kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 10:49:33
(permalink)
ZoranC
kram36
You posted the incorrect PCI-e slot you were using, you incorrectly posted that PE6 was not a x16 slot, sorry if I don't take what you post as gospel.
No sir, there are five x16 slots on the X299 Dark, you simply do not understand what a x16 slot is. An x16 slot is a full length slot, then you have some boards that has a shorter x8 slot and even the X299 Dark has a x4 slot.
Once again, PE6 is -NOT- x16 slot -electronically-, it is x16 mechanically. One can physically fit x16 card in I but it has only 8 lanes to the CPU. And your questions are revolving around how lanes to CPU are utilized, not physical sizing of slots.
Regardless of who is right on that your attitude seems to be that person should be immediately talked down to by you if he makes even a smallest mistake. If that is what you really believe then you are overdue for awakening. Nobody is perfect, even the best ones make mistakes, and you too make mistakes. I could start picking apart your posts but I don’t think that would lead us anywhere (especially because you don’t seem capable of accepting you might be incorrect on something).
Instead I will just point two things (for sake of those that are reading this and actually want to understand what I have been trying to convey):
#1 You claim X299 Dark has five x16 slots. That makes it obvious you are talking about physical -SIZE- of slots which is -NOT- same as five slots with 16 lanes to CPU each. If later was the case that would mean 80 lanes to CPU and one should know that is impossible, these CPUs don’t have that many lanes.
#2 When it comes to your question could PE4 end up using PCH it has been already pointed out to you that would be impossible because PE4 is 16 lanes (x4x4x4x4 when bifurcated) while PCH is only x4.
With that said I will end this conversation and let you have the last word, especially because this and other discussions I’ve seen you engage in leave me with firm impression you seem much more interested in arguing in circles proving you are smartest and know everything best than keeping open mind to what others are saying.
Now let EVGATech_LeeM answer the question as I quoted him for the answer. Also EVGATech_LeeM should specify this can only work on PE1 and PE4 instead of saying "This can be used to set a PCIe x16 slot to x4x4x4x4" if that is how the bios really works. Last word then, there are five x16 slots on the X299 Dark. /end
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 11:16:31
(permalink)
kram36Also EVGATech_LeeM should specify this can only work on PE1 and PE4 instead of saying "This can be used to set a PCIe x16 slot to x4x4x4x4" if that is how the bios really works.
To other readers of this thread that are hoping to get some facts: I have actual hands on experience bifurcating with BIOS 1.23 and Asus card rather than speculating how things might be working. I've tested that card in PE3 and PE4. When in PE3 you get two drives which makes sense because that slot has 8 lanes (x4x4). When in PE4 (and PE3 is empty) you get four drives (x4x4x4x4) because that slot has 16 lanes. I haven't tested that card in PE1 nor PE2 because I didn't see point in spending time on that (PE1 is where my video card is and if I put it in PE2 I would be putting PE1 in x8 mode plus I would be blocking fan on my video card). However there is nothing that would make me think bifurcation would work differently for PE1/PE2 than it does for PE4/PE3.
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 12:08:06
(permalink)
ZoranC
To other readers of this thread that are hoping to get some facts:
I have actual hands on experience bifurcating with BIOS 1.23 and Asus card rather than speculating how things might be working.
I've tested that card in PE3 and PE4. When in PE3 you get two drives which makes sense because that slot has 8 lanes (x4x4). When in PE4 (and PE3 is empty) you get four drives (x4x4x4x4) because that slot has 16 lanes.
I haven't tested that card in PE1 nor PE2 because I didn't see point in spending time on that (PE1 is where my video card is and if I put it in PE2 I would be putting PE1 in x8 mode plus I would be blocking fan on my video card). However there is nothing that would make me think bifurcation would work differently for PE1/PE2 than it does for PE4/PE3.
One thing I forgot to mention: When I tested this I ran Crystal Disk Mark against two drives in stripe. I was hitting over 7GB/sec in sequential test. From all I know that would've been impossible if slot was using PCH.
|
kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 13:10:30
(permalink)
ZoranC
ZoranC
To other readers of this thread that are hoping to get some facts:
I have actual hands on experience bifurcating with BIOS 1.23 and Asus card rather than speculating how things might be working.
