SuperMicro X9SAE/-V - Xeon IvyBridge C216 ATX

Thank you one time again for your reply Fritingo.

My English is poor so this maybe the reason I can't understand this:

"But...page 4-13 in the user manual for the X9SAE has the following: "Boot Graphics Adapter Priority. Use the feature to select the graphics controller to be used as the primary boot device. The options are Slot 6 VGA, Offboard, and Onboard." So it does look like you can select. Would be very odd if you couldn't. Traditionally on boards with integrated video chips (in the old days) or boards that support CPUs with GPUs (currently), there have always been BIOS settings allowing you to select which one is active."

...because It seems that You say it to show that perhaps this board is an exception to this your following statement:

"That's odd. Traditionally, you can't use both a CPU GPU and an offboard card at the same time. Not familiar with any modern board that allows that. If you really need that capability, get two offboard cards."

...but instead it seems to me only a confirmation.

...unless "primary boot device" means a secondary device is contemplated...
 
Last edited:
ps: modern boards allow coexistence of iGPU and discrete GPU but these boards are addressed to non-server users. See LUCID LOGIX VIRTU technology or NVIDIA Optimus/Synergy. But I don't know if server boards as Supermicro X9SAE are able to support this technology (LLVIRTU certainly did not because the manufacturer requires a license which is usually widely publicized while NVIDIA SYNERGY should work on every system ..admitted, I suppose, that the chipset and BIOS support it..).

======================================================

I do not understand why there are three options:

- onboard
- offboard
- VGA slot 6

I suppose "onboard" means iGPU integrated in CPU and until now everything ok but what is the difference between "off-board" and "slot 6"? Maybe because in theory even more of a discrete graphics card can be installed? ...and so "offboard" means any other card in any other slot except slot #6 (16x slot)? Or what else?
 
Last edited:
Sorry, didn't know about the language issue.

What I meant is that the X9SAE and X9SAE-V apparently support having integrated graphics in the CPU and a separate graphics card both installed at the same time. With the ability to switch between them in the BIOS. Not necessarily running them at the same time. But it may in fact allow running both at the same time or switching between them in the operating system (NVIDIA Optimus); I don't know for sure.

If interested in NVIDIA Optimus/Synergy support on the board or confirmation of the above, best bet is to contact Supermicro ( by email, etc.).

I suppose "onboard" means iGPU integrated in CPU and until now everything ok but what is the difference between "off-board" and "slot 6"? Maybe because in theory even more of a discrete graphics card can be installed? ...and so "offboard" means any other card in any other slot except slot #6 (16x slot)? Or what else?

Yes, "onboard" means integrated graphics.

The X9SAE (with one 16x PCIe 3.0 slot) and the X9SAE-V (with two 8x electrical/16x physical PCIe 3.0 slots) both have the same user manual. And perhaps the exact same BIOS. It seems that the difference between "offboard" and "VGA slot 6" in the BIOS settings means selecting from one of two installed offboard graphics adapters if you have the X9SAE-V only (I haven't counted the slot numbers). If you have the X9SAE, only the "offboard" setting would work.

Again, for confirmation, best bet is to contact Supermicro.
 
Yes, X9SAE and X9SAE-V have the same BIOS and your hypothesis could be right but slot6 is the only one present on both motherboards therefore if you have the X9SAE, only the "slot 6" setting would work.

..but, at this point, I do not understand why instead of "offboard", they didn't write "slot 4" for the other 8x slot (8x electrical and 16x physical)...

You are right, the best way to have confirmation (also and especially to ask for clarification on simultaneous support for GPU and iGPU) is to contact supermicro.
 
You forced me to look at the manual again (shakes fist) ;).

Well, both the 16x slot on the X9SAE and one of the two 8x slots on the X9SAE-V are "slot 6". That being the case, would think that the "Offboard" BIOS option means different things for each board.

