ASRock Fatal1ty 990FX Professional
BIOS DOWNLOAD
USER MANUAL
Attachment 119921
Attachment 119922
Attachment 119923
Attachment 119920
Printable View
Fatal1ty UEFI SETUP UTILITY
Attachment 119948 Attachment 119949 Attachment 119950 Attachment 119951 Attachment 119952 Attachment 119953 Attachment 119960
Attachment 119961 Attachment 119962 Attachment 119963 Attachment 119964 Attachment 119965 http://prohardver.hu/dl/upc/2011-09/...111155_thb.png http://prohardver.hu/dl/upc/2011-09/...111203_thb.png
(You can easily create screenshots to an USB pendrive by pressing F12.)
* UPDATE *
From UEFI version 1.44 you can select a different background. So you can say good bye to magnificent John.:D
http://www.xtremesystems.org/forums/...1&d=1316508876 http://www.xtremesystems.org/forums/...2&d=1316508877
The good:
There is a 26-27 millimeters space between the CPU socket and the first DIMM slot (MSI 990FXA-GD80 has 22 mm for example). Thanks to this you can use four modules with heatsinks beside a big cooler (mine is a TRUE Black).
Attachment 119968 Attachment 119969
Anyway I think that this mobo's layout is really good.:up:
If you set the Load-line calibration to 50% then you get (as)rock stable voltages. :D At this precise moment I only have a Phenom II X840 CPU. I set the CPU VID to 1.500V and then I searched
a proper measure point.
Attachment 119942
IDLE:
Attachment 119943
100% LOAD (Prime95 small FFTs):
Attachment 119944
:up:
The FAN controller is very good. Gigabyte should copy that.:p: The whole UEFI is pretty nice as well.
The bad:
- SATA AHCI performace is poor compared to an MSI 890FXA-GD65.
Attachment 119945
Same SSD, same AMD AHCI driver. Link Power Management is disabled. Hopefully it will be fixed soon via an UEFI update.
- It doesn't really like the four DIMM configuration. I have four matching Dominator GT modules equipped with Elpida Hypers. The system is 100% stable with two modules at 1600MHz/CL6 but it is only stable at 1600/CL8 with all of the four modules.:shrug:
Minor issue(s):
- The WOL (Wake on LAN) doesn't work after power loss.
I think that most of the above issues can be fixed via an UEFI update.
....so far. More to come. ;)
Attachment 119946
Looking good! Nearly like Asus :D
I hope you've got BD to test on it as well ;)
nice lookin board with strong specs, thanks for sharing... :)
may i know how does d board perform on memory oc? ;)
Gold Capacitor look good
I think it's better than ASUS. :DQuote:
Originally Posted by Lightman
Unfortunately I haven't got BD.
Quote:
Originally Posted by ah_khoo
Thanks. :) As I stated above the OC is pretty bad with 4 modules unfortunately. I'm going to check it with 2 modules soon.;)
The board looks like rampage 3 formula meets crosshair 4 formula, not really a bad thing i guess.
That fatality background EFI needs to go.......
Loos like only 2 way SLI with triple slot vga's, not really a deal breaker since 2 way is more that enough with the top end cards that would be tri slot.
The caps are interesting, can you do a close up of them?
I hardly see any AMD board does well on 4 dimms oc. Thx, I'm lookin forward to that... :)Quote:
Originally Posted by Oliverda
Attachment 119973Quote:
Originally Posted by chew*
I've got one of these coming for review, but it's still at the site owner's place and he's been under the weather and not been able to get to the post office :( I'm really excited to get it though!
I don't have really good modules to test with unfortunately, timings OR overclocking, but my 2x2GB mis-matched Micron D9 (1.9V) with PSC (1.65V) does 4x2GB 1600 7-8-7 (restriction of the PSC modules, which is a long story itself lol). I was able to do that on my 555BE and 1090T on this old 890FX Deluxe3, so I think it might be a combination of CPU used and a little bit of motherboard thrown in.Quote:
Originally Posted by ah_khoo
I agree with you Chew on the UEFI background :S
EDIT: Olivera-- try getting a reading from this spot, where it says VT8 (or might be a 6, can't tell). I've discovererd those are Voltage Test locations, which I'm sure people knew about but I wasn't aware of haha Anyways, I've noticed they are pretty standard on a lot of boards now (my Giga and DFI S939 both don't), all with the same "VT#" indicators. I think it's worth mentioning that not all of the locations, like on my 890FX Deluxe3, are small double-pad solder points for like SMD resistors or caps, but most of the time just a single round dome of solder.
EDIT2: Holy crap I can't believe I uploaded the version where I didn't point out where I was talking about >_< lol
Attachment 120064
Could you please test memory overclocking (using only PSC) with 1:4 divider? With timings like: 6-8-6 1T, 6-9-6 1T and 7-9-7 1T.
Which CPU do you use? Thuban or Deneb C3?
Thanks. I'll try it on the weekend. ;)Quote:
Originally Posted by Formula350
I only have Elpida Hypers (Dominator GT CMG4GX3M2A2000C8) at the moment but maybe I can get some PCS equipped modules. Is it a PSC kit?Quote:
Originally Posted by superpyton
Anyway with two sticks of Elpida Hypers it's stable at 1600/CL6 but it doesn't really like the higher frequencies. Please note that I have a Propus C3 based CPU at the moment.
Isn't the whole point of those really gay ram clips that the VGA doesn't get in the way? Kind of defeats the point when there is plenty of room and you put them on upside-down.
Very creepy bios.
No Problem :PQuote:
Originally Posted by Oliverda
I hope they are PSC because that's the kit G.Skill is sending me, but I don't honestly know :\ I've tried to pry some of that info, but it might be company policy not to say since it can (and probably will) change over time.
