- Surely, a shiny new
PCI express Gen 4 system is gonna absolutely blow the doors off your old Gen 3 one, right? Well, yes but maybe not for the reason that you would think. MSI sponsored this video where
we are gonna be digging deep into the performance impact of turbo charging your PCI Express speed. Why did they sponsor that
when they're marketing so prominently features PCI
express Gen 4 you might ask. Cause MSI don't give a
(beep) what PCIe Gen you buy as long as it's MSI. So LOL, here we go. (upbeat music) If you've never heard of
our Techquickie channel and you're wondering what... (crashing) I think this back plate
actually took the brunt of it. - Oh, nice - If you've never heard
of our Techquickie channel and you're wondering what the
SAM Hill a PCI Express is, the TLDR is it's the method that modern CPU's use to communicate with virtually everything else in your PC and the total bandwidth available to it for high-speed graphics or storage or networking is
determined by two things. The number of lanes the CPU can allocate to a device and the speed at
which those lanes can operate. Think of PCI Express,
kind of like a highway. Adding more lanes is challenging because it requires more interconnects which makes the components larger and more complex adding costs. By contrast if you could just
make all the cars drive faster that's way more economical, as long as you can keep
them from crashing. So engineers at PCI-SIG and its members have been working hard
to improve data integrity allowing these PCI Express links or lanes to reach higher
and higher clock speeds. Each generation since the first has roughly doubled the speed of each individual lane
with PCI Express Gen 4 offering nearly two gigabytes
per second per lane. That means that our featured card here an MSI GeForce RTX 3090 Suprim X can be fed at nearly
32 gigabytes per second on a compliant motherboard,
a good thing right? Because it not only boasts the most powerful GPU on the market it's factory overclocked to boot. So armed with only the
information you have so far you'd probably think
that PCI Express Gen 4 is an absolute necessity for gamers and power users everywhere. But a theory is just that, a theory. We need to put it to the test. So we'll be using a Ryzen nine 5950X and a top of the line, MSI
X570 GODLIKE motherboard to remove as many other system
bottlenecks as possible. We're also gonna throw an
RX 6900 XT into the mix for team red representation. Starting with Nvidia, our RTX 3090 pulls roughly the same numbers in
Shadow of the Tomb Raider with a slight dip in Gen 3,
95th percentile minimums. What that means is that the experience
is generally the same except that when the
game hitches or stutters, it's a little bit more severe on Gen 3, though we couldn't detect it
with it human eye in this case. In fact, the difference
wasn't much beyond our margin of error in any of the tests that we ran including Forza Horizon four, F1 2020 and Red Dead Redemption two, where Gen 3 frame rates are consistently a framer two behind Gen 4 but no more. Where the pattern
doesn't hold is in CS GO. Check out that 60 FPS
minimum frame rate drop and 50 FPS average loss. For whatever reason, this
game seems to absolutely hammer the PCI Express bus compared to other newer
titles that we tested. So the faster these transfers
can happen, the better. On AMD, the story is much the same Shadow of the Tomb Raider
doesn't see any variation and we're only a frame
off in Forza Horizon four. F1 2020 and Red Dead Redemption two also show only a modest difference
on Gen 4 versus Gen 3. But then again, look
at those CS GO scores. We've gone from over 400
FPS average with Gen 4 to just 311 in Gen 3. That's a difference of almost 30%. From these results then, it's clear that these
kinds of gains are rare but they are out there. If you're running the right applications. Also, we expect this speed
to become more important as direct storage makes
its way into PC games. If you haven't heard of direct storage it is functionally similar to the way that the PlayStation five and the Xbox series is
allow the graphics chip to bypass other system
bottlenecks and directly access your game storage
drive for faster loading times and Sony says with their implementation even real-time asset streaming. That still ways off PC though. So in the meantime, is there anything that can
reliably push Gen 4 today. Productivity maybe? On the Nvidia side, it
starts out pretty bleak. Blender sees only a modest
improvement in KUDA mode while optics is a complete wash but then we quickly found
an improvement in V-Ray. Where we saw measurable bumps in both KUDA and RTX enabled workflows. DaVinci Resolve two, gives us
a modest little performance increase and the same is
true for Adobe Photoshop. Although that last one in particular was somewhat unexpected. Things start to peter out
again with LuxMark four, though where we're looking at a roughly 3% improvement and OctaneBench doesn't
meaningfully improve at all. As for SPECviewperf 2020
most of this is so close that we might as well call
it a tie outside of Maya's stand out 1% increase over Gen 3. Moving over to AMD, the Blender OpenCL result is basically the same
between Gen 3 and Gen 4 but unlike Nvidia DaVinci
Resolve and Adobe Photoshop don't see a major improvement
