for a bit over two years is pretty amazing in and of itself, since I do have a wife and four kids who take up the rest of my time.
Time to ask for a raise
TPU cannot afford to pay me $80k a year, working only 8 months out of a year.
I don't really get paid, so to speak, as I am not employed by TPU. TPU gives me a "per diem" so that TPU owns the reviews, and that's enough. I'm not doing reviews to make money...I'm doing them to ensure realistic perspective is given.
The variance in performance wasn't what I was addressing in your post, I was addressing your claim that when you set the FSB to 100 on an ASRock board it actually is 100. That part just simply isn't true.
That changes from board to board, even, and isn't a brand-specific thing.
That's my point. I mean, I honestly can't even remember a single time when I had a board that ran at exactly the FSB I set in the BIOS.
Most Z77 products do, although you have to manually input the frequency. Even then, some boards offer more granularity than they can actually provide...like you set 100.1 MHz, and you get 100.3 MHz. Set 100.3 MHz, you get 100.5 MHz.
It's really dependant on BIOS, and some board makers have better BIOS programmers than others. Really, after doing so many reviews...that is the one thing I have learned...BIOS, and who makes it, is really what sets board makers apart from one another. ASRock makes a damn good BIOS, both on Intel and AMD platforms, while some other brands do well on Intel, but horrible on AMD.
What sets the FSB then if its not consistent?
Different clock generating chips(other onboard controllers, PCIe buffers), and different "size" of adjustments in BIOS. What sets the clock, really, depends on the platform, too.
Different clock generating chips? That means it could vary from board to board in the exact same model no? Not just vendor?
No. Every board of the same model should be the same. Different models will be different, to a degree...depends on what they are built with.
Specifically with FM2, quite a few boards use outside clock generators, in order to allow more clocking ability(such as the Biostar A85W-HiFi).
That chip is here:
This is a clock gen you'll find on 9-series AM3+ boards. As you can see in the picture, it's uses a quartz crystal to generate it's clocks, and the type of crystal used dictates what steps in frequency are available. If you changed that crystal, you'd have different clocking ability, too.
Separate names with a comma.