Are you trying to just crank the RAM speed and leaving everything else? Because for 1866 it is recommended to run the CPU-NB @ 2800MHz. I can't remember what the formula is to figure it out, but whatever it was it calculates out where 1333=2000 which is default Phenom II, and 1866=2800 which is def for Dozer and I think Llano (sans the dual core models, which the memory controller is only 1600)
They aren't upside down :P The permanently locked ones are indeed at the bottom.Quote:
Originally Posted by El Gappo
Do you have to use the UEFI or does it provide an optional standard BIOS layout?
This can be PSC but i'm not sure. Sorry. :)Quote:
Originally Posted by Oliverda
I've asking Formula350 for testing, but if you also could check this board with different memory then do this please :).
Older ASRocks (like 890GX Extreme3 or 4) with C3 can't boot on 1:4 CL6/7 with Hypers, so there's some progress. :)Quote:
Originally Posted by Oliverda
everyone already complained about that and so they patched in the uefi option to change the background to just black.Quote:
Originally Posted by chew*
The board layout puts a smile on my face. I also can't believe I'm seeing PS/2 connectors. I could actually use this old, but fantastic keyboard of mine with it. The BIOS background of Fata1ity creepily looking at us is a problem, but a minor one. If the price is right, I'd consider this board. I currently have my heart set on an ASUS Sabretooth.
I have this mobo with a AMD Athlon II 455 x3 unlocked to x4 until BD launch :) but i can't unlock L3 cache :S
Anyone know if this mobo can unlock L3 cache ?
Thx!
Probably your CPU is based on Propus/Rana core and it doesn't have L3 cache. ;)
IF the 2x2GB 2133 9-11-9-24 modules G.Skill is sending have PSC, I can test it properly, sure. Otherwise, the PSC-based modules I currently have my 'boss' wasn't able to get them running at their advertised specs of 2200 @ 10-11-10, but I can't verify that since the Deluxe3 only has 1600 divider.Quote:
Originally Posted by superpyton
While unrelated, the Micron D9's I have, and the PSC modules I mentioned above, both do 7-7-7 and 7-8-7 @ 1600 just fine on my old 555BE. Mind you, that is on my Deluxe3, but it still doesn't matter since neither of them are Hypers. hehQuote:
Older ASRocks (like 890GX Extreme3 or 4) with C3 can't boot on 1:4 CL6/7 with Hypers, so there's some progress. :)
EDIT: Oh, hey Oliverda... Could you do me a big favor, since you have an X-Fi? ^_^ All of their 890FX boards (and 990FX unfortunately, by the looks of it) have a NB heatsink that was poorly designed, not allowing for any fitment of a long PCIe x1 card in that first slot. I'm not able to tell definitively by the pics I've seen, but will a sound card fit in the first slot and still clear the NB heatsink? I know it isn't necessarily the ideal spot for a sound card, being in the middle of all that noise, but IF a person were to be silly and use that PCIe x16@x4 for a graphics card then that first slot would be all that is left for discrete audio :\
That point is for the CPU_NB voltage. I found an another VT point (VT1) which is for the cores.Quote:
Originally Posted by Formula350
Attachment 120197
Yes you do. There is no any other layout so far.Quote:
Originally Posted by FlawleZ
I haven't got that PSC kit unfortunately.:(Quote:
Originally Posted by Formula350
I have an earlier PCI X-Fi so I can't say that. Anyway I think you can't put a longer card into that slot so you forced to use one of the longer x16 slot for that purpose.
I sent an email to ASRock support and they sent me an updated UEFI where you can switch on/off the two LAN and the three USB 3.0 controller separately. :up: They sent me and updated version of F-Stream utility as well which now can use the POST CODE led to display the current CPU temperature. I really satisfied with the support so far.
I like those added BIOS options! Sadly my motherboard is mounted on my DIY techbench in a way that will not allow me to see the Diagnostic LED very well (upside down, for starters), BUT I do quite like that option! I think they should have it auto-switch to the temp after the system has been in "Everything is Fine" mode for a certain amount of time, or at least make a small program that can load up with Windows to do it.
Really sucks to hear about the first PCIe x1 though :\ Only thing that'd really fit is a stand-alone audio codec like what MSI and ASUS offer on some boards, which I really think ASRock should've done in the first place, or a discrete LAN/USB3 card... The latter two make zero sense though, as why integrate 2 onboard LAN controllers as well as three USB3 controllers?! :shakes:
Thanks for the info though :)
I asked the support for an UEFI option for that. Let's see what they can do with it.:)Quote:
Originally Posted by Formula350
I think that the 2nd LAN could be useful in some cases, but I agree that the 2nd onboard USB3 connector is unnecessary.Quote:
Originally Posted by Formula350
Anyway I checked the VRM controller chip and it's a CHiL one which can measure many things. I've read that the ASUS uses the same controller under DIGI+ name.:D
http://prohardver.hu/dl/upc/2011-09/13063_chil.png
It would be very good if the mobo could report the CPU's current.:)
I hope they get implemented (I thought you had said they sent you a BIOS with those options already :\)Quote:
Originally Posted by Oliverda
Yes, I had seen that awhile back in a large board photo! I was quite excited about it's use, so I seriously hope ASRock doesn't drop the ball by not taking advantage of all the options available by the chip. So I've got my fingers crossed...Quote:
Anyway I checked the VRM controller chip and it's a CHiL one which can measure many things. I've read that the ASUS uses the same controller under DIGI+ name.:D
It would be very good if the mobo could report the CPU's current.:)
Also, as to not hijack the 'Dozer thread:
I see ASRock didn't make too many changes in the BIOS. That is 100% the same menu option and ordering as in my 890FX Deluxe3 :rolleyes: I hope the Fatal1ty isn't basically a GUI overtop their old 890FX line's BIOS :\ I want near-C5F BIOS options, dag nabbit!! lolQuote:
Originally Posted by Oliverda
Sounds good, but what about the MOSFETs? They advertise with 12+2 phases, but that ChiL chip can only handle 6 or 8. Which chip is it? The 8326 or the 8328? I heard once, that mainboard manufacturers always use a 2way parallel setup, i.e. the mainboard should have 12+2 divided by 2 (real) phases, thus 7.Quote:
Originally Posted by Oliverda
Anyways I am puzzled if these are of the same quality as the chips Gigabyte or MSI are using. Gigabyte is often advertising with ferrit chokes and low-rds mosfets, while MSI advertised "DrMos", which is probably not bad either. I don't like all that marketing BS, but low-rds sounds good, as these mosfets are not heating up as much, if I understood it correctly.
