you wont regret it. its much easyer to handle than Intel imho. I always needed to tweak a lot with Intel since they introduced the e-cores, and even before. Things might have changed with better scheduler in Win and BIOS updates etc, but AMD is just so simple to use. U can easily set a PBO value, or just leave it for the start, as with the 7800X3D and MSFS i didnt really see noticable performance boosts anyway with PBO. Might be different for other games or applications.
I was an Intel “fanboy” for a long time, but since the 7800X3D, Intel would have to pull some insane magic to get me back on their CPUs…
One tip to check out when you get it:
Setup a benchmark scenario, at a specific time, clear skies weather and disable traffic and anything else that isnt consistent, then write down performance.
Then disable SMT (Hyperthreading) and SVM (Virtualisation) in BIOS and do the exact same benchmark scenario again, Sim-Time matters because of shadows! and see if you get a noticable perf.-boost.
For me, i can easily get about 8-10 FPS more with SMT disabled in some aircraft like FENIX or PMDG. But it might differ.
Meanwhile ASUS even implemented a “Game-Boost” option in the latest BIOS which automatically disables SMT for performance in some games.
I’m curious to know if there will be any (performance) advantage to get a X870. At first I thought the X3D Turbo Mode would be a X870 exclusive, but the Asus last Bios implements it for the X670.
I’ve been using MSI motherboards in the Tomahawk version for quite a while, both with Intel and AMD CPUs. I wanted to try a different motherboard model and the Carbon version is a bit better built - it also comes with a newer BIOS UI version and has better VRM.
The only downside to the MSI MPG X870E Carbon WIFI is the shared bandwidth between PCIE_1/PCIE_2 slots and the M2_2 drive slot. This means you need to be careful when setting up your configuration so you don’t accidentally limit the performance of your main GPU slot.
In other words, if you install the graphics card in the PCIE_1 slot, don’t use the M2_2 slot for your M.2 drive. If you do, you’ll reduce the bandwidth of the PCIE_1 slot to x8.
There’s nothing to be afraid of if you know how to build a pc. After 30 years of using Intel, I switched to AMD, and as you probably know, it was because they’ve let us down.
The only thing you need to keep in mind is the LONG initial boot time on an AMD platform after assembling the computer. It might take so long that you might think something went wrong, but just be patient - it’s normal.
I also get better performance with SMT and virtualization disabled.
FYI, in MSI BIOS, to disable virtualization, you need to disable IOMMU and SVM.
Unless they’ve put the 3D cache on both CCDs, I can’t see how it will perform better. I mean, 3D cache on both CCDs + Turbo Mode (which disables SMT) might perform better on the 9950X3D on games that use more than 8 cores. But still there will be some latency penalties using 2 CCDs.
On impulse I bought an 8tb (7000’ish (?) read speed ) WD SSD to go in the first slot and 2 4tb Samsung 990 pro SSDs to go into 2 other slots. I don’t think I will have any SSD slow downs in this scenario but maybe you know different? As I said the 8tb was an impulse buy and I didn’t do any homework on this at all. Hopefully there won’t be any issues reading from/writing to such a large drive.
I might also want to put a separate ((maybe a Creative (?)) sound card though in the third bottom slot too so that I can connect the audio to my oldish 5.1 audio system. Maybe this will have an effect on some m.2 slots? I guess I will have to look at the user guide again before buying.
you should be able to manage the speed of the NVME slots (8x/16x) in your BIOS. So, in case you run into limitations, you could set a rarely used drive to 8x maybe to make sure the more important stuff like GPU gets the full speed.
You can check the speed of your slots in some tools like CPU-Z or GPU-Z i guess or others, to make sure everything runs as intended.
Edit: actually it does say display port in the product specs. I had downloaded the datasheets to compare them and on those it does not call them display ports.
7950X3D was rated for 4.2 GHz base and 5.7 GHz Boost, while 7800X3D’s official spec is 4.2 GHz base and 5.0 GHz Boost.
A 700 MHz difference sounds like a lot but the higher frequencies were only available on the non X3D CCD.
With the newer chips, the X3D Cache is positioned below the processor, closer to the cooling solution.
Could that result in higher achievable clock speeds on the X3D CCD of higher end CPUs?
If so, the gaming performance would be affected, even if they are ‘hybrid’.
Not an expert on VR at all (can’t remember the last time I even used my Quest 2) but wouldn’t you connect the VR headset to the display port/HDMI port on a dedicated gpu not an igpu? Would VR even be feasible on an integrated gpu?
I could be wrong as it’s ages now since I have used VR. Let me know though please if I am wrong
It could be the case that X3D is not even needed for both CCDs or is not adding a relevant performance gain on the top of the series CPUs that justifies the cost increase. I mean… they are going to be 16 cores CPUs. Most games nowadys and in short time won´t even use all of them and with a fast enough clock you could still get crazy results (people are getting them with 9800X3D indeed). Most likely AMD wants to hold some of the eventual improvements for next generations. It´s marketing at the end of the day and CPUs are ahead of what exisiting software really needs so they don´t really need to rush that into market.
