ASUS Crosshair V Formula-Z AM3+ Motherboard Review @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,601
ASUS Crosshair V Formula-Z AM3+ Motherboard Review - ASUS updates its Republic of Gamers lineup with a new version of the Crosshair V Formula. This time adding a Z to the name and a few other improvements as well. Is this board worthy of the Republic of Gamers brand or has ASUS left us with a sub-par version of its original Crosshair V Formula?
 
So these Vishera cores don't like 250+ bus speeds? I remember using this guide to OC mine some months ago. Then this guide came out and he seems to want to keep his at 200 like in the [H] review. I don't have the new Vishera cpu yet, but was very interested in how it OC'd compared to the older chips like mine.

Oh and on my Sabertooth you have to do the load-line calibration stuff before you do the automatic overclocking. Otherwise you end up with your results and a headache every time. The chips from AMD fluctuate wildly in overclocking potential and the cores are prone to shutdown especially in Cinebench when testing overclocks. Now to go back and read the whole review to debate if I want this board or not. Thanks again guys for taking the time to test these for us. :)
 
"The overclocking issues aside, which I can't blame ASUS for,..."

Actually in this case I really think you might have to. I would point the finger at either the board or the board's BIOS before the chip. They can go higher with less effort AFAIK,

...... Then again I wasn't there with the equipment in hand, the [H] team was, so I would just throw my point out there but I wouldn't bet money on it :p
 
Nice review, board seems pretty solid! great network perf too.

Though in that section I think one of these is supposed to say download :)
The large files upload test yielded an average transfer rate of 97.4MB/s and a high of 99.8MB/s. CPU usage did rise to 3% however. The large files upload test showcases the greatest strength of this solution with an impressive 112MB/s average and a maximum transfer rate of 119MB/s. Fortunately we didn't have to pay the price with outlandish CPU usage as it was a mere 2%.
Click to expand...
 
Action_Parsnip said:
"The overclocking issues aside, which I can't blame ASUS for,..."

Actually in this case I really think you might have to. I would point the finger at either the board or the board's BIOS before the chip. They can go higher with less effort AFAIK,

...... Then again I wasn't there with the equipment in hand, the [H] team was, so I would just throw my point out there but I wouldn't bet money on it :p
Click to expand...


No, I firmly lay that at the feet of AMD. Your mileage may vary.
 
Kyle on the mutimedia page the cinebench results for the gigabyte board with the thuban is wrong.

No way a 3.4ghz thuban scores 6.90 in cinebench ;-) The other motherboard shows the correct type of results.

I was hoping you would try and push the memory system. The formula Z has digi power for the ram. Thus supports some crazy memory speeds Overclocked of course.
 
You guys really gotta get a new chip to test FX overclocking cause that one is definitely fubar. 4.4?! My 8150 will do 4.4 on stock voltage and a Hyper 212 and these Piledrivers are supposed to be better overclockers than Bulldozer.
 
MacLeod said:
You guys really gotta get a new chip to test FX overclocking cause that one is definitely fubar. 4.4?! My 8150 will do 4.4 on stock voltage and a Hyper 212 and these Piledrivers are supposed to be better overclockers than Bulldozer.
Click to expand...

AMD chips vary wildly in ability to OC. I can easily take my 8120 to 4.6 with a H80. My nephew who has the exact same basic system as me can't get his past 3.8 with a H100. One of his cores just disappears around 4.2. We were going to swap processors to see if it was motherboard or processor problems but he's much too busy to do it. So we left his system at stock since I'm not going to be able to help him tweak it more. Needless to say he's going Intel next month just from frustration.
 
cageymaru said:
AMD chips vary wildly in ability to OC. I can easily take my 8120 to 4.6 with a H80. My nephew who has the exact same basic system as me can't get his past 3.8 with a H100. One of his cores just disappears around 4.2. We were going to swap processors to see if it was motherboard or processor problems but he's much too busy to do it. So we left his system at stock since I'm not going to be able to help him tweak it more. Needless to say he's going Intel next month just from frustration.
Click to expand...

I don't think they vary that much. I see a lot of people on OCN getting good OC's with their FX procs. I know that's not scientific and I'm sure there are duds out there but it seems there are plenty of good ones out there and I don't see a lot of complaints about bad ones.