Anyways, in short: Can you give a quality rating for AsRock's mosfets?
Thanks
Opteron
P.S: And yes, AsRock's BIOS support is one of the best, you can actually talk to someone, and even better, you get a reply ! ;-)
Quote:
Originally Posted by Opteron146
It's the CHiL 8328 and yes it's "only" seven phases. 6 for the CPU cores and 1 for the CPU-NB. All phases have two chokes (=14 chokes).
Actually I can't describe the quality of the components but it's not DrMOS. I can only tell you that the CPU voltages are rock stable
which is very rare nowadays unfortunately.
Attachment 120229
The chokes are MPFC (Max Power Ferrite Choke), with a capacity of 50A, X58A-OC has also chokes of this type.
The metalic clips are their sign.
Attachment 120241
Asrock implemented also in Z68 EXTREME7 GEN3 these chokes.
There is no such thing as 12-24 phases for CPU VRM.
There are 4, 5, 6 phases controlers, and the number of chokes/phases can be 1, 2, 3, 4( 6*4= 24 Gigabyte VRM).
Mem, IGP, NB have aditional 1-2 phases and chokes afferent.
EDIT: More chokes doesn't mean always better and more stable power to the cpu.
Evga designs and Asus MIV prove that.
About MOSFETs, from what i see Asus CH V uses normal mosfet's?
Attachment 120242
this is part of the reason why asrock is the best motherboard maker right now. they also have unofficial updates for motherboards that are YEARS old.Quote:
Originally Posted by Oliverda
if you guys want a lower end motherboard with fewer features, they do make those. please don't dump on the high end. some of us would like to continue seeing high end motherboards.Quote:
Originally Posted by Oliverda
Sounds good, but out of curiosity I want to know if it is low-rds or not. Can you see the labels of the mosfets? Maybe we can pin down the type.Quote:
Originally Posted by Oliverda
Thanks, is there any visible difference between mosfets and low-rds mosfets? I don't think so, but I just want confirmation.Quote:
Originally Posted by xdan
Furthermore, if we got some part numbers from the label, do you think we can find some information about the part online?
Thanks again
Opteron146
P.S: Maybe the easiest thing would be to ask the AsRock support about it ^^
* * * UPDATE * * *
I've just got an updated 1.44 version UEFI directly from a support guy called Peter.
1. Now you can select a background for UEFI. So you can say good bye to magnificent John.:D
Attachment 120291 Attachment 120292
2. If you set the Onboard Debug Port LED option to ON then the display LED will show the mobo measured CPU temperature after boot up.
Attachment 120293
3. You can enable/disable the two LAN ports and the three Etron USB 3.0 controllers separately.
Attachment 120294 Attachment 120295
4.
From now you can see the actual AGESA version in the CPU configuration section.
Attachment 120296
He promised an updated version of F-Stream Tuning utility which will show the CPU's current (and wattage).:D
ASRock's support is great!:up:
im so torn on this vs the chv or sabertooth... any thoughts comparing them? since they are so close on price. i like the fact the sabertooth has 8 sata on board but two are jmicron which kinda sucks i know but on my intel builds i always end up using them all. i am however staying away (i think) from msi and giga after past issues with them. and seems dfi is out of the game..
what about slot layout do you guys this is a better overall one than the asus' are? since this has 2 pcie 1x which imo have little use and as said that top on is limited. this is my first amd build in years so i dont know a whole lot of exactly what i need/want in a am3+ board..any advice would be great but this board does look sweet and now that we can change the bios im almost sold
Some tests with Thuban and PSC.
