New CPUs means new chipsets, and (even with their better inter-generational compatibility) AMD is no exception. We’re taking a look at the X570 chipset to see what’s new with this type of motherboard!
X570 gon’ give it to ya!
If you’re not up with AMD products, they generally have 2 main board types for supporting their CPUs. There’s the more affordable and humble B-series boards and then there’s the high-tier, full-featured X-series boards with all the bells and whistles.
Yet what are those bells and whistles?
Let’s break down some specifications to see what all the fuss is about:
Required Specifications for X570
- Up to 16x PCIe 4.0 Lanes
- Up to 12 SATA 6Gbps Ports
- 4x USB HiSpeed 480Mbps
- 8x USB SuperSpeed+ 10Gbps
And that’s just the stuff from the motherboard itself. So you would ALSO get all of the PCIe 4.0, USB, and storage I/O from the AMD Ryzen processor. When you throw one of the larger AMD CPUs on there, you’re getting a frankly monstrous set-up for productivity (that’s also fast for gaming and arguably ‘future-proof’).
Yet that’s not the whole story, so let’s dig a little deeper.
Well I’ll be damned, we actually got there.
For a little context, PCIe 4.0 was announced all the way back in November 2011. Call of Duty: Modern Warfare 3 had just been released for PS3 and Xbox 360 to fight against Battlefield 3 for the Christmas top spot. It was a very different time, but one thing that the standard boffins for PCIe 4.0 decided was that it should have a 16 GT/s bit rate.
For those playing the long game with my articles, you might remember when I talked about data transfer rates, that this meant PCIe 4.0 would have double the available bandwidth when compared to PCIe 3.0
So if we’re using 16 lanes for something (x16) on PCIe 4.0, we have a total of 31,600 MB/s available bandwidth.
Now admittedly, your average user isn’t going to see the advantage here for a time. Graphics cards weren’t even saturating PCIe 3.0, and most M.2 drives hover around the x2 to x4 range for their speeds. Yet where this does benefit is in workstation scenarios and multitasking.
Being Smart with PCIe 4.0 and CPU L3 Cache
Like it or not, your biggest bottleneck in any system is getting the data to your CPU. Once it’s there, it’s dealt with INSANELY quickly and then as it’s passed around the system things slow back down. This is where AMD have been smart in stacking the 3rd-gen Ryzen CPUs with more L3 cache. Although their transfer rate is tied to clock speeds (the all important IPC—Instructions Per Clock), your L3 cache typically has transfer rates of up to 175,000 MB/s.
This was why I was wondering if manufacturers were going to push to PCIe 5.0 as we get 63,200 MB/s speeds.
Each protocol is essentially doubling the last, so even when we get to PCIe 6.0 in the crazy future, we still won’t quite be where L3 cache is at right now—but 126,400 MB/s will be getting seriously close. However, as that standard was only agreed upon in principle in June of this year, we’re going to be a ways off!
Based on current rollout speeds, you’re more likely to see something along those lines in 2022/23. So get ready to download Skyrim again, we’ve got time to kill!
I/O Connectivity—The Devil is in the Details
Now, with all this talk of crazy-fast PCIe lanes and bonkers M.2 SSD drives faster than the ones us commoners own… you may have forgotten that in my quick specifications list above, I also mentioned an awful lot of USB port requirements for AMD to give your board the X570 stamp of approval.
For most users, this honestly is the biggest winner from the upgrade to faster standards. We can throw more and faster ports at a system without it impacting overall system performance. That 10Gbps from USB 3.1 is something that has only really come along because of the available system bandwidth, as before there would have been no real benefit (as we would have hit a brick bandwidth wall given enough devices connected to the motherboard). For reference when comparing with the figures listed above, USB 3.1 is 1250 MB/s. That would have been well into the x2 PCIe lane requirement with 3.0, which is a no-no really; we want low system impact for a plug-and-play device, so 1 lane, no fuss please! This is why USB 3.0 has been used, with a theoretical max of 625 MB/s.
This simple doubling of bandwidth becomes very obvious when you’re looking at overall system design, but don’t be too surprised if you see other devices coming along which will take advantage of the faster standards.
It frees up cable-management-conscious builders to not worry about their shiny 8K+ resolution displays needing multiple HDMI leads to mosaic the image on the screen, and GPU manufacturers to know that they will be able have that sustained throughput from the CPU to the display with zero issues.
This is why AMD are in an interesting place right now: they’re the first adopters in the CPU/mobo space, yet they do design GPUs too. Although their first 7nm GPUs essentially run on the existing standards, they aren’t really pushing out enough data to worry about it yet. But it’s something their CPU department has clearly worked with, and developed a platform around. So surely it’s only a matter of time before we see some really monstrous PCIe 4.0 enabled GPUs?
We can but hope.
Yet what do you think? Is X570 a chipset for the future, or dead out of the gate by being overkill? Are you going to take advantage of the higher bandwidth, or are you happily playing Minecraft at 720p?
Let us know your thoughts in the comments, and if you have any questions about the X570 chipset that we haven’t highlighted, feel free to ask!