If you've been living under a rock for the past few weeks and only now decided to build a new gaming PC, we've got some bad news. DRAM pricing has gone a bit crazy, as it tends to from time to time, and this time we're looking at increases north of 100%.
Not long ago, we were asked to test a Ryzen processor with just a single DDR5 memory module. That comment ended up being heavily upvoted, and it struck us as both a timely and genuinely interesting idea. So here we are.
To put the current price shock into perspective: a 32 GB kit of G.Skill Trident Z5 Neo RGB RAM (DDR5-6000 CL30), which we regularly recommend for AM5 builds, has gone from $120 just two weeks ago to an eye-watering $270 – a 125% price hike.
This insane increase has many of you wondering: do we really need two sticks? After all, most games still run fine with 16 GB of RAM. So how much performance are we sacrificing when going from dual-channel to single-channel memory? Today, we're going to answer exactly that.
Obviously, there's no perfect workaround for the current pricing mess. But if we can get away with spending $100 - $120 on a single module now – rather than paying at least twice that for a pair – and then buy the second module later when pricing inevitably drops, that could be a perfectly reasonable money-saving strategy.
Right now, for instance, you can pick up a single 16 GB stick of Patriot Viper Elite 5 DDR5-6000 CL30 for $100. Buy two as part of a 32 GB kit, though, and the price jumps to $255. In that case, buying two individual sticks could actually make more sense.
And considering $100 is about what a full 32 GB kit cost only a few weeks back, getting by on a single module suddenly feels like a much more affordable alternative. Our testing setup is straightforward: we're using a Ryzen 7 9700X, which should represent most of the AM5 lineup – excluding X3D chips, which tend to be less sensitive to memory latency and bandwidth, though capacity can still matter.
In short, we're comparing single-channel and dual-channel performance, with the obvious advantage going to dual channel since it doubles memory capacity. Everything else: frequency, timings, etc. remains identical. We're using the same G.Skill Trident Z5 Neo RGB 32 GB kit for both tests, one run with both modules and another with only one.
Let's get into the data…
Test System Specs
| CPU / Motherboard / Memory | AMD Ryzen 7 9700X | Gigabyte X670E Master [BIOS F38e] G.Skill Trident Z5 RGB 32GB DDR5-6000 CL30 |
| Graphics Card | Asus ROG Astral RTX 5090 OC Edition | |
| ATX Case | MSI Prospect 700R | |
| Power Supply | Kolink Regulator Gold ATX 3.0 1200W | |
| Storage | TeamGroup T-Force Cardea A440 M.2 PCIe Gen4 NVMe SSD 4TB | |
| Operating System | Windows 11 | |
| Display Driver | Nvidia GeForce Game Ready Driver 581.42 | |
Benchmarks
Rainbow Six Siege X
Starting with Rainbow Six Siege, we're only seeing around a 5% performance difference between a single module and two. With this game being heavily CPU-limited at medium settings, that's a reassuring result for anyone hoping to get by with just one stick. Essentially, the 9700X can push close to 500 fps using a single 16 GB DDR5-6000 module.
Battlefield 6
The performance hit is more noticeable in Battlefield 6. In this case, we're looking at roughly a 12% loss. Still, the game remained very playable with just 16 GB of RAM, and the 9700X managed to average over 130 fps with 1% lows around 90 fps. This makes it a viable temporary solution.
Marvel Rivals
Unfortunately, the performance drop in Marvel Rivals is pretty ugly, with frame rates falling from 195 fps to 129 fps at 1080p using the medium preset. However, if you're playing at a higher resolution, with higher quality settings, or using a GPU that usually delivers a bit over 100 fps, the impact of single-channel memory becomes far less substantial.
Since this is meant to be a "make-do" option and a temporary workaround to the current DRAM pricing spike, it's helpful to know that the minimum performance level is still very playable and likely an upgrade over many older systems. Even so, it's wild to see dual-channel memory boosting performance by 51% in this test.
Horizon Zero Dawn Remastered
Next up is Horizon Zero Dawn Remastered, where we see another big drop when using a single memory module. At 1080p, 1% lows fall by 33%, though overall performance remains very playable. At 4K, we still see a 26% decline in 1% lows while using the very high preset with an RTX 5090.
Spider-Man 2
Spider-Man 2 shows nearly a 20% hit to the 1% lows at 1080p, along with a 15% drop in average frame rate when running a single module. That is a pretty big performance drop, although performance is still comfortably playable. At 4K, the gap narrows, with similar 1% lows and a 12% reduction in the average frame rate.
Baldur's Gate 3
Baldur's Gate 3 played quite well with a single memory module. Interestingly, the largest performance drop appeared at 4K, where the 16 GB configuration was up to 23% slower. At 1080p we only saw a difference of up to 10%, and the same was true at 1440p. In all cases, performance remained very acceptable.