Test setup:
- Phenom II X6 1055T
- Zalman CPNS10X Extreme
- ASRock Fatal1ty 990FX Proffesional (EFI: P1.20)
- A-Data Gaming 2000 MHz CL9 2x2GB
- Sapphire HD2600 PRO 256 MB DDR3
- Seagate Baracuda 7200.9 80 GB SATA 2
- Zalman ZM-500HP
- Windows 7 HP x64
Tests - EFI 1.20 (memory divider: 8x/1:4, CPU-NB multiplier: 10x):
- 1600 MHz 6-8-6-21 1T 1,65 V (CPU-NB Voltage: 1,25 V) - HyperPi 32M stable
- 1800 MHz 7-9-7-24 1T 1,65 V (CPU-NB Voltage: 1,25 V) - HyperPi 32M stable
- 2000 MHz 8-10-8-27 1T 1,65 V (CPU-NB Voltage: 1,25 V) - HyperPi 32M stable SCREENSHOT HERE
- 2040 MHz 8-10-8-27 1T 1,65 V (CPU-NB Voltage: 1,3 V) - unstable (12 loop)
- 2040 MHz 9-11-9-30 1T 1,65 V (CPU-NB Voltage: 1,35 V) - unstable (1 loop)
- 2040 MHz 8-11-8-27 1T 1,65 V (CPU-NB Voltage: 1,35 V) - unstable (3 loop)
Now some test with EFI 1.44 - thanks Oliverda :)
Tests - EFI 1.44 (memory divider: 8x/1:4, CPU-NB multiplier: 10x):
- 2040 MHz 8-10-8-27 1T 1,65 V (CPU-NB Voltage: 1,35 V) - HyperPi 32M stable SCREENSHOT HERE
- 2080 MHz 8-10-8-27 1T 1,65 V (CPU-NB Voltage: 1,35 V) - unstable
I like this slot layout. I like having 2 PCI slots because I have a LOT of old PCI devices I could pull out at any moment. I like supporting 2x graphics cards. I like supporting a RAID card on the bottom 16x slot (only 4x lanes). I don't mind the 1x slots because I know there aren't enough lanes to do any better. Already the bottom 16x is actually a 4x. So surrounding the main 16x with two 1x is actually a convenience and is utilitarian, because it leaves room for different graphics and addon configurations, because they could have otherwise left those spots empty or put a PCI slot there which would be wasted because I want a PCI more than a PCI-E 1x. I independently analyzed all these slot configs and ended up agreeing with the one on this motherboard. If I have more PCI-E lanes to work with, or if I don't want PCI legacy support, I would pick a different config. But for me today, this is it.Quote:
Originally Posted by armada
ya know i actually never noticed they dont do 16x 16x 8x. why on earth leave the last slot with only a 4x? look at the msi gd80 for instance. i just dont know if i could buy a gd80 after my last rma experience with msi last year.
http://shareimage.ro/images/3fr2f76l9e89rdacwc9.jpgQuote:
Originally Posted by Opteron146
Low-rds mosfets are not something that spectacular.
They work at 16% lower temperature, providing a lower power consumption and heat than standard mosfets.
But i think that 16% is at their best, the difference i don't thinks that do anything.
MOS mosfets or tantulum mosfets(these are the best), bring some real difference.
EDIT: but from what i see, i just look now on Gigabyte low-rds mosfets page, Asrock has low rds mosfets too...
Low mosfets have 4 pins, standard two...
Or they must be that short too? ...:D
I think that if you can read what is labeled on them, you can find information on net.
Anyway i think that Fatal1ty 990FX Professional has a very strong VRM, just that it is.
In 1-1.1/2 years i think Asrock would implemed DR. MOS too or tantulum, and dual bios, these is just that it's still missing from Asrock mb's, in rest they have all.
http://lab501.ro/wp-content/uploads/...22-580x385.jpg
http://prohardver.hu/dl/upc/2011-09/13063_pwm.jpg
i Hate to come in here and correct you, but there is a way to have something close to true phases. Those are phase doublers. if you see DrMOS configurations, you will see that each of those Drmos chips requires its own PWM durect connection. The low RDS(ON) on the ASUS, ASrock don't they just can use the same driving signal per double teh number of MOSFETs. Actually most ASUS boards do it this way, the M4E uses 4 DirectFET per phase and one inductor, the M4G-Z uses 3 FETs, 2 low side, 1 high side, but the M4G-Z is a bit more precarious.
The M4G-Z only uses 4 PWM phases, but instead of driving 6 MOSFETs with a singal driver, ASUS implements another IC, this one IC contains 2 drivers and a doublers, so for the CPu phases the M4g-Z has 8 phases, 8 drivers, 3 MOSFETs per phase, and 8 inductors.
There are many ways to correctly double the phase count, one way is to double the number of PWm signals, with doublersl ike on the M4g-z and OC board. in the case of the asus, it divides the switching frequency in half. On the Sniper 2 its divided in half as well on most GB boards, but i don't know much about the ASUS one, but i know the GB one is capable of not having to divide the SF in half when only one set of the doublers is used, not 2 like on the UD7,.
http://www.intersil.com/data/fn/fn7564.pdf
The cheap way to double phases is to not use a driver per phase and just use two phases per driver, as we see on many cheaper boards. I count phases by the number of drivers and pwm signals, but a tru phase would be one directly from the PWM, so ussually 4 or 6 or 8.
Latest stuff from the support:
UEFI 1.44
F-Stream utility with current/wattage and additional HDD led support.:D
i always love current and wattage monitoring when its correct, and the Chil is prob one of the most correct out there.
Attachment 120532Quote:
Originally Posted by sin0822
:yepp:
AIDA64 will probably get a support for this too.
I have the P67 Fatal1ty and love it and I was looking forward to a 990FX version for BD ... but my Asus Xonar needs to fit in the Top PCI-E and I dont think it'll fit due to the heatsink below the CPU socket (even though it fits in the P67 fatal1ty, that has a low heatsink) ... looks like the 970 Extreme4 might be the choice for me ...
http://asrock.com/mb/photo/Angle/970%20Extreme4(m).jpg
Are you going to have three x16 PCIe card?Quote:
Originally Posted by SBB
No, just SLI cards, but the sound card in the bottom would block air to the 2nd cards fan, and I dont really want to do that when having the sound card above is a much better solution :(Quote:
Originally Posted by Oliverda
Then buy a PCI sound card.:DQuote:
Originally Posted by SBB
Nice try, but the PCI on this board also has the same problem ;)Quote:
Originally Posted by Oliverda
Really. :)Quote:
Originally Posted by SBB
Anyway are you sure that the soundcard in the last PCIe slot would block air to the 2nd cards fan?
yeah :( it'd be in the slot next to it and i'm using these cards
http://xtreview.com/images/gigabyteg...%203X%2001.jpg
I'm seriously considering this board over an ASUS Sabretooth now. I hope the price is right $149.99 or in that ballpark. You have to love a company that listens to its customers. The high quality components on the board put a smile on my face. Thanks to all the guys that identified the components. One question though, do we know yet if the board has any power protection like over/under current/voltage for those pesky power fluctuations during storms? I recently lost my ASUS M3A79-T Deluxe mobo to a jolt that made it past my Cyberpower 1500VA UPS and PCP&C 750 Silencer II. It hit before the UPS or even breakers could trip and ruined my day.
