When it comes to PC gaming, Alienware (owned by Dell) is a big name, and it's synonymous with both high performance as well as high price tags and excess. Dell's recent Aurora PCs have received good reviews but exhibited key problems when it came to size, cooling, price, and even performance. The latest Aurora R16, which initially launched in August, at least partially addresses all of these issues, and now this powerful prebuilt is coming equipped with Intel's 14th-generation Raptor Lake Refresh chips.

Although 14th-generation CPUs are basically the same as their 13th-generation predecessors (with the notable exception of the Core i7-14700K), the Aurora R16 is getting a little cheaper with this hardware update, and the benefit of a slightly faster CPU is (at least in theory) higher framerates for games where you often find yourself CPU bottlenecked.

The reality is that while the Aurora R16 is a good gaming PC, it suffers from many problems inherent in its old-school prebuilt design. The configuration I reviewed has top-end hardware but not top-end performance, and while end users can do some things to change that, ultimately the Aurora R16 can never catch up to a similarly priced PC built using off-the-shelf parts. Despite all that, the Aurora R16 offers a great gaming experience and does benefit in some ways from being a customized prebuilt rather than a prebuilt using normal hardware.

About this review: The Aurora R16 was sent by Dell for review. Dell did not see the contents of this review prior to publication.

Pricing and availability

The Aurora R16 has been available since August, but only from Oct. 17 will the model with the updated Core i9-14900KF be out on the market. The unit I was sent was configured with the 14900KF, RTX 4090, two 1TB SSDs, 32GB of DDR5 clocked at 5600MHz, and a 1000-watt PSU; this model costs $3,500. That same configuration but with the 4090 swapped out for a 4080 costs $450 less at $3,050, which is about the same price difference between typical, standalone 4080s and 4090s that you can buy today.

Interestingly, the top-end configuration I have is actually $130 cheaper than the 13900F-powered one Dell offers right now, and Dell confirmed that the older 13900F configuration is getting a price cut to reflect this. $130 isn't that much for a $3,500 PC, but it's nice that prices are dropping nonetheless.

There are also cheaper configurations of the Aurora R16, starting at $1,300, and while I didn't get any cheaper models in for review, they actually seem to be priced very well for the hardware, considering this is a prebuilt. Even the $1,300 model with the Core i7-13700F and RTX 4060 is competitive with off-the-shelf parts, though the more expensive models have somewhat worse bang for buck.

The Aurora R16 came with Dell's standard keyboard and mouse, a Wi-Fi antenna, and a regular 3-pin power plug for the PSU. I highly recommend avoiding the stock keyboard and mouse if you're gaming, but for work and productivity, they're actually pretty decent, and the keyboard (despite being a membrane model) was enjoyable to type on.

Case and design

Modern, compact, and a good fit for almost anyone

If you compare the Aurora R16 to the R15, you wouldn't even need to do so side by side to see the big differences. The R16 is practically a totally different PC, designed more practically and normally while still following Alienware's distinct style. The older R15 had a much heavier gaming aesthetic, but I found it to be excessive. The R16 is much more grounded, and although you can tell it's a gaming PC, it also wouldn't be too out of place in a home office.

From the outside, the R16's case is made primarily out of black plastic, which has a consistent matte texture and color. The side panel is made of acrylic with a black background, and Dell offers a completely solid version, as well as one that has some hexagonal holes cut out for airflow. The only bit of metal is on the back, where you'll find all the rear I/O ports, and it, too, is black like the rest of the case. Well, not entirely black, as the PSU isn't painted and is instead its default metallic gray (which probably saves Dell a penny or two per machine).

But perhaps the most unusual part of the R16's chassis is the front panel. Most front panels either take air in from the front or through the sides, and the R16 doesn't really fit entirely into either category. You can see from the above image what I mean: It's open from the side but so open that air intake isn't all that obstructed, which is common for cases that suck in air from the sides rather than directly through the front.

The RGB on the R16 is there but isn't over the top as there are only two devices that actually have RGB: the 120mm exhaust fan and the front panel, which has a strip of RGB lighting on the right side. These RGB lights are set up by default (thanks to the Alienware Command Center app) to change color based on the game you're playing.