Either way, I was only saying I think it would be better if [H] would get a good one for motherboard tests so we could see how well the AMD boards overclock.
 
You should have tried a couple cpu's in the testing to get a gold award or a 8150 that you used in the other reviews that you know what the max overclock was. Testing a cpu for the first time on a board could mean it was the cpu or board. you will never know as well as us.
 
Honest Question:

Why does Asus keep making Motherboards with the SATA pointing directly out to the side? Usually to where you have your drives sitting so it makes it so difficult to plug into those ports. If they were facing strait out I wouldn't complain. Is that a Patent or Copyright for the ports to be facing to the right and not strait up or something?
 
1/2 bent said:
You should have tried a couple cpu's in the testing to get a gold award or a 8150 that you used in the other reviews that you know what the max overclock was. Testing a cpu for the first time on a board could mean it was the cpu or board. you will never know as well as us.
Click to expand...

I only had the one CPU on hand. AMD made it clear that these CPUs could have wildly different overclock ceilings. ASUS' ROG boards tend to be as good or better than any other boards out there on the overclocking front. ASUS has been in this game a long time and I'm going to go out on a limb and state that I believe ASUS isn't to blame for the relatively poor overclock seen here. Neither Kyle nor myself got great results out of this CPU so I think the most likely culprit here is the CPU itself.
 
Last edited:
bmellis1984 said:
Honest Question:

Why does Asus keep making Motherboards with the SATA pointing directly out to the side? Usually to where you have your drives sitting so it makes it so difficult to plug into those ports. If they were facing strait out I wouldn't complain. Is that a Patent or Copyright for the ports to be facing to the right and not strait up or something?
Click to expand...

Because most people that buy these boards have larger cases where there is room for those cables, and cables going out the side is preferred by enthusiasts because it appears a lot neater. Plus, it does not get in the way of large graphics cards. Any upper-range graphics card is going to be sitting on top of the SATA ports, blocking access to them if they pointed straight up.
 
Tsumi said:
Because most people that buy these boards have larger cases where there is room for those cables, and cables going out the side is preferred by enthusiasts because it appears a lot neater. Plus, it does not get in the way of large graphics cards. Any upper-range graphics card is going to be sitting on top of the SATA ports, blocking access to them if they pointed straight up.
Click to expand...

Yep. And plugs that aren't right angled are made fun of by me in reviews. I hate the vertical ports.
 
I didn't even know there was a CHVF-Z until I saw this review. What rock have I been living under?

Dan_D said:
Yep. And plugs that aren't right angled are made fun of by me in reviews. I hate the vertical ports.
Click to expand...

I think more importantly than that is vertical ports can interfere with long GPU's on gaming-oriented boards. It makes no sense to have vertical ports that would be blocked by a big card.
 
Dan_D said:
I'm going to go out on a limb and state that I believe ASUS isn't to blame for the relatively poor overclock seen here. Neither Kyle nor myself got great results out of this CPU so I think the most likely culprit here is the CPU itself.
Click to expand...

So with no proof at all you declare it so. You did the MB and CPU no justice, no one in their right mind would purchase either after this review, as you left it COMPLETELY inconclusive. Are we really to believe the [H] had no resources to obtain another CPU (borrow/steal/buy) even a older 8150/8120 in which to verify your claim that it absolutely wasn't the MB fault?:confused:
 
funkydmunky said:
So with no proof at all you declare it so. You did the MB and CPU no justice, no one in their right mind would purchase either after this review, as you left it COMPLETELY inconclusive. Are we really to believe the [H] had no resources to obtain another CPU (borrow/steal/buy) even a older 8150/8120 in which to verify your claim that it absolutely wasn't the MB fault?:confused:
Click to expand...

I believe them. My 8150 hit 4.4 at stock voltage but got a lot more finicky above that. It'll run 4.6 (my 24/7 clock) easiest with temps in the mid 50's under IBT but I really had to tinker with it to get it higher. I got 4.8 stable enough to pass 20 runs of Intel Burn Test on Very High settings but it took a lot of trial and error and a Vcore of 1.49V and my H100 fans spinning like mad and that's definitely not what I'd call a practical overclock but then some people can hit 4.9 with just a few clicks in the BIOS. So I have no problem believing [H] just got a bum chip but I would've liked to see them get hold of a quality one because it would be fun to see what that beauty of a motherboard would be capable of.
 
