Supermicro Outs Z490 LGA1200 SUPERO Pro Gaming C9Z490-PGW Motherboard with Dual x16 PCIe

[btarunr]

Server motherboard specialist Supermicro has been taking steps into the client-segment for a couple of years now under its SUPERO brand. Its latest creation is a socket LGA1200 motherboard based on the Intel Z490 chipset, the Pro Gaming C9Z490-PGW. There isn't too much bling or color, except some contemporary design cues such as the rear I/O shroud, and diagonal accents on the PCH and M.2 heatsinks. The feature-set of this board looks like it's been drawn up by people who design enterprise boards. The board pulls power from a combination of 24-pin ATX and 8-pin EPS power connectors, conditioning it with a 12-phase "server grade" VRM. Its BIOS ROM chip is socketed. There's equal focus on U.2 as M.2.

The star attraction is the PCIe setup. The board uses a PLX PEX8747 PCI-Express gen 3.0 x48 bridge chip to drive two PCI-Express 3.0 x16 slots at full x16 bandwidth, which can split between four slots in x8/x8/x8/x8 configuration when all of them are populated. Storage connectivity includes one each of M.2-2280 and M.2-22110 slots; two U.2 ports with 32 Gbps wiring; and four SATA 6 Gbps ports. Networking options include a 10 GbE interface pulled by an Aquantia AQC107 controller, a second Intel i219-V PHY handling a 1 GbE port, and 802.11ax + Bluetooth 5 wireless connectivity provided by an Intel AX201 module. USB connectivity includes four 10 Gbps USB 3.2 gen 2 ports, and six 5 Gbps gen 1 ports. The onboard audio setup features a high-grade Realtek ALC1220 codec. The company didn't reveal pricing.



View at TechPowerUp Main Site

 

[dj-electric]

This is a trap i see many people fall to for the wrong reasons.
Reminder - using a third party PCIE switch means adding a tiny bit of latency
Reminder 2 - this old switch only supports 3.0 version of the spec

I could see this board used for ML rigs, as at X8 communication for most types of algorithm techniques remain at top speed.

 

[Ferrum Master]

with that funky rad in front of upper PCIe you can not have two GPUs either way

 

[kapone32]

It would actually be interesting to see a comparison between this board and one of the B550 boards that have PCIe 4.0 x8 across 2 lanes in terms of Multi Gpu support and performance.

 

[TheLostSwede]

But only four USB-A ports...

with that funky rad in front of upper PCIe you can not have two GPUs either way

That should be for the 10Gbps Ethernet controller.

 

[Flaky]

Reminder - using a third party PCIE switch means adding a tiny bit of latency

When was the last time anyone did serious analysis how this added PCIe packet latency affects performance?

 

[MDWiley]

Sorry to go off track but who else is noticing similarities in motherboard designs? Like with layouts and heat sinks? Idk if these companies are using the same board manufacturer but it wouldn’t surprise me.

 

[AnarchoPrimitiv]

with that funky rad in front of upper PCIe you can not have two GPUs either way

What are you talking about? There's nothing on the board that would block the use of any PCIe AIC

 

[Totally]

with that funky rad in front of upper PCIe you can not have two GPUs either way

Why? GPU PCBs don't extend past PCI slots which that funky rad is flush with.

 

[Berfs1]

NGL i like the simple but functional look of this

Sorry to go off track but who else is noticing similarities in motherboard designs? Like with layouts and heat sinks? Idk if these companies are using the same board manufacturer but it wouldn’t surprise me.

um, sure they might be using ATX standards to design? I mean almost all consumer motherboards follow a version of the ATX specifications (except for the OEM motherboards like Dell and HP)

 

[Totally]

NGL i like the simple but functional look of this


um, sure they might be using ATX standards to design? I mean almost all consumer motherboards follow a version of the ATX specifications (except for the OEM motherboards like Dell and HP)

Intel releases a reference design and the manufacturers build their boards off of that

 

[Ferrum Master]

Why? GPU PCBs don't extend past PCI slots which that funky rad is flush with.

The cooling systen often does. Had many cases when even cap causes problems installing a card.

The cooler face plate at most cases.