I've tested that card in PE3 and PE4. When in PE3 you get two drives which makes sense because that slot has 8 lanes (x4x4). When in PE4 (and PE3 is empty) you get four drives (x4x4x4x4) because that slot has 16 lanes.
I haven't tested that card in PE1 nor PE2 because I didn't see point in spending time on that (PE1 is where my video card is and if I put it in PE2 I would be putting PE1 in x8 mode plus I would be blocking fan on my video card). However there is nothing that would make me think bifurcation would work differently for PE1/PE2 than it does for PE4/PE3.
One thing I forgot to mention: When I tested this I ran Crystal Disk Mark against two drives in stripe. I was hitting over 7GB/sec in sequential test. From all I know that would've been impossible if slot was using PCH.
I agree with you 100% on this. That's what I found from my testing, CPU lanes are faster and feel snappier then PCH lanes. Just two Samsung 970 EVO Plus 500GB drives will get over 7GB/s in Raid 0 from the CPU lanes.  WD Black SN750 1TB from PCH lanes.  WD Black SN750 1TB from CPU lanes.  This is the whole reason I'm asking about PE4 staying in CPU lanes when in x4x4x4x4 mode.
post edited by kram36 - 2020/09/07 13:18:59
|
DEJ915
SSC Member
- Total Posts : 544
- Reward points : 0
- Joined: 2013/11/03 21:58:26
- Status: offline
- Ribbons : 11
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 15:20:58
(permalink)
ZoranC
Once again, PE6 is -NOT- x16 slot -electronically-, it is x16 mechanically. One can physically fit x16 card in I but it has only 8 lanes to the CPU. And your questions are revolving around how lanes to CPU are utilized, not physical sizing of slots.
Right, they asked about where the PCIE lanes come then started going off on about the size of the slot which has no bearing on the lanes aside from you can't have more lanes than contacts for them. My Apex IX has 4 full length slots but they are x8 x4 x4 x4 am I magically going to somehow get x16 in all 4 slots? No because the size has nothing to do with the lanes, brands use full size slots for more support for cards and nothing more. Also the PE6 having a graphics card in it has no bearing on the lane allocation it gets. what causes it to switch is what options you enable for your M.2 and U.2 storage.
post edited by DEJ915 - 2020/09/07 15:34:41
|
kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 16:00:03
(permalink)
DEJ915
Right, they asked about where the PCIE lanes come then started going off on about the size of the slot which has no bearing on the lanes aside from you can't have more lanes than contacts for them. My Apex IX has 4 full length slots but they are x8 x4 x4 x4 am I magically going to somehow get x16 in all 4 slots? No because the size has nothing to do with the lanes, brands use full size slots for more support for cards and nothing more.
You are confusing a PCIe x16 slot, which is what EVGATech_LeeM posted " (This can be used to set a PCIe x16 slot to x4x4x4x4 without VROC feature)" with PCIe lanes available to the slot. The X299 Dark has five PCIe x16 slots no matter how you want to talk about it. I never said you could make PCIe lanes magically appear, I simply asked if the lanes stayed on the CPU or went to the PCH. DEJ915
Also the PE6 having a graphics card in it has no bearing on the lane allocation it gets. what causes it to switch is what options you enable for your M.2 and U.2 storage.
Then EVGA needs to update their manual to your specifications. As per EVGA, when PE6 is populated with a video card (anything using x8 lanes) the M.2 PM2 slot gets knocked from CPU lanes to PCH lanes. 
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/07 18:06:06
(permalink)
DEJ915
My Apex IX has 4 full length slots but they are x8 x4 x4 x4 am I magically going to somehow get x16 in all 4 slots?
One can dream though ... 80+ lanes with X299 Dark here I come LOL ;) DEJ915
Also the PE6 having a graphics card in it has no bearing on the lane allocation it gets. what causes it to switch is what options you enable for your M.2 and U.2 storage.