Last edited:
So you believe them with neither experience with the CPU or MB in any situation that is relevant to the specifics to what is being "reviewed" by [H].
Well thanks for the input, but sorry to say it means nothing.
It is a review that consist of one specific MB and CPU. One example of each. The poorest of any example of when a conclusion must be drawn. Yet [H] did. Thanks [H] that is some great tech review. Glad you went the extra mile boys! :D
 
funkydmunky said:
So with no proof at all you declare it so. You did the MB and CPU no justice, no one in their right mind would purchase either after this review, as you left it COMPLETELY inconclusive. Are we really to believe the [H] had no resources to obtain another CPU (borrow/steal/buy) even a older 8150/8120 in which to verify your claim that it absolutely wasn't the MB fault?:confused:
Click to expand...

That kinda what I was saying in my last post except I was more defending the motherboard. They said it may have gotten a gold award if it weren't for the low overclock. They cant blame the board for that and another cpu should have been used to test as well and not a new unused cpu as you just cant tell. Use the same cpu as you did with the older Asus board or test the same cpu on both new and old boards imho.
 
funkydmunky said:
So you believe them with neither experience with the CPU or MB in any situation that is relevant to the specifics to what is being "reviewed" by [H].
Well thanks for the input, but sorry to say it means nothing.
It is a review that consist of one specific MB and CPU. One example of each. The poorest of any example of when a conclusion must be drawn. Yet [H] did. Thanks [H] that is some great tech review. Glad you went the extra mile boys! :D
Click to expand...

I believe them based on countless other ROG board reviews in which they performed stellar with outstanding overclocks. I also believe them based on my own experience with AMD processors and reading posts on several forums showing that there are chips out there that just wont overclock well, I happen to have one. I agree with you that it wouldve been a better review with a good overclocking chip. Im just saying that I believe that the chip they have is a dog and wont overclock very well. All Im saying.
 
MacLeod said:
I believe them based on countless other ROG board reviews in which they performed stellar with outstanding overclocks. I also believe them based on my own experience with AMD processors and reading posts on several forums showing that there are chips out there that just wont overclock well, I happen to have one. I agree with you that it wouldve been a better review with a good overclocking chip. Im just saying that I believe that the chip they have is a dog and wont overclock very well. All Im saying.
Click to expand...

So you don't need any proof that a CPU or MB is to blame?
I'm just shocked at how often the "I'll believe what I'm (confused) told" bracket is willing to raise their head.
Wow, HOLY CRAP, are you serious?
 
As I said before...Use the Same 8150 to test this board as was used in the older revision board.or test both boards with the new cpu. Thats the only sure fire way to test. If not then it could be cpu or board and WE don't know. The [H] is not what it used to be long ago but I trust most my testing to them. Check out other sites.
 
funkydmunky said:
So you don't need any proof that a CPU or MB is to blame?
I'm just shocked at how often the "I'll believe what I'm (confused) told" bracket is willing to raise their head.
Wow, HOLY CRAP, are you serious?
Click to expand...

Youre right. Its a big conspiracy. HardOCP is just covering for Asus. Everybody knows their motherboards are crap. :rolleyes:
 
well for starters hardocp uses Prime 95 as a stability test.

Bulldozer and Piledriver have issues with Prime 95. IBT test or the OCCT suite is much better to use with these processors.

My cpu will fail prime 95 at stock settings, it also fails overclocked. However I pass IBT and OCCT all day long. My system is stable as a rock with well over 2 weeks of uptime as of typing this with out a crash. Can go for months if the didn't install updates and other things which required reboots.
 
funkydmunky said:
So with no proof at all you declare it so. You did the MB and CPU no justice, no one in their right mind would purchase either after this review, as you left it COMPLETELY inconclusive. Are we really to believe the [H] had no resources to obtain another CPU (borrow/steal/buy) even a older 8150/8120 in which to verify your claim that it absolutely wasn't the MB fault?:confused:
Click to expand...