"Offboard" would likely mean the 8x slot 4 on the X9SAE-V only and the 16x slot 6 on the X9SAE. And the "Slot 6 VGA" BIOS option means...the 8x slot 6 on the X9SAE-V and exactly the same thing as "Offboard" means on the X9SAE (the 16x slot 6).

As the BIOS is the same between the two boards, there is no difference between "Offboard" and "Slot 6 VGA" BIOS options on the X9SAE only.

That's the only layout that makes sense here. But again for confirmation, contact Supermicro.
 
Hello Fritingo, thank you but there was no need to check cause I asked to Supermicro (no-reply yet ...maybe Christmas Holidays?).

These days I documented a bit 'on SSDs.
The Samsung 840-PRO is exciting but I saw that the samsung 840pro MDX controller does not use a RAID-like redundancy and this, in terms of reliability is not really the best.
This fact can be seen also as the capacity is reported: it is in fact expressed in binary rather than decimal (although the 840 [non-PRO] capacity, unlike the 840 PRO which uses the same MDX controller, instead is expressed in decimal, but only because here the overprovisioning is enabled by default while in 840-PRO it is not ...according to what I read on a review).
What do you think about?

======================================

Looking, I found this product: Comay Venus PRO 3.
It's an SSD produced in China that uses SandForce SF-2281/2282 controller and Intel 25nm synchronous NAND.
It is guaranteed for three years and 2 million MTBF.
The main feature is that it is equipped with a supercapacitor (CapXX produced in USA/Australia) that can protect against power failure and this is a feature present only on enterprise class devices (in mainstram class, you can find it only in intel 320 series) although a capacitor does not cost more than $10 (and for companies that make large purchases perhaps much less).
It is rated to reliability more than performance but also these are not bad at all. It therefore seems quite stable.
Support AES 256 encryption and RAISE (SandForce raid-like redundancy).
Is priced at $ 180 and looks like a very interesting product.
Here is a review:
http://www.tweaktown.com/reviews/4996/comay_venus_pro_3_240gb_ssd_review/index2.html
 
Last edited:
Never heard of Comay before. They don't sell in the North American market yet, thus I wouldn't bet on them ;).

If you are concerned about reliability of your entire system, you will have a UPS (uninterruptible power supply) to which your system is plugged into. Eliminating the concern about things going haywire when the power fails.

"samsung 840pro MDX controller does not use a RAID-like redundancy" Huh? You talking about that SandForce RAISE junk? Yes, Samsung uses its own components, not SandForce. As per Tom's Hardware: "SandForce designed RAISE to allow SSD vendors to use lower-quality NAND chips with its more premium controller without sacrificing longevity. Whatever. If you think that's worth more than what Samsung's (or Intel's) own controllers give you, the more power to you. I don't consider it any advantage; it looks like nothing more than standard over-provisioning with marketing speak added in for good measure. It also looks like SandForce's controller using a drive's storage chips (made by other vendors) to cover their own butts if their historic unreliability pops up on a drive...heh.

If you are having second thoughts about the Samsung 840 Pro, the only other modern drive I'd recommend in terms of reliability is the previously mentioned enterprise-class Intel DC S3700. But you'll pay a very hefty premium for it (70%+ more than the Samsung).

There may be other good ones, but in terms of SSD reliability, its still pretty much a crapshoot in terms of finding out what's reliable. They don't have the history regular hard drives do. So according to my own reliability guidelines, I look at who makes the SSD controller on a drive, what features that controller has and what the historic reliability of that controller brand has been. Also what "class" the SSD is (consumer or enterprise). Little else matters. In the old days, multi-level cell (MLC) flash memory wasn't "as reliable" as single-level cell (SLC) flash memory, but controller advancements and features have eliminated the difference. That Intel drive in fact uses MLC memory.
 
Last edited:
As you can see from the already linked review, now Comay (CoreRise) has a reseller in North America too.

I have no experience with SSDs and I learn the things I know, on the forums or through the reviews.
Therefore, I do not know how much the SandForce controller can be good, however I think it's not just shit. After all, it was adopted even by Intel.