Yup. I must admit though sentence about PE6 could've been phrased much better, it took me some reading and digging first time I saw it before I figured out what they meant.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 10:37:30
(permalink)
Im using an asus pci 16x on the dark in pe1 and it recognize the 3nvmes i have on it. In bios it shows and everything not in the top corner off the bios where it shows the speed and gen tho it only recognize 2 of the 3nvmes it says 4x/4x "it doesnt show the 3rd drive" it does show all nvmes of the pci 16x on the nvme section tho.. I havent install the 4th one as im using it still on the micro2 ... But in the gui in the right corner it shows 4x4x "nada" and the gpu i have in that slot as testing... It also boots 4200mhz it drops a stick but i know thats the 7940x cpu with the 7980xe i should have no problems.. The 7940x behaves the same in the micro2 board that i remember.. It loves loosing a stick at that speed lol.. 4000 0 problems..
I got also a 64gb xtreem 3733 kit thats running at 3800 on 16gb sticks dr ...probably can do 3900+ but thats that.. On the micro2..
Got me the pci card for hdd cache purposes and shes a nasty ittle fella and hell fast in conjunction with primocache and ram.... oh my, almost 60gbs on writes lol almost 30gns on reads i need to try the whole thing first but its looking promising.. But im swaping mobos and changing my rig to an open build bcuz got tired on postponing hardware swaps bcuz of the size of my build.. I need to watercool somehow the dark vrms as they running kind of hotter than my micro2 and im not liking it...
Idk but im more inclined towards the micro2 board than the dark. Dark has more features but the micro2 as giving me a hell of a ride so far loving that little board.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 10:41:51
(permalink)
Btw here "da nasty" lol ps5 got crap on her lol
Attached Image(s)
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 10:51:32
(permalink)
zGunBLADEz
In bios it shows and everything not in the top corner off the bios where it shows the speed and gen tho it only recognize 2 of the 3nvmes it says 4x/4x "it doesnt show the 3rd drive" it does show all nvmes of the pci 16x on the nvme section tho.. I havent install the 4th one as im using it still on the micro2 ... But in the gui in the right corner it shows 4x4x "nada" and the gpu i have in that slot as testing...
It would help if you could post a screenshot of that and let us know what you have in PE2.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 10:58:15
(permalink)
Pe2 is empty only pe1 is used with the pci nvme card 16x and pe4 with a gpu for display purposes..
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 11:03:50
(permalink)
zGunBLADEz
Pe2 is empty only pe1 is used with the pci nvme card 16x and pe4 with a gpu for display purposes..
Interesting. It makes me wonder could it be due to fact you are trying to use 3 drives, not 2 or 4. P.S. Do you mind sharing the screenshot, please?
|
DEJ915
SSC Member
- Total Posts : 544
- Reward points : 0
- Joined: 2013/11/03 21:58:26
- Status: offline
- Ribbons : 11
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 11:28:41
(permalink)
I think there was a discussion in a thread about that display before that since the bifurcation/non-vroc was added later the pcie lane display at the top does not show it properly.
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 11:35:55
(permalink)
DEJ915
I think there was a discussion in a thread about that display before that since the bifurcation/non-vroc was added later the pcie lane display at the top does not show it properly.
Your post made me realize I might have been mislead with how he phrased his post. I interpreted it as "he can't see 3rd drive at all when card with 3 drives is in PE1". That requires clarification ... @zGunBLADEz Is -Windows- able to see and use all 3 drives?
|
arestavo
CLASSIFIED ULTRA Member
- Total Posts : 6916
- Reward points : 0
- Joined: 2008/02/06 06:58:57
- Location: Through the Scary Door
- Status: offline
- Ribbons : 76
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 13:19:47
(permalink)
There's supposedly a beta BIOS that allows all of the NVME drives to be seen in the BIOS (including the ones in a bifurcation card). Apparently there is a limit to how many are shown in the current, non-beta BIOS'.
This is just me going off of memory of what I've read in this thread.
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 13:28:11
(permalink)
arestavo
There's supposedly a beta BIOS that allows all of the NVME drives to be seen in the BIOS (including the ones in a bifurcation card). Apparently there is a limit to how many are shown in the current, non-beta BIOS'.
Yes, V 1.23 of BIOS doesn't show NVMe info for all drives when one has more than 4 drives but that is -not- what he is talking about, is it?
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 19:52:05
(permalink)
DEJ915
I think there was a discussion in a thread about that display before that since the bifurcation/non-vroc was added later the pcie lane display at the top does not show it properly.