The R16 measures 7.76 inches wide, 16.5 inches tall, and 18.05 inches deep, which is pretty similar to the size of midrange to higher-end mid-towers and perhaps even a touch smaller. This is far smaller than the massive R15, which has the same 25.2-liter internal volume as the R16, but the difference here is that Dell decided not to put a normally sized case into what is essentially an even bigger one. This, thankfully, means the R16 is a much more normal PC regarding size and weight than the older R15.

The last thing I want to touch on is ventilation, which on paper doesn't seem amazing (but we'll see later that it's actually quite good). There's just a single 120mm fan for intake at the front and three 120mm fans for exhaust, one at the rear and two at the top through the liquid cooler. The GPU can also serve as intake if you have the vented side panel. I strongly suspect thermal performance in the R16 won't be great if you get the solid panel instead of the vented one, though I couldn't test that.

Overall, I'm a fan of the Aurora R16's design. It doesn't really overdo it in any areas, it's normal enough while also being distinct, and is designed well on a technical level, all of which are fairly unlike the older Aurora R15. A reset in this area was definitely needed and achieved.

Ports

Good but not for this kind of PC

One area where the Aurora R16 is lacking is the port selection. Front I/O is pretty good, though, with a headset jack and four USB 3.2 ports, one being USB Type-C. At the rear, there are eight USB 3.2 ports (two being Type-C), a 2.5 gigabit Ethernet port, and lots of audio jacks.

To be clear, this is a pretty good amount of ports and should be enough for most people, but for comparison, my ASRock Z790 Taichi Lite has 12 USB ports at the back, two of which aren't just Type-C but also USB4/Thunderbolt 4. Front I/O, meanwhile, can be up to eight USB 2.0 and eight USB 3.2 ports, though most cases will only offer about as much as the R16, often less. I would expect a high-end motherboard in a $3,500 PC, and while USB ports don't kill the gaming experience, I would have preferred more of them instead of a ton of audio jacks.

Both the RTX 4080 and 4090 that Dell puts in the Aurora R16 have the standard three DisplayPorts and one HDMI port, which is ideal in my opinion. I don't know what the other GPUs Dell offers will come equipped with, but hopefully, it's also three DisplayPorts and one HDMI.

Customization

Depressingly locked down

Easily, the worst part about the Aurora R16 is how much you can't customize it. While this obviously isn't unprecedented for prebuilts (especially when it comes to Dell), more and more PC builders are using off-the-shelf parts. The Aurora R16 is about half standard and half proprietary, which I guess is better than 100% proprietary, but it's still not good enough for today's standards.

The most custom part of the R16 is the motherboard and the case, both of which don't use standard ATX mounting, and the motherboard even has a pan handle that extends to the front I/O instead of just using a daughter board like a normal PC. This means two things: You can't use this board outside of a compatible Dell PC case, and you can't put a different motherboard inside the R16 case. That would be fine if this were an amazing motherboard, but sadly, it isn't for a couple of reasons I'll get into later.

Upgrading the GPU in this PC is possible without using a Dell GPU, but it will have to be small enough to not interfere with the latch at the front that normally holds the GPU in place. Plus, while the GPU power plugs are standard, your choices are either a 500-watt or a 1000-watt PSU, the former being very underpowered and the latter being very overpowered. Why Dell doesn't offer a more normal 850-watt option isn't clear to me; it would work just fine with the 14900KF and 4090.

The CPU, RAM, storage, and CPU cooler are all completely standard thankfully, and you can upgrade them as normal. There are two M.2 slots for NVMe SSDs (only up to PCIe 4.0 though) and a SATA slot for a HDD or SATA SSD. There are only two RAM slots, which isn't great (and is one of the two reasons why this motherboard isn't very impressive), and while you can install non-Dell certified RAM, their XMP profiles won't work and they can only go up to 5,600MHz. However, you can at least manually set RAM up to 5,600MHz, which is better than nothing.

The upgrading process in general is pretty normal for a PC, except for the GPU, which can be removed without tools. The excellent cable routing also makes it easy to work inside the R16, which was a nice change from my test bench and my usual desktop. If you're a computer novice, the R16 should be much less intimidating than the typical PC, though you still need some know-how to upgrade and perform maintenance.