Regarding the RAISE system, if I'm not mistaken, the Intel 710 and 320 series (for example) with proprietary controller, implement a similar redundance system. And I think Intel doesn't use memories of poor quality.

In any case, I too have always wondered if RAISE and overprovisioning technologies are substantially the same thing and if it is redundant to have both at the same SSD (but I didn't understand if there is or not any difference).

From what I understood, to lower the cost of manufacturing processes, manufacturers use a smaller and smaller litography.
This leads, perhaps, to lower energy consumption and lower prices but also to a shortening of rewriting cycles (P/E) and decreasing in performances.
Samsung PRO 840 uses a 21nm process on its toggle-NAND (the smaller ever seen) and this means high reliability and endurance is given mostly by an excellent controller and its firmware, not by memories themselves.

This means SandForce controllers are not the only ones designed to offset the use of bad memories.
After all, as I have read somewhere, the difference between mainstream-class and enterprise-class SSDs, currently depends mosly (or only) on the controller's firmware.
So the statement of Tom's Hardware, I fear it is valid not only for SandForce (if we also consider the introduction of TLC NAND on Samsung 840 not-PRO series managed by the same MDX controller).

This is not to say that, of course, what you think about the SandForce controller is wrong. It is only to discuss. I imagine that your opinion is based also on bad experiences and this matter more than anything.

It should also be said that if the world of traditional hard drives, in recent years has dramatically worsened, the world of SSD seems a disgusting evolving: it is not possible that a mainstream class product can theoretically corrupt the file system on my PC if a power failure occurs or if it isn't sufficiently stable ...and it is not possible that the producers (including samsung) are not obliged to clearly show what is the endurance and reliability of their devices according to the JEDEC standards.

It is not written anywhere that a mainstream-class product should be less reliable than an enterprise-class one. This differentiation is only artificial (if we consider that often depends only on the firmware). This approach is typical of thieves and this world is increasingly of them ...but unfortunately this is not a bug only in the computer world but in all the fields of our life.
 
Last edited:
There is a difference between a "distributor" and a retailer. Distributors normally are the middlemen who buy products from manufacturers and then sell to retailers who then sell to you. Few, if any, distributors sell directly to end-users. But if that Comay is something you really want and they do sell to end users, great. I'd be worried about warranty returns, though.

It is not written anywhere that a mainstream-class product should be less reliable than an enterprise-class one. This differentiation is only artificial (if we consider that often depends only on the firmware). This approach is typical of thieves and this world is increasingly of them ...but unfortunately this is not a bug only in the computer world but in all the fields of our life.

I totally (and completely) agree :).

When the reliability of a product becomes something that differentiates between market price levels and/or market "class", its only the consumer that loses. And unfortunately overall product quality with many things has been decreasing for several years.

Heck, I have still-working-great ancient Maxtor IDE hard drives over a decade old. Be interesting to see in the future if any Seagate or WD consumer spinning-disk drives purchased today can last as long.

Regarding the RAISE system, if I'm not mistaken, the Intel 710 and 320 series (for example) with proprietary controller, implement a similar redundance system. And I think Intel doesn't use memories of poor quality. In any case, I too have always wondered if RAISE and overprovisioning technologies are substantially the same thing and if it is redundant to have both at the same SSD

The 710 (expensive enterprise) & 320 (consumer) use Intel branded controllers that are going on four to five years old. If you want a modern, fast drive, those models are outdated. Both have nowhere near the performance of modern drives. OK for reliability, though. The 710 has been succeeded by the DC S3700 (mentioned by me here of course), which also uses an Intel controller. The 320 uses the same (old, slow) physical controller as the 710 (3 gbps vs 6 gbps SATA connector for current drives, etc.). Its successor is the 330, which uses a SandForce controller.

As far as I can tell, RAISE and over-provisioning technologies are substantially the same thing. "RAISE" is a marketing term/trademark only used by SandForce .The exact details on what each controller manufacturer does for over-provisioning may differ a bit, but the end results are the same. And every controller & drive manufacturer uses over-provisioning ("redundancy"). Including Samsung.