This... They all 3x show on the bios nvme section and on windows.. Its just in the top right corner where issue is at but a quick trip to nvme bios section will show you the drives. @ZoranC yes windows see all 3 nvmes have them in raid 0 with windows storage spaces Also, i got a AX200 intel wifi card for the board and 0 issues it works just fine..
post edited by zGunBLADEz - 2020/09/09 20:05:51
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 20:25:40
(permalink)
Never mind this one that pe6 shared slot is a nightmare lol
post edited by zGunBLADEz - 2020/09/09 21:11:01
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/09 21:59:49
(permalink)
zGunBLADEz
DEJ915
I think there was a discussion in a thread about that display before that since the bifurcation/non-vroc was added later the pcie lane display at the top does not show it properly.
This...
They all 3x show on the bios nvme section and on windows.. Its just in the top right corner where issue is at but a quick trip to nvme bios section will show you the drives.
@ZoranC yes windows see all 3 nvmes have them in raid 0 with windows storage spaces
Also, i got a AX200 intel wifi card for the board and 0 issues it works just fine..
Then that clarifies it. Upper right corner is supposed to show number of lanes and PCIe version for each slot but when slot is bifurcated it will only show x4 regardless of how many are used in that slot (there was a thread between me and EVGA about that and one more interface oddity when this was rolled out). Even NVME info might not show everything correctly (if one has more than four NVMe drives, counting U.2, in the system) so as long as Windows sees all of the drives one is good. Glad to hear it is working fine for you.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/16 05:44:57
(permalink)
Someone with a evga dark that can confirm something about the 12v readings and the 80port code screens and 12v readings using that and hwinfo as well got the same reads... im getting sporadic readings/drops to 7v & 4v this will show on the 80port and as well hwinfo they sporadic readings drops.. Even under stress load or idle it will do this behavior .. Funny tho if you get a drop like that even on idle your pc would crash no question asked. I tried different psus and still show same behavior another user on overclock.net capture this drop as well on hwinfo... I try another BIOS same issue is present. Also, why the 80port led "it's not showing actual " hottest core" "package temp" on the led when using sensor 1/2 together? Its showing Tcase instead to make it show actual package temp i have to select cpu die only and thats all it shows but if i select Sensor 1/2 lets say cpu die and vrms the temp is the TCase and not the actual package temp.... Edit: actual package temp will display CORRECTLY ON BIOS ONLY OPTION in the 80port led as one reading.. THE 80 PORT settings on bios are all wrong needs to be fixed.. Why i cant show independently in both screens at the same time one side cpu and the other side vrm temps? https://www.overclock.net...ssion.1632870/page-538Here
post edited by zGunBLADEz - 2020/09/16 06:03:12
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/16 13:23:15
(permalink)
zGunBLADEz
Someone with a evga dark that can confirm something about the 12v readings and the 80port code screens and 12v readings using that and hwinfo as well got the same reads... im getting sporadic readings/drops to 7v & 4v this will show on the 80port and as well hwinfo they sporadic readings drops.. Even under stress load or idle it will do this behavior .. Funny tho if you get a drop like that even on idle your pc would crash no question asked. I tried different psus and still show same behavior another user on overclock.net capture this drop as well on hwinfo...
I have seen that behavior monitoring my X299 Dark and two power supplies (Corsair AX1200 and AX1600i) with HWInfo64. My instinctive reaction was to be concerned but eventually figured that: a) using any piece of software (whether it be HWInfo or BIOS) to readout voltage depends on accuracy of software itself, plus software depends on sensor itself, so there is a chance software and/or sensor are malfunctioning, and ... b) it doesn't matter what software thinks voltage is, what matters is actual voltage and only reliable way to know actual voltage is by measuring with voltmeter. Measuring with voltmeter would be going way too deep down the rabbit's hole for me so at that point I stopped myself, rethought things and (just like you) realized if what HWInfo was telling me was actually true my system would be crashing left and right all day long instead of passing many many hours of P95 without a single problem. So I decided I should ignore those readouts.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/16 21:25:51
(permalink)
I agreed, But problem is and thats why im iffy about it. i noticed first on the 80port mobo code screen as we speak the mobo is on a open bench the mobo is almost in my face when using it... not in Hwinfo, hwinfo just confirms it for me that what i saw on the board the app can record it. I had the 2 screens displaying just the 12v sensor and it did read out 4v & 7v drops mind you is quick and sporadic and you have to pay close attention to it... Then i decided to monitor it with hwinfo on a graph and it did it again "i just stared at the mobo code screen till it did it" and hwinfo confirms on the software side.. So its the board doing it not hwinfo..