Performance

Fast, but it could be even faster

As I mentioned before, the Aurora R16 is very customizable, but the unit I reviewed was specced out like so:

Additionally, I received Dell's RTX 4080, which I did a little testing on, but unless stated otherwise, I used the 4090 in my benchmarks. I've tested various applications on the R16 to understand how the CPU, GPU, and SSDs perform. I'm not just looking at gaming performance here but the full package. But I didn't want to just see how well the R16 performs, I also wanted to know how well it performs compared to my test bench using the Core i9-14900K, the same RTX 4090 that came in the R16, and some hardware you could reasonably expect to be in a $3,500 PC that anyone could build.

I was particularly interested in how the 14900KF would perform, and to be clear, the 14900K is the same CPU but with integrated graphics. One of the first things that I noticed on the R16 was that it only has 12 power stages in the VRM, which is where the CPU gets power from; for comparison, my Z790 Taichi Lite has 24 power stages just for the CPU. Fewer power stages generally mean the CPU has less power to access, and less power means less performance.

Unless stated otherwise, these benchmarks were run on the R16 with the 4090, and I used the default software settings. This means the power profile was set to Performance, the Alienware Command Center overlay was running, and that VBS was enabled. While it's easy to disable VBS, it's not exactly common knowledge that disabling it results in better gaming performance in many titles, so I left it on for the R16 to be more realistic, and disabled it on the test bench to show what the maximum possible performance is.

For my test bench, I actually didn't really need to change anything since the Taichi Lite comes with all of the performance-boosting settings enabled by default, which is common for these sorts of upper-midrange and high-end boards. I switched the power plan to high performance to ensure the 14900K would perform the best it could, just like how the R16 used Dell's custom Performance power plan.

Both machines used the latest updates to all software I used, from Windows 11 to the games to the GPU drivers, as of Oct. 15. Anyways, with all the testing methodology stuff out of the way, here's the data I collected.

CPU benchmarks

For measuring CPU performance, I tested two rendering apps (Cinebench 2024 and Blender) and two games (Counter-Strike 2 and Forza Horizon 5). This isn't nearly the full suite of CPU benchmarks I test, but that's not necessary and would make this review much longer than it needs to be.

In Cinebench 2024, the multi-threaded performance between the Aurora R16 and my test bench was pretty high at nearly 13% in favor of the test bench. That's a fairly high performance difference, considering these are supposed to perform exactly the same, and I'll explain why this is the case when we get to the power consumption part of the review. Single-threaded performance, thankfully, was equal.

Blender Benchmark is also multi-threaded, and here, the test bench beat the R16 by almost 16%. This is not a great look for the R16 considering how expensive it is to get the 14900KF.

While my test bench is slightly ahead of the R16 here in Counter-Strike 2, it's probably just down to VBS, normal run-to-run variations, and maybe slightly lower CPU clock speed. I would count this one as a tie.

In Forza Horizon 5's in-game benchmark, both the test bench and R16 are pretty close in average framerates, but there's a significant gap in the 99th percentile framerate. The 99th percentile is a quick way to judge how consistent frame pacing is, which is what determines how smooth gameplay feels. The R16's 99th percentile framerate isn't crippling slow, but it indicates that the 14900KF is bottlenecking gaming performance in high-framerate gaming scenarios in addition to heavy multi-threaded apps.

The performance results here are disappointing. In multi-threaded programs you're looking at roughly 15% lower performance with Intel's top-end CPU, and in games where you're trying to get a high framerate, you might see noticeably worse performance, too. This just isn't what you'd want in a PC that costs $3,500 and boasts top-end hardware but not top-end performance.

As a side note, whenever I had the UI of the Alienware Control Center open, CPU performance declined even further by an additional ~5%. I contacted Dell about this, and a spokesperson confirmed that this is unexpected behavior. Thankfully, merely closing the AWCC window is sufficient to fix this issue, and it can run in the background without causing issues. The overlay (which pops up whenever you open a game) is also not an issue.

GPU benchmarks

For testing the R16's graphics performance, I tested only a synthetic benchmark and a game, but since I'm comparing the same GPU in different PCs, I don't think I need to actually test much else.

The performance difference between the test bench and the R16 is essentially 0 in 3DMark's Speed Way test, which is 100% dependent on the GPU. Even VBS didn't impact the results here.