I didn't claim anything absolutely. You should understand that overclocking is a lot of the time luck of the draw. Sometimes you win big and sometimes you lose big. I had a Celeron 300A that couldn't do 400MHz. That was pretty rare back in the day and not in a good way. Your mileage may vary. I didn't think that needed explanation or clarification.

Given that I've seen CPUs from AMD that wouldn't overclock worth a damn and AMD even said that these CPUs could have a + or - 300MHz difference in the overclock ceiling it's pretty easy to blame the CPU. This is by no means 100% conclusive and I didn't mean to give the impression that it was. Don't forget that AMD even told us what cores to disable to achieve realy high overclocks. That right there tells me they have little confidence in the CPUs ability to clock well consistently. But we never play that game. I wouldn't disable Hyperthreading on an Intel CPU even though I know it gets higher clock speeds.

As for the motherboard I've seen a lot of ASUS boards and I've NEVER in all the years I've done this, saw one that didn't overclock fairly well. And beyond that every ROG board I've seen overclocked like crazy. In fact most of the time the best overclock's I've ever achieved were on ROG boards. And other reviews and regular people also had better overclocks with these using less effort as someone else already stated. ASUS is very very consistent with the quality of their boards. At least that's always been my experience and has been for 15+ years.

So yeah, based on information from AMD, I'm giving ASUS The benefit of the doubt here. No question. And I'll be the first to admit that my conclusion is based on only what I've worked with. Could my opinions change with a different CPU? Absolutely. But as always I have to go off of what I had to work with and what my experiences were. I only had the one FX-8350 on hand. I'd have used a different one had another been available to me.
 
My FX didnt like Prime95 or OCCT at all. I got 4.85 @ 1.425V 20 run stable in IBT and on a whim last night after reading this thread, I fired up Prime95 and 3 cores dropped out within seconds! Tried OCCT and it crapped out in seconds. :eek:

I stayed up til 3AM trying to get it to work. Ended up 4.6 was the highest I could get stable under Prime95 and OCCT even trying voltages as high as 1.50 which is WAY over what is recommended max safe. At 4.6 and 1.46V I could run OCCT at 60C. This morning I got up and decided to try different fan/rad configurations and that included re-seating the proc and re-applying some TIM. Well the stuff that comes pre-applied on is apparently pretty good cause I shot up to 64C in seconds the first time. Pulled the water block off and re-applied the TIM. Did the spread method the first time and did the rice grain method the 2nd time. I have re-seated the CPU FIVE times and tried a couple different variations of fans and rad and after about 10 hours of work (I really need to get a life) Im at 4.5 and 1.44V because I hit 63C at 4.6 now that Ive taken the pre-applied stuff off my H100. Hopefully Ill get a little better temps when this Arctic Silver 5 cures and I can bump back up to 4.6 cause Im feeling pretty stupid for buying this H100 last week only to go from 4.4 to 4.6. :(

Apparently IBT and linpack in general doesnt bother FX procs but Prime95 and OCCT just rip it a new one. And it seems like there is something about FX procs that makes them cruise right on up to something like 4.4 on stock voltage like a boss then just become a dumpster fire (almost literally with the heat) at .1 beyond that. I mean how can you run 4.4 at 1.375V but need 1.46V to hand 4.6?

Ill keep tinkering with it and see if I can get a little better performance but it looks an awful lot like my own personal results are the same results Dan was getting with his FX. Now where's that $99 2500K thread.........
 
That's interesting. I just ran Prime95 and OCCT like I always do. Otherwise it worked for general tasks at 4.6GHz (even did a pass at wPrime and Hiper Pi) and even to a degree at 4.8GHz. As for the heat wall that you reach on the AMD processors, it's the same with Intel. Sandy Bridge-E and Ivy Bridge both are good with minimal adjustments in voltages to about 4.4GHz. After that you'll have dramatically increase voltages to go higher. Ivy Bridge will run at 1.2v or less at 4.4GHz and require 1.25v+ to hit 4.8GHz. On the AMD side our processor needed 1.48v to do 4.6GHz with any kind of stability.
 
I'm not giving up in it yet. When you drop a C note on a cool new cooler, you get motivated to get something out of it. Maybe if I play with RAM timings I can squeeze a little more headroom out of this thing.
 