post edited by zGunBLADEz - 2020/09/16 21:31:30
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/16 21:41:47
(permalink)
zGunBLADEz
I agreed, But problem is and thats why im iffy about it.
i noticed first on the 80port mobo code screen as we speak the mobo is on a open bench the mobo is almost in my face when using it... not in Hwinfo, hwinfo just confirms it for me that what i saw on the board the app can record it. I had the 2 screens displaying just the 12v sensor and it did read out 4v & 7v drops mind you is quick and sporadic and you have to pay close attention to it... Then i decided to monitor it with hwinfo on a graph and it did it again "i just stared at the mobo code screen till it did it" and hwinfo confirms on the software side.. So its the board doing it not hwinfo..
That is why I said it could be -either- software -or- whatever both pieces of software depend on to find out what current value is (sensor?). I don't like seeing it (whenever something doesn't work as it should my level of confidence in things experiences a drop) but in the end I had to be practical about it. Still it would leave a very good impression if EVGA would look into it and identify reason.
|
zGunBLADEz
New Member
- Total Posts : 58
- Reward points : 0
- Joined: 2014/11/26 07:05:08
- Status: offline
- Ribbons : 1
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/17 00:25:35
(permalink)
Yep i want and answer for this and the 80port code cpu die reading as well. Cant unserstand how this was overlooked after this yrs and nobody said something about it same as the per core overclocking that i have to literally open a topic and ask for it as one of the best features of x299 is that in particular and it was practical absent.. At least evga added per request so goodies to them.. This board have all this readings and 12v is one of the most essential one and of course i have to nit picking on why i cant show simultaneously 2 temp sensor readings on 2 80port screens instead of cycling thru both like mind blown.. Also i have no cpu power draw whatsoever even with svid enable which my Micro2 board does with svid disable and it does very well... it display on hwinfo the actual cpu power draw "heat" which is very helpful on my cooling approach while benching.
post edited by zGunBLADEz - 2020/09/17 00:29:17
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/17 10:41:43
(permalink)
zGunBLADEz
Cant unserstand how this was overlooked after this yrs and nobody said something about it ...
Neither can I. And that's not the only thing I can't understand.
|
kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/20 10:40:38
(permalink)
ZoranC
kram36Also EVGATech_LeeM should specify this can only work on PE1 and PE4 instead of saying "This can be used to set a PCIe x16 slot to x4x4x4x4" if that is how the bios really works.
To other readers of this thread that are hoping to get some facts:
I have actual hands on experience bifurcating with BIOS 1.23 and Asus card rather than speculating how things might be working.
I've tested that card in PE3 and PE4. When in PE3 you get two drives which makes sense because that slot has 8 lanes (x4x4). When in PE4 (and PE3 is empty) you get four drives (x4x4x4x4) because that slot has 16 lanes.
I haven't tested that card in PE1 nor PE2 because I didn't see point in spending time on that (PE1 is where my video card is and if I put it in PE2 I would be putting PE1 in x8 mode plus I would be blocking fan on my video card). However there is nothing that would make me think bifurcation would work differently for PE1/PE2 than it does for PE4/PE3.