All right. Unfortunately the 970 Extreme4 isn't capable of x16/x16 CF or SLI. You should get a 990FX mobo.Quote:
Originally Posted by SBB
Don't hesitate too much. :) It's better than Sabretooth and your X-Fi Fatal1ty will be love it. :DQuote:
Originally Posted by Mechromancer
I don't know anything about such protections of the mobo but I would recommend you a Seasonic X PSU.;)
I appreciate the recommendation Oliverda. I already have another Seasonic M12 700w that I suppose I should put back in. They are great PSUs!
The 970 Extreme4 supports 8x/8x on SLI, which is the same as my current P67 board, would I notice any improvement going to 16/16?Quote:
Originally Posted by Oliverda
Few percent. If I were in your place then I would go for an 990FX mobo.
I think the biggest question is: Do you get above 60FPS in the games you play right now? Otherwise I'm not sure it'd be all that worth upgrading if that's the only reason :)Quote:
Originally Posted by SBB
Hopefully this mobo is as good as the Fatal1ty, because this one looks much better imo.
http://www.asrock.com/mb/photo/990FX%20Extreme3(m).jpg
ASRock 990FX Extreme3
Maybe on color-scheme,cause the VRM looks Waaaaaay too weak...
http://img225.imagevenue.com/loc959/..._122_959lo.jpg http://img272.imagevenue.com/loc104/..._122_104lo.jpg http://img196.imagevenue.com/loc92/t...3_122_92lo.jpg
http://img229.imagevenue.com/loc81/t...4_122_81lo.jpg http://img206.imagevenue.com/loc223/..._122_223lo.jpg http://img211.imagevenue.com/loc1153...122_1153lo.jpg
http://img157.imagevenue.com/loc110/..._122_110lo.jpg http://img281.imagevenue.com/loc24/t...0_122_24lo.jpg http://img159.imagevenue.com/loc249/..._122_249lo.jpg
http://img245.imagevenue.com/loc206/..._122_206lo.jpg http://img244.imagevenue.com/loc247/..._122_247lo.jpg http://img168.imagevenue.com/loc837/..._122_837lo.jpg
http://img245.imagevenue.com/loc579/..._122_579lo.jpg http://img298.imagevenue.com/loc459/..._122_459lo.jpg http://img180.imagevenue.com/loc681/..._122_681lo.jpg
http://img190.imagevenue.com/loc537/..._122_537lo.jpg
Pictures of my Fatal1ty 990FX Pro
Cpu, waiting for BD?
Still with a solid 555BE as B55 X4 4GHz,
waiting for the BD...
Please put that board through its paces. I want to see how well it treats you before I make a purchasing decision.Quote:
Originally Posted by Valg
New UEFI is on the web: 1.30.
Lacking some features of 1.44.
Just saw that (new?) MSI board:Quote:
Originally Posted by xdan
http://www.msi.com/product/mb/990XA-GD55.html
Quite cheap, but it seems to have the GD80's VRMs, i.e. Tantalum, DrMos and SFerrit Chockes. high-res picture:
http://www.msi.com/pic/product/five_...0916115927.jpg
What do you think about that? $135@newegg:
http://www.newegg.com/Product/Produc...82E16813130620
wow those MSI VRM look good
Msi 990fxa-gd80 or Asrock 990FX Professional ? That is the question :D
I heard this board is good with Zambezi. Anybody have experience with this combination yet?
ASRock Fatal1ty 990FX Professional !
i have been with asus boards and seems like i cant get away no matter what, all the reviews i have read and see this ASRock Fatal1ty 990FX Professional board seems to have some MAGIC to it.
im ready to buy and switch to 1100T x6 phenom.....i still have my crosshairIV extreme.
can i hit 4.3 4.4 or better ghz with a 1100T phenom im phazechange cooled -44c
thanks ASrock people.........
I am really waiting to see results from the Fatal1ty 990FX Pro compared to the Crosshair V with Bulldozer CPU before making a purchase.
Want to see if the reported higher performance in Fatal1ty 990FX Pro compared to other mobo manufacturers is true or not.
Unfortunately for me I dont like the look of the Fatal1ty 990FX Pro where I think the Crosshair V looks fantastic.
Hope to see some results soon!
Latest BETA UEFI: 1.47
(I'm going to make a comparison review about Crosshair V, 990FX Sabertooth, MSI 990FX-GD80 and ASRock Fatal1ty. Unfortunately it can only take place in November.)
Quote:
Originally Posted by Oliverda
Thank you for the provided BIOS. Please, don`t forget to update them like now. :)
Thak You for the BIOS update! :)
Did You notice any changes yet?
I noticed few so far:
- mouse dont work in UEFI any more (jumps to up-left corner and clicks by itself when draged over a button) -> but it is calibrated right
- the good night LED turn of the diagnostic LED over night time (no more temperature is displayed anymore) -> must be enabled in UEFI first
- Cool and Quite can be left on while overclocked and unlocked the CPU, it causes no random BSoD -> STILL BSOD with CnQ enabled!
