Discussion in 'Reviews' started by cadaveca, Jun 3, 2011.
To read this review go to: http://www.techpowerup.com/reviews/ASUS/M5A99X_EVO/
Not to rain on anyone's parade, but this review is essentially a copypasta of the M5A97 EVO review with a few bits tweaked. Also, the phrase "hit the next page" gets really annoying really quickly when it's used on EVERY PAGE OF THE REVIEW.
Yeah, the board is almost identical to the M5A970 EVO, so of course the review is near identical. Only so much can be said about the little bits that go on the board.
I still had to put in near 25 hours of testing, and I do a review every week, which amounts to near 40 hours per board, with writing, pictures, graphs, and getting it on the site.
You'll also find no fluff, no marketing speak, or anything like that.
First off thank you for the awesome review.
Second a quick question;
I was looking through your review of this board particularly & was pondering purchasing one of these. Would it be possible to do any core unlocking on it? Under Advanced CPU Configuration in bios maybe?
You bet there is, as well as the option to individually enable/disable cores. I checked to see if this screws with the temp sensors, and of course, it does.
The whole 970/990X/990FX pci-e bandwidth thing made me wonder if a NF200 chip could be put on a 970 board to get you 2x 16. And if so, would a board like that be cheaper than a 990FX board? If so maybe that'd make a good Asrock product or something.
Ok, os here's my take on it:
970 = Singlecard use.
990X = medium-performance MultiGPU, Crossfire or SLi
990FX = MultiGPU(highend), up to 4 cards.
From what I know, NF200 introduces some latency that leaves it's x16/x16 connectivity a bit under-performing comapred to x8/x8, and can even point out that sometimes, P55 even, gets better performance than X58, with x16/x16.
If X58 shows some performance deficits on occasion, then taking NF200, and turning a x8 connection into a x16 connection, doesn't really offer anything, except another chip "in the pipeline". The only time I could excuse the NF200 is when it is paired to chipset, not to make x8 lanes into x16, but when it add a completly whole new slot, offering 4-card connectivity to a chipset that doesn't natively support such configurations.
As to costs, i cannot comment on that one. To me, it seems that just simply using 990FX instead of 990X would be the cheaper option, but I'm not privy to such pricing information.
My thoughts on the NF200, if your planning on tri-sli get a board with one. To me SLI is almost a waste on its own, let alone tri SLI. That's just my personal opinion though and not many people listen to what I got to say
Thank you for looking into that for me & answering that question! Now I can move on and look into this board as an option.
I typically check out everything a board offers, BIOS and software included, so really, I already knew the answer before you even asked. ASUS has both of those things done really well(software and BIOS), but I've spouted enough love about that in other threads, for sure. Being able to capture screenshots of the BIOS by just pushing F12 saved me tonnes of time in taking pics of the BIOS and editing them to size, which is something I really really appreciate.
I cannot think of how many times people have asked me for subtimings for memory, and usually the first thing I ask for is such screen shots, and getting them can be a pain, but ASUS fixed that problem, 100%.
I don't see where in your review you list the parts on the M5A99X that are higher quality than the M5A97. As far as I can tell, the only differences are slot configuration, two JSATA ports, and crossfire/sli 8x/8x. Can you be more specific? I see you also got a nicer overclock on the IMC, processor, and memory. Were those not attainable on the M5A97?
No, they were not, and partially because power consumption was higher on the M5A99X, IMHO. I can only place that as being better VRM components, even though they are of the same brand and type.
Yes, the only differences between the two are the extra JMB362 controller (and associated parts), the slot config (and associated board parts), the addition of the power-eSATA port, and the SLi support.
The only explanation for the better clocks, because my CPU hits the VRM's limit for power, is that somehow the VRM is built with a higher bin of parts. The VRM is so customizable, it's hard for them to pull off some weird BIOS mojo to pull this sort of stuff off.
I actually wrote this review a bit over a week ago, and since then, Newegg has a video up with "JJ" from ASUS showing the 4 boards, and he mentioned that the parts were a bit better too, so I do beleive this to be the case.
Even the auto-clock utility ended up with better clocks on the M5A990X EVO.
Awesome review. I'm loving your reviews on the new (AMD 9x0) motherboards.
Any chance you'll be reviewing a Sabertooth eventually?
Because that's the one I'm eye-balling and I trust your reviews.
Lots of people likin' ASUS, I see. More interesting to see who likes what.
I guess I should just tell ASUS to send me one of everything.
I really do love your reviews and am still hoping to see a Crosshair V review from you someday.
Your reviews of the lower end boards have cemented the fact that I am going Asus for my next board though.