My FX didnt like Prime95 or OCCT at all. I got 4.85 @ 1.425V 20 run stable in IBT and on a whim last night after reading this thread, I fired up Prime95 and 3 cores dropped out within seconds! Tried OCCT and it crapped out in seconds.

I stayed up til 3AM trying to get it to work. Ended up 4.6 was the highest I could get stable under Prime95 and OCCT even trying voltages as high as 1.50 which is WAY over what is recommended max safe. At 4.6 and 1.46V I could run OCCT at 60C. This morning I got up and decided to try different fan/rad configurations and that included re-seating the proc and re-applying some TIM. Well the stuff that comes pre-applied on is apparently pretty good cause I shot up to 64C in seconds the first time. Pulled the water block off and re-applied the TIM. Did the spread method the first time and did the rice grain method the 2nd time. I have re-seated the CPU FIVE times and tried a couple different variations of fans and rad and after about 10 hours of work (I really need to get a life) Im at 4.5 and 1.44V because I hit 63C at 4.6 now that Ive taken the pre-applied stuff off my H100. Hopefully Ill get a little better temps when this Arctic Silver 5 cures and I can bump back up to 4.6 cause Im feeling pretty stupid for buying this H100 last week only to go from 4.4 to 4.6.

Apparently IBT and linpack in general doesnt bother FX procs but Prime95 and OCCT just rip it a new one. And it seems like there is something about FX procs that makes them cruise right on up to something like 4.4 on stock voltage like a boss then just become a dumpster fire (almost literally with the heat) at .1 beyond that. I mean how can you run 4.4 at 1.375V but need 1.46V to hand 4.6?

Ill keep tinkering with it and see if I can get a little better performance but it looks an awful lot like my own personal results are the same results Dan was getting with his FX. Now where's that $99 2500K thread.........
Click to expand...

depending on when your H100 was made they used to use Shin-Etsu which is along the best tim available. So the pre applied Tim was better than arctic silver 5. They switched recently to use a different type of Tim however it is still quite good. Arctic silver 5 is old and is bested by numerous newer compounds these days. Most of the high end coolers come with a decent Tim nowadays.

Arctic silver 5, while its better than white silicon greese, it doesn't really compare to the newest stuff available. I only use it now when i make batches of thermal epoxy, to mount heatsink's to ram modules and mosets. Don't ask that's for another day.

Dan not quite sure is up with your overclocking efforts on the FX processors. Maybe tweak one for a weekend or two. They are quite picky. Even ram timings that are stable with one overclock may make your system unstable when the cpu reaches higher speed. For example, at 4.5ghz i can set my ram to run 1t, at 9-9-9-24 @ 2006mhz. However at 4.9ghz 1t ram is impossible for stability, i have to go to 2t, same timings though of 9-9-9-24@ 2006.

To be completely honest i was stuck at 4.6ghz for awhile. Kept thinking it was the max for my 8120. Until one day I decided I would make it go higher. After hours and hours of tweaking and stability testing, i finally landed at 4.9ghz. Was a pain in the ass, and i just about gave up for past 4.6ghz, but finally made it work. Keep the processor under 50c full load if you can, that helps alot.
 
ebduncan said:
Dan not quite sure is up with your overclocking efforts on the FX processors. Maybe tweak one for a weekend or two. They are quite picky. Even ram timings that are stable with one overclock may make your system unstable when the cpu reaches higher speed. For example, at 4.5ghz i can set my ram to run 1t, at 9-9-9-24 @ 2006mhz. However at 4.9ghz 1t ram is impossible for stability, i have to go to 2t, same timings though of 9-9-9-24@ 2006.

To be completely honest i was stuck at 4.6ghz for awhile. Kept thinking it was the max for my 8120. Until one day I decided I would make it go higher. After hours and hours of tweaking and stability testing, i finally landed at 4.9ghz. Was a pain in the ass, and i just about gave up for past 4.6ghz, but finally made it work. Keep the processor under 50c full load if you can, that helps alot.
Click to expand...