Well, two god things, one bad, so far i know :)
Didnt test the OC.
I didn`t tested the 1.47 today. I hope tomorrow..I will give my feedback also
Oliverda, tell me please why with the 1.44, and 1.47 bios, is the 2133Mhz memory frequency missing? How can I set the 2133Mhz without raising the fsb?
Dear Oliverda and Asrock please make a bios with 2133 and 2400 mhz memory frequency. Is very sad to have a maximum of 1866 mhz:mad:
How are the voltage increments/steppings on this board?
Ex. ASUS Sabertooth 990FX all starts at .005v increments.
Thanks for the reply.
I thought this was their high end board?Quote:
Originally Posted by Vladutz20
1866 is the MAX you can select?
I was considering buying this board, but geeeeeeeez gimme a break, that's pathetic at best.
Well, if you need higher mem clocks, then just raise the reference clock ;-)
1866 is the highest memory frequency that you can set in bios. I have tested this mainboard with 1.3 ,1.44 and 1.47 bios version.
Other brands has in bios by default 2133 and 2400 mhz. I don't want to raise anything !!! I want to have 2133 mhz by default :). I have Patriot Sector 5 2400 mhz 9-11-9-27 memory kit :)
In short, you bought OC RAM but you dont know how or are too lazy to overclock ... bad luck then.
Mister Opteron ! For example Asus Crosshair V Formula has by default 2133 and 2400 memory frequency. Asrock 990 FX Professional has a maximum of 1866... That is the problem.
I Know how to overclock my memory but i don't want to raise anything ...
Yes I know that there is a multiplier on other boards, I just wonder why you make that BIIIG problem out of it. Who cares, if there is a problem try to solve it yourself that's the fun with oc, to tweak and tune until you get a satisfying performance. If you know how to oc your memory then do it. Why wouldn't you want to raise anything? Anything above DDR3-1866 is not officially supported. It does not matter, if you oc by HTT or by Multiplier. I really don't get it. Don't complain and help yourself. Oc HTT to 250, use the multiplier of 1866 and you will get DDR3-2333.
Anyhow, that mode could be quite useless, because the uncore might limit the bandwidth as with the K10. As long as you dont oc the NB clock, it would be easier to run your memory @1600 with (very) low latency, e.g. CL5/CL6.
Good luck.
Good luck hitting 1600 with cas 5 :D
Hypers with 1.8v+ might do 5-6-5 XDQuote:
Originally Posted by Alex-Ro
5-8-5 is doable on PSC...
Anyone know what's with the 1.30 dated 10/20? I emailed them asking what is the difference from the older dated version and never got a response. The only thing I noticed that i didn't see before was the updated mouse support which is sorely needed, but other than that if it's a revision why not call it 1.30b or something?Quote:
Originally Posted by Valg
Been using the 1.44 uefi update and the nb multiplier doesn't seem to work and thanks for the 1.47 version it is now corrected. Still memory multiplier option is missing hope they'll include it in the next update.:)
Hello XS!
I just bought this board along with a FX-8150 and I'm having some weird problems.
I made a small overclock to 4200 MHz (21x200) at 1.3375V and whenever I shutdown my computer completely and turn it on again the frequency changes from 4200 to 4080 (21x194.6). In other words, whenever I shutdown and turn it on again the bus speed changes. If I restart after getting the 4080MHz frequency, the baseclock changes back to 4200MHz (21x200) again, in other words: the base clock returns to what I set it to in the BIOS.
The frequency change can be observed in CPU-Z, under computer properties and in F-Stream.
I should probably also tell you that I've tested many different BIOS versions: 1.20, 1.30 and 1.47 and the problem is apparent in all of them. I've also tried all the different LLC-settings aswell as C6, C1E etc and the problem remains.
What should I do? Is something broken? Am I missing something?
Please help me!
Edit:
Even at stock settings it changes to 194.3 MHz Bus Speed after a coldboot. One of the Cool'n'Quiet profiles for 1400 MHz (200x7) changes to 1360.2 MHz (194.3x7), also the RAM settings which I set to 1866 MHz in the BIOS change after a coldboot sometimes, to 647.7 MHz instead of 933 MHz. What am I supposed to do? Is there something wrong with my motherboard?
Warwian, try to disable all te CPU power management function(Cpu Feature), disable also spread spectrum. Or try hitting the F9 key in bios, save the bios, and test it agian.
I think I've localized the source of the problems now. Because they seem to have vanished when I left the RAM settings alone. In other words: when I left the RAM at 1333 MHz CL9. I'll see if I can get them running at 1866 tomorrow with a bit looser timings or something.Quote:
Originally Posted by Ramguy
Try running at 1866 then manually set the mem timings according to your ram specs, the 1.47v bios has a better and detailed ram configurations, try to save your settings as automatic and not on default mem profile. Set all your overclocking config to manual it may correct the issues you have encountered.Quote:
Originally Posted by Warwian
So I asked this a few weeks back and no one owns this board knows? I was under the impression that someone here owns it would like to give some details about the BIOS?
Quote:
Originally Posted by raisethe3
I will try that when I get home. I'd just like to summarize my problems in a few sentences first:Quote:
Originally Posted by jedm33
Basically any changes I make in the BIOS (for example changing RAM frequency to 1600 MHz or upping the CPU multiplier) ends up with me getting a black screen telling me that the systems has failed to boot several times before and that I may enter the BIOS to change settings. This always happens after shutting down and turning on the computer again. If I ignore this message and enter Windows, I get these strange values. For example, I set the CPU frequency to 4200 MHz and the memory to 1600 MHz with very loose CL9 (almost CL10 timings which my memory use for 2133 MHz) and turned it off and on again. I now got the black screen which I ignored and the CPU frequency had changed from 4200 (200x21.5) to 3705 (190x19.5), the voltage had changed from 1.3325V to 1.432V and the RAM frequency had changed back to 1333 MHz with auto CL9 settings (9-9-9-24) when I entered Windows.