Glad to see they are going full UEFI!!!
cadaveca could you include the 7XX-8XX chipset boards in your next review so we can see the difference in performance
also i can't wait to have TPU in my rig
Not unless someone sends me some. They'd be in there if I had them. Currently every board I've got for review, and at least one I bought myself is there.
Yeah, there's some TPU in each and every board.
someone send this man a mobo
To be completely honest, I do not think ther would be much difference...maybe a bit in drive and USB performance, but really, the rest is only BIOS-level timing tweaks, as the CPU is just the same. The chipset isn't going to influence performance very much, unless there is something wrong.
So i expect little-to-no difference.
There is several ahrdware features that are focused for upcoming BUlldozer CPU, and BIOSes will be tuned specifically for that platform. 7-series for AM3 quads, 8-series for Thuban 6-cores, 9-series for Bulldozer.
I had to buy CPUs for these reviews; i think I'm going to get rid of the 970 board to recoup some of that cost, and maybe the older 11556 boards I reviewed, and the CPUs....but I have to be ready to purchase Bulldozer @ launch. I'll 100% for sure be keeping this 990X board, and will revisit it with Bulldozer in the near future.
As it is, I'm not too concerned with performance, no matter the platform, and would almost like to remove that from my reviews, but I know you guys like to see the numbers, so I've got a few tests there. The way things are going, boards really are going to ahve little-to-no perforamcne impact at all, like I said, unless something is wrong...what's mostly left on boards nowadays, is basically all "dumb logic".
Other reviewers will cover a wide variety of benchmarks, multiple instances of the same type of benchmark, etc, but I'd much rather show you guys WHY there might be a performance difference.
I think ASUS need to send me the highend boards, the ROG expander, 4 Nvidia GPUs, and 4 AMD GPUs. The next thing I'm going to spend some time working on is having a board fully populated...all drive plugs and slots filled...seee how it goes. I'm currently looking at boot times and such, trying to find an effective method of showing some differences there.
Ive got the M4A79XTD EVO so im guess this is the spiritual successor to that?
That's the board I've got in my game box that I'm going to replace with this new board I just bought from NewEgg. And thats based upon this review. Perfect timing on this review Cadaveca.
Cadaveca, are you going to come out with more 9 series reviews?
I'm still interested in fan control, and what seems to be glossed over in almost all reviews is comparisons between different boards.
I was thinking that the 990x was pretty much the 990fx with less PCIe lanes, and unless you were running more than one card there was no difference. I'm now seeing there's power consumption differences... so I'd love some kind of comparison between the different boards once you've reviewed em. It'd also be fantastic if you could compare MSI, Gigabyte and Asus board families against each other so that us consumers could pick a manufacturer and board that does what we want, without paying for stuff we don't need.
And yeah, I know, it's a lot to ask!
You know what..It's not alot to ask, at least not of me.
I live in Canada. Right now there is a postal strike. FedEX, UPS, DHL, still all delivering. That may have thrown a wringer into the works here.
I was expecting some products from other OEMs, but nothing has arrived on my doorstep.
However, I do have another review done, for Z68, which will be live soon. Other than that, I have nothing waiting to be done, so it's up to the OEMs.
It takes me a week to do a review, and then another few days for W1zz to do his editting part, and other things, while he has his own stuff to do as well. So, my next review has been done for a while, but at the same time, with me doing basically a review a week for the last little while, I haven't had much time for myself, so I've been enjoying the time off.
I guess tomorrow I'll have to work on getting more product to review for you guys, but today, I'm spending some time with my kids.
ASUS seems "on-board" with my reviews, MSI was supposed to send product a month ago, but nothing has arrived, and I think Gigabyte is afraid of my reviews.
I'd very much like to add more boards, as with only ASUS 9-series boards, the picture painted by the results doesn't offer much, does it? I mean, I could just tell you, these two boards are cookie-cutter copies of each other, with some functional differences, and that's a very good thing...these boards have left me fairly impressed. I'm going to be playing with this board, the M5A99X EVO, for the next week or so, even.
Hmm. I realise you spend a great deal of time on the reviews, but some of what you review is at least to me of little or no interest.
For example the CPU performance isn't of interest to me, because I know CPU performance across different boards is identical and any differences in benchmarks are both statistically insignificant, and likely to be unrelated to the boards. Thus I'd prefer if you put a section in saying "After testing with WinRAR, wPrime and CineBench I confirmed CPU performance was what it should have been." That might save you a bit of time?
If you're testing the board and not the CPU then you, I think, need to confirm when it meets expected performance, and report in detail where it is below or above performance. Power consumption is something you'd need to go into in detail on each board... I think... as it's likely to be different. And of course overclocking is where boards do differ.
How does that sound?
Separate names with a comma.