Avatar Frontiers of Pandora
Avatar: Frontiers of Pandora is mostly GPU-limited, so the performance hit is not too severe. Even so, 1% lows dropped to around 113 fps. At 1080p the single-module setup was up to 23% slower, but by the time we reach 4K where frame rates fall well below 100 fps, performance is effectively the same.
ARC Raiders
ARC Raiders runs quite well with a single memory module, though like many of the other games we've tested, the 1% lows take the biggest hit. They dropped by 27% at 1080p and around 20% at both 1440p and 4K. The average frame rate at all three resolutions was similar; however, you will notice more frame-time inconsistencies with a single module.
Assetto Corsa Competizione
Next we have ACC. The game is not particularly memory sensitive, so using just a single 16 GB module is not an issue. At worst you sacrifice around 9% performance while still achieving well over 160 fps.
Cyberpunk 2077: Phantom Liberty
When comparing DDR4 and DDR5 performance on the 12400F, we saw a significant uplift with DDR5. However, reducing available DDR5 bandwidth by half on the 9700X appears to have almost no effect on Cyberpunk 2077 performance, which genuinely surprised us.
Counter-Strike 2
Like ACC, Counter-Strike 2 is not very memory sensitive. Running with one or two memory modules makes little difference. At most, we're looking at about a 5% drop when using a single DIMM.
The Last of Us Part II Remastered
The Last of Us Part II Remastered performed well with a single memory module. At 1080p we observed a 13% drop, yet the average frame rate still exceeded 150 fps and 1% lows were above 100 fps. This pattern was also seen at 1440p. At 4K the average frame rate remained just over 130 fps with 1% lows around 110 fps, and both configurations delivered the same performance. For most players, a single-module setup will work fine.
Space Marine 2
Finally, we have Space Marine 2, a very CPU-demanding game. Even with one memory module we were able to average around 90 fps, with 1% lows near 70 fps, which is still very playable. However, performance dropped by up to 22%, and interestingly, as with Baldur's Gate 3, the largest difference appeared at 4K.
13 Game Average
Here's the 13-game average data, calculated using the geomean. At 1080p, the single-stick configuration was 12% slower when comparing average frame rates and 16% slower for the 1% lows. This isn't a huge difference overall, though the results did vary quite a bit from one game to the next.
Increasing the resolution to 1440p didn't change much in our testing, although we are using an RTX 5090. It wasn't until we moved up to 4K that the margins began to shrink. At that point, the single-module setup was just 9% slower for average frame rates and 13% slower for the 1% lows.
Is the One-Stick Strategy Actually Worth It?
So after all that, is buying a single 16 GB memory module now, with plans to add a second once pricing cools off, actually a good idea? To figure that out, we need to look at the real-world savings. As noted at the start of this review, the 32 GB G.Skill kit we typically use has climbed to $270, up from $120 just a few weeks ago.
To our knowledge, you still can't buy those exact modules individually, but some DDR5 options are sold as standalone 16 GB sticks. The Patriot Viper Elite 5 we mentioned earlier is one example, and we also found a single 16 GB G.Skill Flare X5 DDR5-6000 CL36 stick for $116.
If you can pick up a 16 GB module for $120 or less now and then grab a matching stick later for about half that price, you're looking at a total of roughly $180 instead of $270 upfront. Meanwhile, Newegg is selling V-Color Manta DDR5-6000 CL30 32 GB kits in the $220 – $230 range, which might be worth snapping up while stock holds.
Whether the single-module strategy makes sense depends on how tight your budget is. Saving around $100 on a new build is nothing to sneeze at, and for some buyers it could be the difference between finishing a system today or waiting weeks. There are caveats though. The memory you buy now may not exist in retail channels later, and mixing different modules can be finicky unless you're comfortable opening the BIOS and manually tuning primary, secondary, and tertiary timings.
Overall, the results from our testing are surprisingly encouraging. Running a single 16 GB stick in a single-channel configuration didn't tank performance in any title we tested. We never encountered frame rates that dropped far enough to make a game meaningfully worse or unplayable. And since this is intended as a temporary workaround, the small drops at 1080p matter less than what happens at 4K. In this case, what happens at the higher resolution is actually more relevant than typical CPU-bound testing.
Realistically, most people exploring this approach just want to know whether their games will still run well. Based on the data, the risk is small. If performance doesn't meet expectations, the fix is simple: bite the bullet and purchase that second stick. Just keep in mind that mixing modules can sometimes introduce stability issues if the pair doesn't match perfectly.
There's even a clever workaround if you have a friend building a PC at the same time. Buy a dual-channel kit together, each use one stick, and when prices finally drop, put the original pair into one system and split the cost of a cheaper kit for the other. It's a small hack, but in a pricing spike like this, every bit helps.