If I were to simply restart after getting these strange values the correct ones I'd set in the BIOS would have been loaded (for example 4200 MHz, 1.3325V and 1600 MHz on the RAM) but they become strange again if I shutdown the computer and turn it on again.
I also noticed that if I restart and try to run Prime95 with the correct BIOS values (for example 4200 MHz, 1.3325V and 1600 MHz on the RAM) the CPU tends to clock itself down after a while to 3300 MHz. Even though I have turned off CnQ, C1E, C6 etc.
Does anyone have any more help? It's true that I have left some values on auto when trying these overclocks, I'll try to put in everything manual when I get back home.
What bios version are you using? 1.20, 1.30? or the modded 1.44, 1.47?...There are few users of fatality in this thread that carry the fx-8150 as of now that's why we can't get any feedback yet. If manual overclock and mem settings can't fix these issue then it might be a bios related problem after all. Or better yet try using the latest officially released bios update version 1.30 it has some cpu code updates according to the website. Btw, the 1.44v is a bit buggy so I won't recommend it using with your fx-8150. Goodluck ^_^Quote:
If I were to simply restart after getting these strange values the correct ones I'd set in the BIOS would have been loaded (for example 4200 MHz, 1.3325V and 1600 MHz on the RAM) but they become strange again if I shutdown the computer and turn it on again.
I also noticed that if I restart and try to run Prime95 with the correct BIOS values (for example 4200 MHz, 1.3325V and 1600 MHz on the RAM) the CPU tends to clock itself down after a while to 3300 MHz. Even though I have turned off CnQ, C1E, C6 etc.
These problems occured with the 1.30 BIOS but I also experienced them with the 1.47 one. I'm going to try the 1.47 BIOS again when I get back home, I'm also going to try different RAM.Quote:
Originally Posted by jedm33
The only settings which reliably work after a complete shutdown are the UEFI defaults, with RAM at 1333 MHz.
The big question to me here is why my settings change after a complete shutdown (when logging into Windows after the black screen I explained above, which tells me that the system has failed several times before and that I may change my settings in the BIOS). For example depending on which RAM settings I chose the baseclock seems to change. For example:
The 1866 MHz RAM setting seem to change my bus speed to 194 MHz and the 1600 MHz RAM setting seem to change the bus speed to 190 MHz (I'm not completely sure of this but I've seen it a couple of times). And also the CPU voltage is very different from what I set it to. For example as I said before I set my CPU to 4200 MHz with 1.3325V and RAM at 1600 CL9 and it booted in with the CPU at 3705 MHz with 1.432 V and the RAM at 1333 MHz.
I almost feels like it's changing my OC settings after a complete shutdown to some sort of fail-safe settings. I also had a theory that there might be some automatic overclocking function that is disturbing my settings. What else could it be that changes volts and frequencies? Worth mentioning again is that if I enter Windows with strange values after a shutdown (for example 3705 MHz) they correct themselves to what I set them to in the BIOS after a simple restart (back to 4200 MHz which I set in the BIOS). But if I turn off and turn on the computer again they change back to 3705 MHz and I have to restart to get my 4200 MHz back again.
Hmmmm, I think your problem is in the mem timings. I've experienced that too with incorrect or invalid ram timings, try setting it to 1866 as I've said above and experiment with different ram timing until it boots correctly without the black screen nags, mine worked at 8-8-8-24-31 ddr3 1600 cl8 1.5v ripjaws it's intel optimized I can't get it to work on cl6 timing though.Quote:
The big question to me here is why my settings change after a complete shutdown (when logging into Windows after the black screen I explained above, which tells me that the system has failed several times before and that I may change my settings in the BIOS).
I tried a different kit of RAM this time with no improvements what so ever. Instead of the Mushkin 8GB 2133MHz CL9-kit which I used before I tried an older Corsair XMS3 4GB 1600MHz CL9-kit.Quote:
Originally Posted by jedm33
I also took some pictures this time!
This is a picture with the correct settings which I set in the BIOS:
http://i1139.photobucket.com/albums/...reshutdown.png
Better picture that I sadly did not manage to link:
http://imageshack.us/f/685/beforeshutdown.png
And this is what happened with the very same BIOS configuration after I shut down the computer and turned it on again and got the blackscreen:
http://i1139.photobucket.com/albums/...lackscreen.png
Better picture:
http://imageshack.us/f/18/aftershutdownandblacksc.png
Can anyone explain this?
I don't understand what's going on, there seems to be something wrong with this motherboard and I'm regretting buying it in the first place. God I miss my old Gigabyte-board now!
EDIT: The only way to make the computer boot with my Mushkins at 1866 is with the timings at CL10. I tried some really loose and weird ones and they seem to work: 10-10-10-30-40, even after a shutdown without getting a black screen. Worth mentioning again is that my XMS3s didn't make their rated specs either after a shutdown as you can see on the pictures above. The 1600 MHz CL9 changed to 1333 MHz CL9 after a shutdown and getting the black screen.
Another EDIT: I just sent a message to ASRock support to see if they have any ideas. It might just be a question about BIOS maturity after all, considering FX-8150 is fairly new and all.
ASRock has always been behind as well as Biostar for overclocking ram on AMD platforms...
Seems they have some compatibility issues to work out, ASUS and Gigabyte boards are doing fine.