As noted before, I wouldn't touch a SandForce drive currently with a 10 foot (soggy) pole. But will hold out hope that the new owners LSI will improve SandForce's miserable reliability record.

One other thing about reliability not previously covered: firmware updates. The speed of providing (and quality) firmware updated by a drive manufacturer when issues are detected is also important.

My recommendations for SSDs remain the same as when we first started the conversation. For best overall value (reliability + price + performance), I'd get the Samsung 840 Pro. For ultimate reliability regardless of price (but not as good a value), I'd get the Intel DC S3700.
 
Last edited:
Heck, I have still-working-great ancient Maxtor IDE hard drives over a decade old. Be interesting to see in the future if any Seagate or WD consumer spinning-disk drives purchased today can last as long.

I can say the same for my old mainstream-class Quantum Hard Disk (acquired in year 2000).

Yet, there have always been told that free markets and competition should help increase quality and lower prices ...obviously there must be some anomalies
 
That's odd. Traditionally, you can't use both a CPU GPU and an offboard card at the same time. Not familiar with any modern board that allows that. If you really need that capability, get two offboard cards.

Thats definitely not true. All newer boards with an H or Z chipset I used have an option to enable both. That includes brands like Asus, Gigabyte, MSI and Intel. Usually that option is called something like "iGPU multimonitor" or similar. I currently use that feature on an P8Z77 WS and it works like a charm (I do not use Virtu, it is not compatible with SLI). The problem was entirely with the OS in the past, because Windows XP did not allow multiple graphics drivers. Starting with Vista this became possible and Linux allowed that for a longer time already.
 
Thats definitely not true. All newer boards with an H or Z chipset I used have an option to enable both.

Heh. Like I said "Not familiar with any modern board that allows that". Which doesn't mean "there aren't any". Unlike some people, I don't know everything (damn it) :D. Am very familiar here with multiple offboard cards (of the same make) providing multiple display outputs; they have been around for years.

But haven't looked into the particulars of recent Intel (or AMD) boards providing the same capability with both onboard/on-CPU graphics + an offboard card with different drivers. Thus my recommendation then to contact Supermicro if interested in using such technology with this C216 chipset board. Since its apparently available now on other chipset boards as you say, great.
 
Here, finally, the response from the Supermicro technical support:

My question:

I would like to know for sure if the motherboard in question (X9SAE) is able to support the simultaneous coexistence of a iGPU (graphics integrated in CPU) and a discrete video card so that you can possibly make use of NVIDIA optimus/sinergy technology or run the switch between the two GPUs without having to restart your pc and change BIOS settings.
In practice, I would like to install a XEON E3-1245V2 CPU (with HD4000 integrated graphics) and a NVIDIA QUADRO 2000 graphics card being able to use either the one or the other card, if necessary, without having to restart the computer and change the BIOS settings.

No matter if the currently, NVIDIA Synergy technology is still under development (this is only an example). I just want to know if the operating system (windows or linux) will be able to detect the presence of both the graphics processor and use them correctly loading drivers of both GPUs.


Their answer:

Currently there is no simultaneous support of onboard iGPU and video card. You'll have to make
the switch through BIOS settings.



It must be said that the responses of the technical support (regardless of the manufacturer) are always extremely verbose ...it seems to speak to the sibyl
 
Last edited:
It must be said that the responses of the technical support (regardless of the manufacturer) are always extremely verbose.

:D:D

Sorry to hear, though. Had a suspicion this was the case with no mention of simultaneous support in the BIOS, but good you were able to get a definitive answer.

True server boards (from Supermicro) often have integrated Matrox graphics that can either be enabled or disabled as is traditionally the case; no simultaneous usage.