Performance was just about equal in Cyberpunk 2077's in-game benchmark, with the test bench being slightly in the lead. I suspect this is just down to VBS, since the 4090 performed the same in Speed Way regardless of being in the R16 or the test bench. It's very unlikely to be a CPU bottleneck given how low the framerate is, too.

I also tested the 4080 in Speed Way and Cyberpunk 2077, and I just wanted to see if either GPU was performing weirdly. Thankfully, the results aren't strange, with the 4090 being roughly 30% faster than the 4080 in both benchmarks as expected. This means that Dell's custom GPUs aren't poorly designed or limited in some way (in fact, they perform very well) and also that the R16 doesn't exhibit strangely low performance in GPU-bound scenarios.

SSD benchmarks

Finally, we have our SSD benchmarks, CrystalDiskMark and IOMeter. The first showcases absolute maximum performance, while the second I use to test performance when the drive is filled up.

Scores are organized by read/write and are measured in MB/s.

It's a bit of a mixed bag here for the SK Hynix PC801 SSD that Dell uses for the Aurora R16. It actually wins against Corsair's MP600 Pro NH (one of the fastest PCIe 4.0 SSDs) in a single workload, but the rest of the time, the PC801 ranges from being slightly slower to completely in the dust. In particular, the random performance on the PC801 is really bad, worse than even Solidigm's dirt cheap P41 Plus. While SSD performance doesn't impact gaming (yet), it just doesn't really belong in this kind of PC, does it?

Scores are measured in MB/s.

I used IOMeter to hammer the PC801 SSD with write operations to see how well it could maintain the good sequential writing seen in CrystalDiskMark. However, the level at which the SSD is filled also impacts writing performance. SSD cache (the smaller but faster part of an SSD's storage) gets smaller as an SSD is filled, leading to that cache getting exhausted increasingly faster with less space. Additionally, the fastest way for SSDs to write data is to store them in empty blocks, but obviously, when an SSD is filled, there are fewer empty blocks to go around, further hurting performance.

The above chart and table show how the PC801 performs at 10%, 50%, and 90% full, and the results are decent enough. You can expect at minimum 2GB/s in large file transfers (like .zip files) even when the drive is nearly full, which is good to see and is about in line with the MP600 Pro NH.

Overall, the performance on the Aurora R16 is not as good as it should be. With a $3,500 budget, a Core i9-14900KF, and an RTX 4090, we should at least expect top-end gaming performance, but that's clearly not what you get here since the CPU is underperforming. Thankfully, CPU bottlenecking is usually only an issue in games where you're hitting more than 120 FPS, so if you prefer 4K gaming with the highest graphics settings possible, you probably won't run into CPU bottlenecking even with the RTX 4090.

As for productivity performance, the CPU and SSD are significant limitations. This is a gaming PC, but lots of people do both work and play on a single device, such as myself. That being said, the R16 isn't totally incapable; it's just worse than you'd expect for the money and hardware involved.

Noise, power, and thermals

A mixed bag

The final thing I want to discuss here is noise, power, and thermals, which are both important for the user experience in terms of comfort and also can have an impact on performance (for example, an overheating CPU tends to be one limited in performance). Dell also provides three custom power plans through the Alienware Control Center, which I tested individually for this section.

Performance when using different power plans

I want to start with testing the performance of the three power plans, Quiet, Balanced, and Performance (which is enabled by default). These plans differ when it comes to how fast they'll let the fans get and how much power they'll let the CPU consume, and more power and better cooling means more performance. I tested all power plans in Cinebench 2024 to see the difference.

In Cinebench 2024's multi-threaded test, the Performance plan is 8% faster than the Balanced plan, which is 4% faster than the Quiet plan. Performance in the single-threaded test was the same, and gaming performance didn't change in either Counter-Strike 2 or Forza Horizon 5. So, if you're worried about gaming performance changing if you switch off the Performance plan, you don't need to worry (probably).

Noise testing

I tested the Aurora R16 in a room with an ambient noise level of 40 decibels, and tested how loud it would get at idle and while running Cinebench 2024, 3DMark Speed Way, and both tests at the same time just to really stress out the R15.