ASRock boards almost always look good on paper though...but until Ket got on things about BIOS support and become the official fanb...I'll quit there...
I dunno, their intel boards have been doing a lot better so I can see why they are gaining respect. This board looks pretty.
I agree, my gigabyte a75 ud2h is doing well with my ram configs before. I hope ASrock fixes the RAM issues ASAP, this board OC's so well and voltages are rock stable.:)Quote:
ASRock has always been behind as well as Biostar for overclocking ram on AMD platforms...
Seems they have some compatibility issues to work out, ASUS and Gigabyte boards are doing fine.
Right, so after doing some more testing with the 1.47 BIOS I decided it was time for an update!
Here are some pictures of some of the many different OC settings I tried today (I know the pictures are worse than horrible and almost impossible to read but it's the best I could do for now):
http://i39.tinypic.com/szxsgy.jpg
http://i40.tinypic.com/534njo.jpg
http://i41.tinypic.com/iwt4ro.jpg
Please note that this was just an example and I have no idea if those voltages are perfectly stable or not. It was just to show what happens with those settings after a complete shutdown.
Anyways, here are the very same settings after returning to Windows:
http://i41.tinypic.com/juzm9f.jpg
Looks great huh? Yeah.... but wait! This is what happens after shutting down the computer and turning it on again (after getting the black screen I've mentioned so many times before, remember that the black screen only appears if I turn off and on the computer and never during a normal restart):
http://i39.tinypic.com/2h6eq76.jpg
Wath? What trickery is this? Bus speed yet again set to 194 MHz for some reason? And what about the DRAM frequency? 772 MHz at CL6? Lol :p
I figured that there might be something wrong with my OC settings, so I decided to have a go with CPU stock settings but trying for 1866 MHz on the RAM.
Here comes some more horrible pictures of the settings I used:
http://i39.tinypic.com/2hf007s.jpg
As you can see in this picture the 1.47 BIOS was able to read my RAM's SPD settings and load them correctly:
http://i41.tinypic.com/4t9a2h.jpg
And as usual they worked perfectly before turning on and off the computer:
http://i44.tinypic.com/2j5b3mw.jpg
But once again they fail horribly after a complete shutdown:
http://i40.tinypic.com/280uusw.jpg
Bus speed changes to 194 MHz yet again and the RAM returns to the same fail-safe 1333 MHz CL9 auto setting I've mentioned before.
Lastly, I decided to try to take a picture of the black screen I've mentioned so many times. There is nothing special about it really, you have probably all seen it before. Many times! The strange thing with is of course that is shown after only changing the RAM settings to a mere 1866 MHz which should be cake for Bulldozer.
The camera had immense problems focusing on the screen so the picture is absolutely horrible, but still:
http://i41.tinypic.com/33v1axc.jpg
After doing these tests I saw that ASRock support had posted a new BIOS on Tweaktown (http://forums.tweaktown.com/asrock/4...tml#post408002)
I tried this aswell, but it seems to be identical with 1.47. The reason why I think it might be identical to 1.47 is that my computer sometimes fails to show even the BIOS screen at all after turning off the C6 states setting. This problem doesn't exist on 1.30 but has happened on both the 1.35 and the 1.47.
And of course, as it is identical to 1.47 the same OC/RAM problems were happening there too, I tried for 1866 MHz there aswell but with no results.
Warwian -
Insert an USB stick into your MBO,
and in the UEFI press F12 to take a screenshot.
Thanks for the tip!Quote:
Originally Posted by Valg
I actually decided to give the new 1.35 BIOS ago again and to also get a chance to take some new proper pictures. It seems the 1.35 isn't quite like the 1.47 BIOS after all! There are even more problems!
First of all! Here are the settings I used for the 4300MHz tests above (taken in the 1.35 BIOS instead of the 1.47 BIOS though, not that there is any difference; they still don't work. And yes! I have tried other settings and the results are similar: doesn't work properly (I did some more tries after these by just changing the voltages, multipliers, base clock around aswell as trying on many values on "auto" etc)!
The settings in the pictures were just an example!
http://i40.tinypic.com/16koc0.jpg
http://i41.tinypic.com/21omsth.jpg
http://i41.tinypic.com/efp16b.jpg
As you can notice on the last picture the DRAM Frequency is at DDR3-1333, but that's just because I booted into BIOS with that. As you might have guessed. But these are the settings that I used on the pictures above running the 1.47 BIOS. You can also see the results of these settings in those screenshots above.
Anyways, I tried to boot Windows with these OC settings (I never tried any real OC settings when I tried the 1.35 BIOS before, I only tried to get the RAM to 1866 MHz while keeping the CPU at stock)
And of course, they didn't work at all. I got the black screen after the computer was turned on off again by changing the CPU/DRAM frequency settings (the computer of course shuts itself down after saving and exiting the BIOS depending on what settings I've set) and I was welcomed by the black screen yet again. This time I didn't bother to boot up Windows but instead went back into the BIOS. And as you can see on my next picture the DRAM Frequency settings had changed from 1866 MHz (that I set it to) to 800 MHz:
http://i40.tinypic.com/126ecyt.jpg
Yet again, the system has problems with the settings I chose in the BIOS. This time the strange values were actually shown in the BIOS though. I've only seen them in Windows before and they've always looked correct in the BIOS. Until now of course!
When I tried to have a look what the BIOS had changed the timings to after reverting to DDR-800 MHz I was met by this error:
http://i43.tinypic.com/539rbk.jpg
Anyone know what that means? It just stopped being able to read my SPD profiles after that and it would only show that whatever changes I tried to make to the RAM.
Any thoughts?