Discussion in 'Reviews' started by cadaveca, Jul 2, 2012.
To read this review go to: http://www.techpowerup.com/reviews/ASRock/Fatal1ty_X79_Champion/
Nice review mate. This motherboard isnt released yet, right?
Sorry but 8.9? im quite shocked, this board deserves whhhaayyy more IMHO
Yeah, I do feel like this is the best X79 board I have had yet when it comes to clocking and regular usage, however, the outstanding issues do prevent me from giving it a higher score. My USB 3.0 enclosure did not work at all, CPU performance is rather low, and a few other things I ran into like stock turbo not scaling properly are things that cannot be ignored.
At the same time, a new BIOS can fix all of that, and then it really would be killer...but now, it's just damn good.
a viable alternative to a rampage???
IMO yes, indeed
I cannot tell ya, as I never got one of those. What I can say, is that the Fatal1ty Champion is really good, even with the issues I have, and I've been using it for several weeks in my daily PC(including right this second).
I'm just waiting on a new BIOS.
Sweet review captain!
It depends on what you want to do with your motherboard, I personally bought Asus Rampage IV for the hotwire.
What about the VRM efficiency? It looks like it draws almost 20 more watts at stock speeds than the P9X79 Deluxe under load, then you mentioned that there were turbo issues. Is the loaded power consumption without turbo and if it is, that usage is terrible in comparison to other X79 boards. You wouldn't happen to have noticed the VRM temperatures, would you?
It looks like a well designed and well thought out motherboard, but even with such a number of small issues, it does make me wonder what else they skimped on or didn't put enough TLC into.
i am see u do some new nice points adding in this review, the review looking so cool keep on it, great work
As I mentioned in the review, my test CPU has degraded, so the numbers kind of give the wrong picture there. I recently got a second 3960X, and will be re-testing and reporting power consumption numbers with a bit more detail.
The issues...aren't really issues to me, rather than differences in BIOS programming. There's nothing that affects stability, even under large power-load OCs, (like 5 GHz on 12 cores and multiple VGAs), and it lets me use lower CPU voltage than others, so while certain things were not quite what I expected, overall it's pretty damn good.
View attachment 47777
please use a true multi-threaded benchmark for your synthetic memory bandwidth tests at least for the x79 chipset. aida64 has a single threaded memory test wich cannot saturate the quad-channel memory bus on x79 chipsets. it really skews the picture here if it is compared to the dual-channel chipsets.
how about sisoft sandra? it has a true multi-threaded synthetic memory benchmark...
Correct. However, although X79 does offer more bandwidth than the tests show, that bandwidth does not show much benefit except in specific applications.
I did use Sandra in past memory reviews, so I more than aware that it's basically the only benchmark out there that does use all of X79's bandwidth.
I will probably add it in my re-bench that is coming shortly.
Nice. Though I think a test that better reflects the average TPU user would make these reviews all that more meaningful.
What to simulate:
- Browser with 2 dozen tabs open. Compulsively refreshed between matches.
- The game being played.
- Blu-Ray transcoding.
- Hulu, Netflix, or previously ripped movie playing on second screen.
- BOINC or FAH making use of leftover resources.
I think something close could be scripted out to represent this common load.
I already put over 40 hours each week into every review. If you want to make me a script, that spits out results in .txt format, and the apps require no installs, I'm game. No guarantees I'll use it though, as testing does need to be able to show differences between products, and have repeatable results. Do keep in mind these are board reviews, and not platform reviews. That type of review, like my memory reviews, would have far different testing.
Do keep in mind as well, I make $0 doing these reviews, so if you guys want changes, that's no problem, but do consider how much work is involved, and how freely I already give my time to TPU.
I do get some small compensation, but all that goes to purchasing hardware for reviews. I had to buy the 3770K used for Z77, bought all the SSDs used, the PSU, case, everything...except the boards and a few memory kits. Any changes that require additional time or expense are not likely at this point, as I have already invested in current testing changes. All the testing I do does cover the type of workloads you have mentioned. Do also keep in mind that all tests I use are currently purchased software, and not free apps, so there is costs involved there as well. Doing reviews isn't just about doing some tests and posting results..it's actually quite expensive.
No, I don't think you cover my workload. That's why I asked (partially tongue-in-cheek, but it is something I've always sought from all the review sites for many years)
However, since I've thought up how to implement most of that, I'll get something together.
ACMARK, memory reviews, usage testing. I do have a personal set of tests I do, and I do use each board for most of the week. I do collect numbers that I do not report as boards tend to make little difference that can be numerically valued.
B:Game tests. I report with apps that react to motherboard differences.
C:Handbrake(I encode custom-made AVI to H.264, takes 5 minutes per run, I do 5 runs each test)
D:not covered, but I do exactly this testing while gaming. I usually sit in the BF3 TS when doing this testing, so many members do even hear of issues as they happen. If there is something that sticks out, It's mentioned in the review, but since that is more a VGA test than a motherboard test, and I am not reviewing videocards, it doesn't take a priority among things I report on. I use second screen for TS, AIDA64 for monitoring, and usually some sort of video from links dropped by TS members. It's testing, in a way, but really, that's just how I use my PC, as do most people!
E:Although not exactly in graph form, power consumption testing is done exactly so, but using Prime95 as the app to load the rest of the chip. Actually reporting this as a numerical value is kind of hard, but that testing does get done.
Really the issue comes down to being able to put this sort of feedback into a graph.
The point I think that was missed is that I'm going to handbrake while I game, watch a movie, and FAH is in the background. Alt-tab between matches, refresh all tabs, alt-tab to get back in the game. You know, usual stuff A+B+C+D+E! (come on, you couldn't possibly think that I wanted a stand-alone test that just refreshed webpages )
Speed tests are fine... though they've kinda lost meaning for me over the years. I'm looking more for a capabilities benchmark. Wrote up an outline for how I'd go about it, now for some research and testing.
No, I do understand. But making a number for a graph out of that is the hard part.
I just be in the minority. I only use maybe two programs at a time and only on thats stressful.
Cool. I wasn't sure earlier.
And did you say challenge? It's what I live for!
If I can execute it as imagined, such a benchmark could be telling to how much "system" someone would need for the intensity of multitasking they desire.
I agree, it is hard to give a score for that, although it could be a meaningful real-world measure for overall system performance if it gives reproducable results.
So how do you think this ASRock BIO's is vs. ASUS BIO's?
Yeah, that's the issue. I've spent some time trying to create a repeatable metric, but gave up and just do the testing anyway. I mean, if I find a weird bug, like my USB 3.0 issue with this boiard, I can report it to the board maker, and they can fix it, so I do use each product for a minimum of a week ehile doing evrything else I doo PC-wise.
Both are great, and both have features the other does not, so really, there is no valid compare.
Separate names with a comma.