Another option, perhaps, is combining one of the new AMD A8 or A10 "Vishera/Piledriver" CPUs (with integrated ATI graphics) and a performance oriented 990X/990FX chipset motherboard that has support for ECC memory (and those newer CPUs with updated BIOSes). Such boards I've only seen from Asus. The newer FM2 boards for those CPUs I haven't seen with ECC support yet. And I know ATI Linux video support hasn't been top notch, along with not knowing here whether simultaneous graphics (with an Nvidia or ATI offboard card) is even possible with those boards.

I'd look into it, though, if that's a crucial feature. Or even looking at the non-AMT, less than ideal Asus P8C WS again...if it does have that support (I don't know if it does). Or finally one of the many non-ECC supporting motherboards that allow such support. But that's not ideal.

Personally, I'd get the X9SAE and go through rebooting if I wanted to use different video for power-saving. Or if simultaneous video is more important, go with the X9SAE-V and use two offboard cards, switching to the CPU video when you want to save power. As you plan on an expensive Quadro card(s), the 8x speed of the -V board slots is sufficient. Frankly, Quadro cards are not worth the money in my opinion unless you earn your living off PC video/CAD production work. Most and/or all of the differences between consumer and Quadro cards are in the drivers, not the hardware capabilities.

Good luck finding that simultaneous onboard+offboard video capability on ECC-capable motherboards.

Edit: if you do find one, let us know what it is :).
 
Last edited:
...Or if simultaneous video is more important, go with the X9SAE-V and use two offboard cards, switching to the CPU video when you want to save power...

I cannot understand this statement. As already said, simultaneous video is important only between an integrated iGPU and a discrete video card in way to use discrete GPU only when I need more graphics power or CUDA. The X9SAE-V wouldn't resolve this problem.

I will try to fall back on a motherboard with chipset Z77 or Q77. According to the specifications of the Z77 chipset, Vt-d should not be supported, however, I see a lot of motherboards with this chipset that support VT-d. I wonder how it is possible (perhaps they mount a dedicated additional controller?)
 
Last edited:
I cannot understand this statement. As already said, simultaneous video is important only between an integrated iGPU and a discrete video card in way to use discrete GPU only when I need more graphics power or CUDA. The X9SAE-V wouldn't resolve this problem.

Sorry. Thanks to the length of this thread + my limited time, I didn't recall this particular usage requirement. I just recalled "power savings" and "simultaneous video". Later, I recalled "ice cream" :D.

That particular "more power" usage is still possible with either the regular board + one discrete card or the -V + two discrete cards. Just would not be able to use the integrated video at the same time as the others, of course.

I will try to fall back on a motherboard with chipset Z77 or Q77. According to the specifications of the Z77 chipset, Vt-d should not be supported, however, I see a lot of motherboards with this chipset that support VT-d. I wonder how it is possible (perhaps they mount a dedicated additional controller?)

Don't know about that. Even some Intel branded Z77 boards support VT-d. Looking at the block diagrams of a couple doesn't show any additional "VT-d controllers". What determines VT-d support is apparently a board's BIOS and the CPU, not the board chipset.

One thing for sure, though: no "K" series CPU supports VT-d.
 
Last edited:
Has anyone figured out how to get iAMT 8.0 bios KVM to work while a discrete graphics card is installed? I would love to have the option for BIOS to use onboard GPU if the CPU has one (so remote BIOS access can work) then have the discrete graphics adapter take over later on in the boot process.

A main reason why I bought this board is remote BIOS access (AMT 8.0) and I do not want to give up my Geforce GTX 660ti.

If SuperMicro reads this please give it some thought. Thanks!
 
I finally go into contact with SuperMicro and they're charging me $50 for bent cpu pins even though the pins weren't bent and the socket cover was installed properly.
 
Here, finally, the response from the Supermicro technical support:

My question:

.........In practice, I would like to install a XEON E3-1245V2 CPU (with HD4000 integrated graphics) and a NVIDIA QUADRO 2000 graphics card being able to use either the one or the other card, if necessary, without having to restart the computer and change the BIOS settings.