in performance either. Curiously though, where Nvidia stumbled at LuxMark, AMD shines with a hefty 27% improvement in the Food Render and
11% in the Hall Bench. That is definitely nothing to sneeze at if you use LuxCoreRender. Finally, SPECviewperf brings me back to they are the same picture, territory with little to
show for the extra bandwidth that Gen 4 offers. Again, Maya stands out
as maybe slightly better but it's nothing that
you're going to want to replace your motherboard for. These numbers, both agree
with and also contradict some of the earlier results
found not only by us when we reviewed the Radeon RX 5700 series but also by GamersNexus, who did a full video
on gen four performance back in September. What this shows us is that
there's been some quiet movement between then and now with
advances in CPU performance on the platforms that support Gen 4 perhaps being the biggest. We used a Ryzen 5950 X today. That is a significantly faster CPU than the 3900 X that we used last time. And the 3900 XT that
GamersNexus used in their video. So, could it be that Intel's
upcoming Rocket Lake CPU's could hold the key to even
better Gen 4 performance. We will be checking it out. So get subscribed to make
sure you don't miss it. Whatever Intel it has
up its sleeve though, PCI Express Gen 4 and up for that matter, really weren't made for consumers anyway, at least not directly. I mean, it's great for hooking up more devices to a motherboard chip set. A faster link here means
faster storage ports faster network ports, more USB
ports, et cetera, et cetera. But for most people, the benefit of Gen 4 is pretty small compared to
simply having a faster CPU. It's just that you can't
really get a faster CPU without going Gen 4. Now, the real reason PCI
Express keeps getting faster is because of the data center. And then we just get their
trickle-down technology. In a data center, it's
not so that they can just shove faster devices into their racks. I mean, that is a thing, but it's more about being
able to shove more devices into them and using
features like bifurcation and devices like PCI Express
switches to split the lanes. If you've got faster lanes,
you not only can have faster devices for the
same number of lanes. You can have more devices
with the same number of lanes without sacrificing the
speed that you already have. Incidentally, that's
exactly what we're doing with our high-speed MVME storage server. We've got a gen four carrier card that has eight gen three SSDs
on it running at full speed. If we put that card into a Gen 3 slot they would end up being
bottlenecked because natively that slot only has enough
lanes for four SSDs. Which actually got us thinking, what if I'm not just a gamer? What if I want to use my
system as a workstation by day and gaming rig by night and
plug in like two graphics cards for some heavy workloads. Ah, okay. So some motherboards like our MSI X 570 GODLIKE are capable of bifurcating
PCI Express lanes. That means they can take a 16 lane slot like this one up here
and split those lanes out to multiple physical slots, allowing more than one
device to share the bandwidth that otherwise would have
gone to a single device. To find out if this helps, we installed a second
RTX 3090 in our system and ran our benchmarks, at
least the ones that benefit from dual GPU, in Gen 3
and Gen 4 modes again. And turns out that there's not
much benefit here right now either, The few benchmarks where
there were improvements with Gen 4 we're V-Ray by about two or 3% and Lux Mark by about a
single percentage point. That's it, a little anticlimactic. But if you think about it,
these kinds of workloads aren't going to scale linearly in terms of bandwidth required anyway, and where we'd be more likely to see a significant improvement is in crunching deep learning data sets. But that's a workload that does skew quite a bit closer to the data center than to the desktop. Also in the data center they
might have quite a lot more than two devices per 16 lanes. Anyway, don't fret. It took a while for Gen
3 to be fully utilized when that took over from Gen 2 as well. So now that AMD has had
a generation of support and the Intel's bringing it with their upcoming Rocket Lake CPU's we are headed in the right direction just in time for direct storage to start making proper use of it, we hope. And besides, a few percent improvement
is nothing to sneeze at. You start with some Gen 4 PCI-E toss in some re-sizable BAR
support, sprinkle on faster RAM and a pinch of community optimizations. And all of a sudden, you've got a pretty compelling generational upgrade. So, you know what I say, keep it coming. Just like I keep the sponsors coming. Big thanks to NSI for
providing us all the equipment and sponsoring this deep
dive into PCI Express Gen 3 versus Gen 4 performance. Go check them out. We use their top end stuff here but their gaming X trio cards
are also great for performance and they have a wide
range of motherboards, X 570, B 550 to satisfy your rise and fix. Oh, and of course there's
Z490 boards are already wired for Gen 4 when Rocket Lake launches. So, we're going to have
all that linked down below. Thanks for watching guys. If you enjoyed this one,
maybe go check out our video on the server that Liquid
sent us for a taste of just how fast PCI Express can go when you load it up with enough devices.