Unfortunately I do not have the time on a review to spend days and days trying to fight for higher overclocks. I do what I can in the time alloted, get it stable, beat on it some and then move onto the next review.
 
nice review kyle the board looks good however when will we get a more complete picture from you on the fx 8350? we know it isn't perfect and does not match intel in ipc when the intel chips are overclocked and the amd chip is left stock. however is it better than say a k10.5 or k11 for a current amd user that does not have the cash to get locked into intels dead ivy bridge socket? looking for you to toss the amd crowd at [H] a bone with some meat on it rather than leave us out in the cold hungry.

anyways Ive seen many 8350's hit 5ghz+ over here http://www.overclock.net/t/1318995/official-fx-8320-fx-8350-vishera-owners-club
 
Last edited:
Dan_D said:
Unfortunately I do not have the time on a review to spend days and days trying to fight for higher overclocks. I do what I can in the time alloted, get it stable, beat on it some and then move onto the next review.
Click to expand...

You can send it to me and I can crank on it for ya! Free of charge! :D :p

I actually think I might have found some of the problem. I was overclocking the CPU/NB at the same time as the base clock despite knowing better. I set the CPU/NB clock and voltage back to default and am running OCCT right now and showing 56C at 4.5 GHz @ 1.45V. Still having to use a lot of juice but my temps are a lot better and thats with my H100 on Medium. Im not using AUTO on the CPU/NB voltage but manually went in and set it to 1.15 thinking maybe the board was bumping up the voltage a little with LLC. So far so good. Im gonna let OCCT run for another hour and if it passes, Ill keep going and see what I can do. Maybe its the CPU/NB thats adding the heat and maybe even affecting the stability.
 
Dan,

In my experience with AMD procs, the NB can quickly be the weak link. If you run the NB up, and apply very little added voltage, and use the LLC, you may have better luck. I have also found that the NB clocking can become a limitation on max cpu OC, due to heat as well as timing issues. I have never had a great OC without some level of NB OC and asome LLC on the NB. It seems that, for me, the NB and CPU clocks have some relative ratio required for stability.

I really do believe that you are spot on in assessing that AMD processors plus your relative new-ness to 990FX overclocking is the fault here.

Hopefully this difficult experience is viewed as an opportunity to learn a bit, not as a chore to be worked through, by the staff.

Not everyone (few ppl probably....) enjoys the 90+ reboots and fails required to learn your specific board and procs "sweet spot" for a mere 10-20% gain.

I do. I hope you do as well :)
 
I did play with the north bridge settings. I had tried maxing out load-line calibration and stepping it down a couple notches below that as well. I'm not a stranger to overclocking AMD systems in general nor the 990FX. Vishera was the new element for me.
 
The CPU/NB is definitely the bitch of this. I set mine back to 2000 MHz, the voltage at 1.125 and I'm OCCT stable at 4.615 hitting 58C with my H100 set to medium. I can't get 4.7 to pass OCCT even with voltages up to 1.49. I did find that I seemed to get better stability with LLC set to Auto believe it or not. Hard to believe anything works best on Auto but I guess that's why Asus rules! ;)

Cranking the CPU/NB up any at all really made it harder to get stable and changing the voltage up to just 1.25 made the temps shoot up a good 4-5C!

So after a good 2 weeks of banging on this thing I'm gonna say 4.615 GHz at 1.41V with stock CPU/NB is the best I can do. Not shabby I guess as its a 1GHz overclock and its more than enough CPU muscle for my gaming rig and resolution but I still wanted 4.8 at least. :(
 
I'm waiting on a rev 2 Sabertooth, and 8320 from the egg. Once they are in, I think i'll shoot for 4.1/4.2, with a 2800NB and 24-2800 HT. I believe that that will give me a better balance for the time when 6+ cores are pretty loaded, and memory transfers and I/O become contention points for slowdown. I need to read up on whether the vishera NB starts piling on latency and when. If there is a point where going faster on NB/HT forces added latency, then all that extra heat and power gets a diminishing return.

If ~2800 doesn't add latency, then I will gladly trade some core clock for more consistent performance under varying computational load types.
 
A guy on OCN did a bunch of benches with a 8150 with a bunch of different combinations of core clock, CPU/NB and HT Link speeds and the best results came with them both at 2500'ish and both running the same speed. Above that and there were no gains.

Getting 2400 on mine was what caused the issues cause it took more voltage and that really drove the heat up.
 
You must log in or register to reply here.
Back
Top