AMD Radeon RDNA2 Navi (RX 6500, 6600, 6700, 6800, 6900 XT)

Every motherboard released in last 2-3 years has PCIe 4.0, so every new system has at least PCIe 4.0 slot.
I thought Intel skipped PCIe 4.0 and was on 3.0 until the very latest Alder Lake?
 
I thought Intel skipped PCIe 4.0 and was on 3.0 until the very latest Alder Lake?
Nope, mobile had it since Ice Lake and desktop since Rocket Lake (also Comet Lake was supposed to have it but they axed it when they couldn't get it stable)
 
Nope, mobile had it since Ice Lake and desktop since Rocket Lake (also Comet Lake was supposed to have it but they axed it when they couldn't get it stable)

Didn’t rocket lake launch less than a year ago? Intel hasn’t had PCIe 4.0 for very long. Either way a lot of people are still running PCIe 3.0 systems.
 
Rocket Lake launched in Mar 2021, that was Intel's first consumer PCIe 4.0. Otherwise Ice Lake for servers also was PCIe 4.0 which launched shortly after. Ice Lake mobile was only PCIe 3.0. Tiger Lake was when Intel mobile moved to PCie 4.0.

Using motherboard releases is a bit misleading in terms of actual PCIe 4.0 adoption.

X570 did release in July 2019 with Zen 2 but it was considered the "premium" tier for boards and the high end option. Even the cheaper B550 motherboards (a year later in June 2020) did not actually replace B450 motherboards. B450 have been selling the entire time as a significantly cheaper alternative. This along with A320 motherboards mean plenty of PCIe 3.0 motherboards are in use with AMD CPUs that do support PCIe 4.0 (Zen 2, Zen 3). Desktop APUs are still PCIe as well regardless of motherboard.

Comet Lake (10xxx series) on Intel's side is also still commonly sold. It was not fully replaced by Rocket Lake (Intel's <6 core 11xxx CPUs are also Comet Lake still and therefore PCIe 3.0), nor by Alder Lake yet.
 
The other additional thing to consider here is reviewers tend to test all max (well except RT, but lets not go there :devilish:) settings while individual setting adjustments are available. If it's memory pressure causing causing the most performance drop then it can be a situation in which this particular card would need to sacrifice more memory sensitive settings (typically texture quality related) relatively more so than other settings to bring performance up when compared to others. However at the same time we could also debate the issue of those max likely texture settings.

I feel though in general this is an example though of a situation in how while the traditional review format can reflect the academic performance but doesn't necessarily convey the actual usability performance.

On a semi related note. Has anyone looked into whether or not PCIe limitations may affect games that rely more on streaming of data to manage memory load in terms of things such as more texture pop in?
Performance is work done per unit of time. If you want to comparatively quantify one, you need to keep the others constant.
Taken to the extreme: Yes, *any current GPU* is a great choice for playing minesweeper or CS 1.6.
 
Apparently you missed "worst case" in my posting. RTX 3050 performance is lost on PCI Express 3.0 systems in some games.
Nope, I did not miss it, but tests presented in this thread seem to indicate that this single, minium fps result is very much an outlier result. Are you into worst cases being treated as the basis for a discussion? Me, I'm not.

Also, the 8% performance loss (it's a 9% gain) gets smaller with higher VRAM-utilization (FHD -> QHD in that very same article), without sky high frames per second.
 
Nope, I did not miss it, but tests presented in this thread seem to indicate that this single, minium fps result is very much an outlier result. Are you into worst cases being treated as the basis for a discussion? Me, I'm not.
4 out of the 6 games listed have a performance loss due to PCI Express 3 at 1080p.

The narrative that RTX 3050 is immune to PCI Express 3 performance loss at 1080p is a lie.
 
4 out of the 6 games listed have a performance loss due to PCI Express 3 at 1080p.

The narrative that RTX 3050 is immune to PCI Express 3 performance loss at 1080p is a lie.
Well, it's not any narrative of mine or even one which I particularly care about. To be honest, I haven't even seen this narrative mentioned here in the thread except by you, so I'm not sure about the point you're trying to make.
 
Well, it's not any narrative of mine or even one which I particularly care about. To be honest, I haven't even seen this narrative mentioned here in the thread except by you, so I'm not sure about the point you're trying to make.
I've seen it mentioned in quite a few softball reviews where they talk about how "you'll probably not have any issues with only 4 PCI lanes gaming at 1080 as you'll rarely exceed the VRAM limit", (I'm paraphrasing, not from any particular review), and I believe that's the narrative they're talking about.

It's the, "Well it's good enough for the most part", syndrome going on with this card. It's really not, not at all, it's just if there's nothing else at least it's something.

I repeat: the 6500 XT is a terrible product, I would damn near get anything else over it if possible.
 
Fortunately the "GT" products aren't designed to be full fledged gaming products. And yes, the 6500 XT is a terrible gaming product.
 
AMD Radeon RX 6900 XT Graphics Card With 3.15 GHz Clocks Achieves Top Position In 3DMark Fire Strike & Time Spy Hall of Fame

https://wccftech.com/amd-radeon-rx-...ghz-clocks-achieves-top-position-in-3dmarkrk/

Due to new CPU used for overall score, but hoping for similar normal overclocked numbers with the 6nm refresh.
They've already confirmed RDNA3 for this year and there's no roadmap with N6 RDNA2. You really think they're gonna refresh the RDNA2-lineup?
(Yes, I know Navi24 is N6)
 
Back
Top