Decklink card slows down playback

Hi all,

I'm almost done with my new workstation (CentOS is finally working). But today I've discovered a weird bug. When I enable my Decklink 4K Extreme 12G in DaVinci preferences and I play any media (just a WAV for example), I'm stuck at 19 fps (on a 25 fps timeline). So the playback is very choppy.

This only occurs when monitoring output is set to UHD/4K, not HD/2K, on both Windows and CentOS O.S. So I thought first about a limited bandwith on the PCI slot... but it doesn't seem to be the case:

4b:00.0 Multimedia video controller: Blackmagic Design DeckLink 4K Extreme 12G
Subsystem: Blackmagic Design DeckLink 4K Extreme 12G
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0
Interrupt: pin A routed to IRQ 123
Region 0: Memory at b1e00000 (64-bit, non-prefetchable) [size=1M]
Capabilities: [40] Power Management version 3
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1+,D2+,D3hot+,D3cold-)
Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [48] MSI: Enable+ Count=1/1 Maskable- 64bit+
Address: 00000000fee00000 Data: 0000
Capabilities: [60] Express (v2) Endpoint, MSI 00
DevCap: MaxPayload 512 bytes, PhantFunc 0, Latency L0s <64ns, L1 unlimited
ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset- SlotPowerLimit 0.000W
DevCtl: Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
MaxPayload 512 bytes, MaxReadReq 512 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
LnkCap: Port #0, Speed 5GT/s, Width x8, ASPM L0s, Exit Latency L0s unlimited
ClockPM- Surprise- LLActRep- BwNot- ASPMOptComp-
LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- CommClk+
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed 5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
DevCap2: Completion Timeout: Range B, TimeoutDis-, LTR-, OBFF Not Supported
AtomicOpsCap: 32bit- 64bit- 128bitCAS-
DevCtl2: Completion Timeout: 50us to 50ms, TimeoutDis-, LTR-, OBFF Disabled
AtomicOpsCtl: ReqEn-
LnkCtl2: Target Link Speed: 5GT/s, EnterCompliance- SpeedDis-
Transmit Margin: Normal Operating Range, EnterModifiedCompliance- ComplianceSOS-
Compliance De-emphasis: -6dB
LnkSta2: Current De-emphasis Level: -6dB, EqualizationComplete-, EqualizationPhase1-
EqualizationPhase2-, EqualizationPhase3-, LinkEqualizationRequest-
Capabilities: [100 v1] Device Serial Number 00-00-00-01-01-00-0a-35
Kernel driver in use: blackmagic-io
Kernel modules: blackmagic_io

I tried another software which uses my decklink card: same bug. I enabled "Above 4G Decoding" in the BIOS without any success. Forcing to work in PCI 3.0 instead of PCI 4.0 neither.

It looks like a buffer configuration / bug in the drivers (latest, released today) because I can get smooth playback during around 1-2 sec.

Hardware specs:
Gigabyte Aorus TRX40 Xtreme
AMD 3990X
64 GB RAM
SSDs
Radeon VII (x2)

Any help appreciated.

Thanks.

Strange discover:

The log created from Decklink Video Setup says:

PCIe Status

Link Width
1x

Link Speed
5 Gbps

It should be 8x.

Why does the O.S. detects 8x but not Decklink Video Setup?!

Might need power to the card besides from PCIe slot.

The power cable is plugged. I've just found the problem:

The motherboard can't handle 2 GPU + 1 Decklink (above x1) + Thunberbolt Add-in card at the same time.

It displays "d4" error which means too many lanes when I swap the Decklink and the Thunderbolt card, even by enabling 4G decoding and disabling CSM in the BIOS. This is surprising because there should be enough with the TRX40 chipset. I suspect a BIOS bug.

How many displays you running?
Also M2 drives share the PCIe lane space so calculate how many lanes you are using. Two GPUs at x16? 1 x Decklink at x8? 1 x Audio card at x4? 2 or 3 M2 nvme drives? It soon adds up.