I got two ASUS Hyper M.2 V2 cards in. I want a bootable RAID 0 for Windows, so I have to use VROC and Intel NVMe drives. First I tried with three Intel P760 512GB NVMe drives on the Hyper M.2 card in PE4 and put my two Samsung EVO Plus NVMe drives on the other Hyper M.2 card in PE3. No matter if I had PE3 in VROC or non VROC, the motherboard would only see two of my Intel NVMe drives in PE4, like you said, it cuts down PE4 to two x4 lanes when PE3 is used. So then I figured I would just put my Samsung NVMe drives on the motherboard. However when doing that, those slots must be set to VROC since I had PE4 in VROC mode. If I did not do that, the system would hang with post code 94 and required me to clear the bios to boot again. This is not an optimal use of my Samsung NVMe drives as VROC does not run as fast. So I ended up putting the two Samsung NVMe drives back into the Hyper M.2 card and placing the card in PE4 with the non VROC mode, put two of my Intel NVMe drives on the motherboard in VROC, put the other Intel NVMe drive in my other Hyper M.2 card along with another smaller 120GB Corsair NVMe drive (using it as a Temp folder drive), placed the card in PE3 and set it to VROC. Now the motherboard booted and saw all three of my Intel NVMe drives and I could put them in Raid 0. Was worried splitting the Intel NVMe drives up like this might cause an issue, but I have not seen any so far and the system is running fantastic!  So yes, bottom line is to use 4 drives in the ASUS Hyper M.2 card, the card must be in PE4 and nothing in PE3. I assume like you that PE1 would work also, but that's where my video card is. However, EVGATech_LeeM should be more clear in his post about how this works. BTW: I did start out with just two of the Intel NVMe drives and the performance just wasn't where I wanted it to be. So I ordered in another Hyper M.2 card and a third Intel NVMe drive. This was with just two Intel NVMe drives. Not sure why the Read SEQ1M Q1T1 drops like it does with three Intel NVMe drives? 
post edited by kram36 - 2020/09/20 10:49:58
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/20 11:55:07
(permalink)
kram36
First I tried with three Intel P760 512GB NVMe drives on the Hyper M.2 card in PE4 and put my two Samsung EVO Plus NVMe drives on the other Hyper M.2 card in PE3. No matter if I had PE3 in VROC or non VROC, the motherboard would only see two of my Intel NVMe drives in PE4, like you said, it cuts down PE4 to two x4 lanes when PE3 is used.
… which makes sense because, per manual, PE3/PE4 share between them 16 lanes to CPU in either x0/x16 or x8/x8, one can’t put 5 drives in them for total of 20 lanes and expect to see them all. kram36
So then I figured I would just put my Samsung NVMe drives on the motherboard. However when doing that, those slots must be set to VROC since I had PE4 in VROC mode.
For those interested my personal layout is like this: a) U.2 drive (or, to be precise, IcyDock adapter with Samsung 970+ NVMe that converts it into U.2) on PU2 (using CPU lanes) as boot drive. b) M.1 and M.2 populated with Samsung 970+ NVMe. Both configured to use CPU lanes but no special storage configuration (like VROC/non-VROC setting) needed. That doesn’t mean they can’t become part of a “RAID” because I’m using Storage Spaces as my storage solution and advantage of Storage Spaces over VROC/RST(e) is that you can mix and match (it can use any drive Windows sees, whether it is on U.2 port or in M.2 slot or on bifurcated PCIe card or hanging off mb’s SATA port, as part of storage pool). c) Two Samsung 970+ NVMe on Asus PCIe card in PE4 configured as non-VROC and “RAIDed” through Storage Spaces. I’ve populated just two because I have plans for PE3 and don’t need more than two. Those that don’t have plans for PE3 and need more than two drives can use four, it will work fine.
|
kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/20 12:09:16
(permalink)
kram36
So then I figured I would just put my Samsung NVMe drives on the motherboard. However when doing that, those slots must be set to VROC since I had PE4 in VROC mode.
ZoranC
For those interested my personal layout is like this:
a) U.2 drive (or, to be precise, IcyDock adapter with Samsung 970+ NVMe that converts it into U.2) on PU2 (using CPU lanes) as boot drive.
b) M.1 and M.2 populated with Samsung 970+ NVMe. Both configured to use CPU lanes but no special storage configuration (like VROC/non-VROC setting) needed. That doesn’t mean they can’t become part of a “RAID” because I’m using Storage Spaces as my storage solution and advantage of Storage Spaces over VROC/RST(e) is that you can mix and match (it can use any drive Windows sees, whether it is on U.2 port or in M.2 slot or on bifurcated PCIe card or hanging off mb’s SATA port, as part of storage pool).
c) Two Samsung 970+ NVMe on Asus PCIe card in PE4 configured as non-VROC and “RAIDed” through Storage Spaces. I’ve populated just two because I have plans for PE3 and don’t need more than two. Those that don’t have plans for PE3 and need more than two drives can use four, it will work fine.