No matter if the currently, NVIDIA Synergy technology is still under development (this is only an example). I just want to know if the operating system (windows or linux) will be able to detect the presence of both the graphics processor and use them correctly loading drivers of both GPUs.


Their answer:

Currently there is no simultaneous support of onboard iGPU and video card. You'll have to make
the switch through BIOS settings.



It must be said that the responses of the technical support (regardless of the manufacturer) are always extremely verbose ...it seems to speak to the sibyl


That sucks. My basic boxdh67 Intel mother board supports both at the same time and I remember this on much older boards too. Bios only asked to select the primary GPU.

For me this won't be such a big deal since I am using it as a server BUT if this was my workstation I would definitely want more than 2 monitors. It is a workstation for goodness sake! Anything less than 3 screens for work always has me wishing I had at least one more screen.

Well I guess there is a reason that it is $200 and I am glad for the AMT.
 
X9SAE-V-O is the retail package. MBD-X9SAE-V-B is the bulk package. The bulk packaging will probably be a plain box with some of the accessories.
 
I plan to utilize my server mostly as a virtual host, so I went with the X9SAE-V and a Xeon E3-1245V2. I'm going to play with Hyper-V in Windows Server 2012 first, but have been thinking about going with ESXi as my host hypervisor, and at some point, I'd like to look into running a ZFS file server.

Other than a few hiccups during the setup and installation, I really like this system and so far I'm glad I went with a workstation board over a consumer chipset.

How's everyone else's running? Any thoughts on which hypervisor I should settle on before I get too far down the rabbit hole?
 
Last edited:
So, I had a plan, and it changed several times, and now I see that motherboard that seems perfect for me (the V one).

My new plan is to buy the mobo along with the cheapest Intel CPU there is (and is so latest tech I have to order it in another country), the Celeron G1610.

The beauty of that plan is that I can use an old kit of 3*2GB DDR3 from my 1366 desktop rig and start my OpenIndiana server with that strange combination. Then in a month I can order a couple of 8GB ECC sticks, and after that if the Celeron is inadequate I can go the i3 route or splurge on a Xeon.

Wadayathink ?

Any trouble with OpenIndiana ? Any issue with the IBM M1015 HBA ?

:)
 
Last edited:
I have my HTPC running with the x9sae-o and m1015 flashed to lsi9211-8i with 8 drives connected, 24/7. Runs great! I have not tried the latest bios yet and am curious whether this will solve the cacophony of beeps during boot.
 
Thanks. Is there anything in the bundle worth the trouble of figuring out if I'm getting it or not ?
 
Here, finally, the response from the Supermicro technical support:

My question:

I would like to know for sure if the motherboard in question (X9SAE) is able to support the simultaneous coexistence of a iGPU (graphics integrated in CPU) and a discrete video card so that you can possibly make use of NVIDIA optimus/sinergy technology or run the switch between the two GPUs without having to restart your pc and change BIOS settings.
In practice, I would like to install a XEON E3-1245V2 CPU (with HD4000 integrated graphics) and a NVIDIA QUADRO 2000 graphics card being able to use either the one or the other card, if necessary, without having to restart the computer and change the BIOS settings.

No matter if the currently, NVIDIA Synergy technology is still under development (this is only an example). I just want to know if the operating system (windows or linux) will be able to detect the presence of both the graphics processor and use them correctly loading drivers of both GPUs.


Their answer:

Currently there is no simultaneous support of onboard iGPU and video card. You'll have to make
the switch through BIOS settings.
I dont really understand. Can you switch between 1245v2 graphics and the Nvidia card - on the fly? Without rebooting?

When you boot up the PC, you will see the splash screen and then the BIOS. Who shows this splash screen? The graphics on the 1245v2 cpu, or the Nvidia card? If the cpu shows the splash screen, what happens if I switch to Nvidia card, will the mobo instantly switch to the Nvidia card, or will it reboot so the Nvidia takes over? Do I need to reboot before switching graphics system?
 
With nvidia synergy the main display is plugged on the IGP and the switching is handled by the driver so in the BIOS the IGP should be the one displaying.