The results here basically tell us that the Quiet and Balanced plans aren't going to let the fans spin much faster due to even a super heavy CPU workload like Cinebench 2024, while the Performance plan will respond to high CPU heat with increased fan speed. In gaming, all three plans had about the same noise level, probably because these plans don't impact GPU performance (each plan returned almost identical Speed Way scores). These power plans and the case fans respond exclusively to the CPU as far as I can tell.

Power testing

For testing power, not only am I looking at how the three power plans impact the power consumption of the 14900KF, but also how the 14900KF in the Aurora R16 compares to the 14900K in my test bench.

This chart here explains exactly why the R16 is so far behind the test bench in CPU-bound workloads: The 14900KF isn't allowed to consume nearly as much power. Even using the Performance plan, the R16 averages at 250 watts over the course of Cinebench 2024, while the 13900K in my test bench averaged 300 watts. Despite this, the R16 averaged at just about 80 C in this test, while 14900K in the test bench gladly hit 100 C to keep the 300-watt power consumption going.

The reason why Dell limited the 14900KF's power usage is that the motherboard the R16 uses has a mere 12-stage VRM, which is nowhere near enough to keep the 14900KF going at 300 watts. The best it can do is 250 watts on average it seems. Additionally, the R16 also has much less consistent power draw than my test bench, and although I don't know why this is the case on the R16, I suspect this might be a problem for games, which can be very sensitive to CPU clock speeds, especially if you're trying to hit a very high framerate like in the Counter-Strike 2 benchmark.

The difference between the three different power profiles can be seen in this chart. All three consumed 250 watts initially in Cinebench 2024, but the Quiet and Balanced plans dropped to 176 watts and 186 watts on average respectively for the rest of the benchmark. That level of power consumption is nearly what the Core i5-14600K consumes, though thankfully the performance on the 14900KF is still better since it has more cores and can use this amount of power far more efficiently.

Thermal testing

My CPU temperature test results were taken from the same Cinebench 2024 runs I did for the power testing. For the RTX 4090, I used the Speed Way stress test, which runs on loop 20 times.

On average, the Balanced plan was the coolest at 77 C, while both the Quiet and Performance plans averaged 82 C. Of course, the Quiet plan got hot because it didn't raise the fan speed, while the Performance plan did but boosted higher instead of just running cooler. Strangely, despite being cooler the Balanced plan apparently wasn't any louder, which probably indicates that either my decibel reader isn't quite high-end enough or that the ambient noise of 40 decibels is too loud to get good data.

GPU temperature testing across all power plans In Speed Way resulted in the 4090 leveling out at 74 C, despite consuming over 400 watts continuously. This is a great result, and it's good that Dell knows how to design a high-end graphics card. However, I think instead of using the 4090's great cooling performance to make it cool, it would be better to reduce the fans speed and let it run at a normal 85 C.

Should you buy the Alienware Aurora R16?

You should buy the Alienware Aurora R16 if:

• You want great gaming performance
• You want a unique-looking PC
• You want a PC that's good out of the box

You shouldn't buy the Alienware Aurora R16 if:

• You want the best gaming performance possible in high framerate games (like competitive shooters)
• You don't want to pay a premium for a prebuilt
• You want to be able to upgrade your PC and reuse its parts

I really liked the Aurora R16 in general as I was testing it out. It performed great, it wasn't too loud, it didn't overheat, and there were literally two preinstalled apps that could hardly be called bloatware. When I put it on the desk to take photos of it for this review, I actually liked having it there and wouldn't mind switching to it from the mini-ITX desktop I built myself, which is definitely much slower.

However, for $3,500, the R16 isn't really worth it. You're not even getting the true performance out of the Core i9-14900KF due to the inadequate motherboard with its measly 12-stage VRM, and the SSD isn't amazing either. A similarly speced PC built using your own two hands would cost less and provide substantially more CPU and SSD performance, not to mention actually upgradeable parts that work universally. The top-end configuration of the R16 is definitely functional, but I don't see why you would spend top dollar to end up with second-place performance.

While the top-end model of the Aurora R16 isn't great and is instead just good, I do really like the base $1,300 configuration from the look of the spec sheet. It's priced extremely competitively, and you can upgrade pretty much every important component down the line, though I can't recommend it since I didn't test it. As for the model I did test, maybe opt for the Core i7-13700F or a different, lower-end CPU since you're not going to get 14900KF performance out of the Aurora R16, and you might as well save a little money.