Want to run a test an see if the special bios they sent you fixes the VROC issue I had with the Samsung NVMe drives on the motherboard? Save your current bios settings because if it hangs, you have to clear the cmos to boot up again and then set only PE4 to VROC. See if your system boots (just to the bios, don't let it go further, unless you have VROC drivers installed) or if it hangs with post code 94. I was really surprised that I had to have the Samsung NVMe drives in VROC mode on the motherboard for the system to boot.
|
ZoranC
FTW Member
- Total Posts : 1099
- Reward points : 0
- Joined: 2011/05/24 17:22:15
- Status: offline
- Ribbons : 16
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/20 12:34:50
(permalink)
kram36
Want to run a test an see if the special bios they sent you fixes the VROC issue I had with the Samsung NVMe drives on the motherboard? Save your current bios settings because if it hangs, you have to clear the cmos to boot up again and then set only PE4 to VROC. See if your system boots (just to the bios, don't let it go further, unless you have VROC drivers installed) or if it hangs with post code 94.
I’m not in position to experiment any more with that, my build has been completed and system moved “into production use”. However, version of BIOS I am using at the moment has nothing to do with it. All of my “will bifurcation and StorageSpaces work” tests have been conducted with V 1.23. Only “issue” I had was that NVMe info did not -show- info for all drives when one had more than 4 NVMe drives total -BUT- system -DID- correctly see and use all of them. I got beta BIOS -after- that to address only that. I think you should be able to easily reproduce my setup and confirm that. Take your VROC key (if you have one, I don’t) and all drives out, put 970+ in M1 and M2, put two NVMe drives in first two positions on Asus card and put card into PE4, clear your CMOS, make sure M1 and M2 are using CPU lanes, storage configuration for PE4 is non-VROC and for all other slots disabled and that should be it.
|
kram36
The Destroyer
- Total Posts : 21477
- Reward points : 0
- Joined: 2009/10/27 19:00:58
- Location: United States
- Status: offline
- Ribbons : 72
Re: BIOS Updates for X299 Dark (1.23) / FTW K (1.23) / Micro ATX 2 (1.16) / Micro ATX (1.2
2020/09/20 12:54:12
(permalink)
ZoranC
kram36
Want to run a test an see if the special bios they sent you fixes the VROC issue I had with the Samsung NVMe drives on the motherboard? Save your current bios settings because if it hangs, you have to clear the cmos to boot up again and then set only PE4 to VROC. See if your system boots (just to the bios, don't let it go further, unless you have VROC drivers installed) or if it hangs with post code 94.
I’m not in position to experiment any more with that, my build has been completed and system moved “into production use”.
However, version of BIOS I am using at the moment has nothing to do with it. All of my “will bifurcation and StorageSpaces work” tests have been conducted with V 1.23. Only “issue” I had was that NVMe info did not -show- info for all drives when one had more than 4 NVMe drives total -BUT- system -DID- correctly see and use all of them. I got beta BIOS -after- that to address only that.
I think you should be able to easily reproduce my setup and confirm that. Take your VROC key (if you have one, I don’t) and all drives out, put 970+ in M1 and M2, put two NVMe drives in first two positions on Asus card and put card into PE4, clear your CMOS, make sure M1 and M2 are using CPU lanes, storage configuration for PE4 is non-VROC and for all other slots disabled and that should be it.
I had the same issue in VROC mode of all the drives not showing in the bios, but were usable in Windows. Why would I want to reproduce your setup if I already know on my system it doesn't work as I must use VROC? Also the VROC key is pretty much worthless on X299 motherboards. Even if you have a VROC key, you can not Raid any non Intel NVMe drives with X299 through the bios. Yes you can use the software to do it in Windows with non Intel NVMe drives, but it's only for 90 days. I used Microsoft Storage Spaces to put my Samsung NVMe drives in Raid 0 and it gave me zero benefit. I do some system benchmarking and Microsoft Storage Spaces gets picked up as a Ram Drive and the benchmark tosses out the score for those drives. So I just use the regular Windows Disk Manager to put my Samsung NVMe drives in Raid 0 and the benchmark has no issue with the drives.
|