 

[Woomack]

I can only tell you that a gfx card with a backplate still has about 3mm space to that heatsink above the top PCIe slot. No one is manufacturing anything thicker because it's a market standard. The only exception can be quite exotic cards with passive coolers on the back but I'm not sure if there was anything new with that design recently.

 

[Ferrum Master]

I can only tell you that a gfx card with a backplate still has about 3mm space to that heatsink above the top PCIe slot. No one is manufacturing anything thicker because it's a market standard. The only exception can be quite exotic cards with passive coolers on the back but I'm not sure if there was anything new with that design recently.

Strix and WF card top cooler shroud often overlaps PCIe outline boundries and and get stuck against larger caps. If you have them there in the first place.

If you haven't had such problem, good for you but that is normal, becase seldom any mobo maker does something like that unless it is OEM or server. SLI users more often got issues with that with the more bottom slots.

For upper PCIe slot design, having something there is a bad and unusual design practice in the first place for consumer boards.

That is what you get from guys wo are used to designing blade server parts.

 

[basco]

i think if ya run 2 gpu´s its 16\0\16\0

and i think the angle we look at is off so it looks taller.
maybe its not higher then the m2 metal plate

is this really a cooler for intel lan chip same on z390 mobo?

 

[Woomack]

Strix and WF card top cooler shroud often overlaps PCIe outline boundries and and get stuck against larger caps. If you have them there in the first place.

If you haven't had such problem, good for you but that is normal, becase seldom any mobo maker does something like that unless it is OEM or server. SLI users more often got issues with that with the more bottom slots.

For upper PCIe slot design, having something there is a bad and unusual design practice in the first place for consumer boards.

That is what you get from guys wo are used to designing blade server parts.

Motherboards used to have chipset coolers in that spot for years and it's nothing unusual. Also IO cover is closer to the PCIE slot than this radiator (and it's still in the same place as on every other motherboard). You are just looking for an issue that doesn't exist. There is at least 5mm space between the graphics card PCB and the heatsink. Using XFX RX5600XT THICC III Ultra in the 1st slot, there is still about 3mm space between the backplate and the said radiator. I can tell you that as I have this motherboard on the desk next to me.
I've never had any issue with Gigabyte or ASUS graphics card having too thick backplate or anything else interfering with motherboard components (other than mentioned IO covers).

 

[Ferrum Master]

Motherboards used to have chipset coolers in that spot for years and it's nothing unusual. Also IO cover is closer to the PCIE slot than this radiator (and it's still in the same place as on every other motherboard). You are just looking for an issue that doesn't exist. There is at least 5mm space between the graphics card PCB and the heatsink. Using XFX RX5600XT THICC III Ultra in the 1st slot, there is still about 3mm space between the backplate and the said radiator. I can tell you that as I have this motherboard on the desk next to me.
I've never had any issue with Gigabyte or ASUS graphics card having too thick backplate or anything else interfering with motherboard components (other than mentioned IO covers).

Well you cannot overtake the fact that this is an bad design. You are kinda stuck on the backplate issues, I didn't even mention it... Please you are free to show any other lacking design example like this. You are free to look up to the last preview boards... NONE, I mean NONE have something that distant that can compromise upper slot clearance, even BIOSTAR made it right. All designers have noted to put that place empty as they HAVE experience... Putting a rad there where's the GPU hot exhaust is also a fine idea, as there is no other source of airflow there. Like putting a plaster on a dead body. They could not even manage to add proper spacing for the screw hole and the I/O panel shroud. It reeks of hasted bad beta design.

If your setup is fine, leave it be, but you have no point to defend it only because you have it in you shelf, as it could be a unusual flaw for many system builders, I am not the only one having stumbled on such issues during past years exactly on such funky design motherboards.

 

[MDWiley]

NGL i like the simple but functional look of this


um, sure they might be using ATX standards to design? I mean almost all consumer motherboards follow a version of the ATX specifications (except for the OEM motherboards like Dell and HP)

True. Reason I say it is because the similarities are most prominent in newer ITX board heat sinks. Not complaining of course, just found it interesting.

 

[Berfs1]

Intel releases a reference design and the manufacturers build their boards off of that

I know, I was explaining why there are "similarities in motherboard designs".