Then in windows if you launch a 3D application the nvidia card will do the work, send it to the IGP through the PCIe, and the IGP will display it.
 
Can anyone else confirm what ram they have successfully working and not working? I know the kingston and hynix were mentioned earlier but since this will be my first ecc mobo and dont have any spares I want to make sure I purchase a confirmed set.
 
I dont really understand. Can you switch between 1245v2 graphics and the Nvidia card - on the fly? Without rebooting?

No, you have to select it in the bios. It's just like any computer with onboard graphics, and you install a gpu afterwards.
 
Can anyone else confirm what ram they have successfully working and not working? I know the kingston and hynix were mentioned earlier but since this will be my first ecc mobo and dont have any spares I want to make sure I purchase a confirmed set.

I also was able to get my Samsung 1600MHz low power non-ECC memory working, I also put in my i3-2120 in my newer X9SAE-V and my Kingston ECC memory worked. I was unable to verify if ECC was working properly or not, but it was functioning.
 
Can anyone else confirm what ram they have successfully working and not working? I know the kingston and hynix were mentioned earlier but since this will be my first ecc mobo and dont have any spares I want to make sure I purchase a confirmed set.

I bought 32 (i.e. 4*8) GB Samsung DDR3-1600 ECC with Module Manufacturer Part # M391B1G73BH0-CK0, which is not certified for the X9SAE, but it passed two runs of memtest with no problems at all.
 
I have a few specific questions, and I would really appreciate if someone who owns this board would TRY the following.

Add a discrete video card to one of these boards that is also populated with a XEON E3-12x5V2 CPU. Set the integrated GPU to be the main one is BIOS. Then, in Windows, can you attach displays to both the integrated port and the add on card, and have Windows see/use both? There has been a lot of back and forth about this topic, and many people talking about virtu and such. I have no interest in that. I must drive 5 monitors, and my add on card can drive 3. I want to know if the board will run the other two. I also do not want to just add another discrete card (since someone will probably suggest that), because I want to make use of Quicksync.

Now my next question, has anyone tried to use Quicksync with this board, and does it work?

I had an Asus P8C WS working great since last year, but the motherboard completely died on me yesterday, so now I am looking for a replacement. It was able to do all of the things I listed above without problems (except dying within a year of purchase). It was actually quite nice because regardless of what I set in BIOS, I was able to tell Windows to set a monitor on the iGPU as the main one, and then use Quicksync. I could also easily flip it the other way in Windows display properties and play games or whatever if I wanted the horsepower of the discrete card (all without any reboots or cares for what the boot up video was set to).
 
I have upgraded the bios to 2.0a on my x9sae board, now Solaris 11.1 hangs on boot up. Has anyone experienced this problem?
 
I bought 32 (i.e. 4*8) GB Samsung DDR3-1600 ECC with Module Manufacturer Part # M391B1G73BH0-CK0, which is not certified for the X9SAE, but it passed two runs of memtest with no problems at all.

I never managed to find memory problems with memtest. Prime95 worked much better at that.
 
I have upgraded the bios to 2.0a on my x9sae board, now Solaris 11.1 hangs on boot up. Has anyone experienced this problem?


Managed to get fixed with help from guys in the Oracle forum. System now boots fine if I disable Intel VT-D in the bios;
-> Integrated IO Configuration
-> IIO Revision L1/P0/E1/NO
->Intel VT-d [Disabled]

;)
 
I also was able to get my Samsung 1600MHz low power non-ECC memory working, I also put in my i3-2120 in my newer X9SAE-V and my Kingston ECC memory worked. I was unable to verify if ECC was working properly or not, but it was functioning.

Hey I would just like to confirm with you that you did actually manage to have an i3 running ECC ram with an X9SAE motherboard. I am rather hesitant to purchase the motherboard because it does not officially support ECC on i3/i5/i7 processors but I read that you managed to have a system running with such a configuration? Also, which Kingston ram did you use specifically?
 
Back
Top