Introduction, Specifications, and Pricing
Along with the set of DDR4 we looked at recently, Kingston was sure to pass along some of their DDR5 for testing. While they look similar outside, what is under the hood is much different. We get things like the included PMIC for better signaling and voltage regulation and new ICs that pack in more speed than what was available in DDR4 kits. Not to mention that DDR5 comes with on-die ECC and that all DDR5 is read as quad-channel, even with only a pair of sticks.
Kingston's Fury Beast DDR5 starts at 4800 MT/s just like all the rest, and as for the feature set, Kingston lists a few things worth mentioning. First comes a notation for improved stability for overclocking, which is nice for those who like to surpass what the XMP 3.0 profile offers. They also say there is increased efficiency, that these kits are indeed XMP 3.0 certified, that the world's leading motherboard manufacturers qualify them, and that they come with a low-profile heat spreader design so as not to conflict with various coolers.
With this being our first look at DDR5 from Kingston, we have no idea what to expect. However, with the rated speed and timings of this set of RAM, we do have some idea of how it should perform on our test system. Even so, with DDR5 being new and things like locked and unlocked PMICs, we do have some tinkering to see just what Kingston offers and how well it stacks up to a couple of other kits we have tested thus far. Stick around and see just how well the DDR5-5200 of Kingston Fury Beast set stands up to the competition and find out if this is the right kit for your next build.
There isn't much information in the specifications chart we took from the product page, but all the vital statistics are presented. Initially, the chart covers the density of the sticks and kits, where we are shown that you can buy 8GB, 16GB, and 32GB individual sticks. For those looking for a kit, you can get 16GB, 32GB, and 64GB sets, of which we have a 32GB version. Next comes speed, and we all know that DDR5 starts at 4800 MT/s, but Kingston also offers 5200 MT/s which we have, 5600 MT/s, and even 6000 MT/s. At 4800 MT/s, the sticks come as CAS38, but for the rest, you will see them at CAS40. Voltages vary by speed as well.
To run 4200, they need just 1.1V, whereas the 5200 and 5600 MT/s kits require 1.25V, and if you opt for the 6000 MT/s option, the XMP 3.0 profile will set the VDIMM to 1.35V. Operationally, you can run the Fury Beast within a temperature range from 0° to 85°C, and while not shown in the chart, these are backed with a limited lifetime warranty. The last thing shown is the size of the Fury Beast, where we can see they stand just 34.9mm high.
More specifically, in the set of KF552C40BBK-32, the part number spells out quite a bit. First, these are Kingston Fury DDR5. The 52 alludes to the 5200 MT/s speed, whereas the C40 refers to the timings. BBK is for the color black, and the -32 describes the density. Visually, the Fury Beast has a familiar aesthetic while also being an all-new design. The aluminum heat spreaders are anodized, but the FURY found on them is made of exposed aluminum for a bit of pop against all the black.
Along with that, the Kingston name, the Beast series, and the notation of DDR5 are all expressed with white painted letters and numbers. As to the design, there are embossed shapes on the sides, as well as what appear to be ventilation holes near the top, along with some notches at the top to harken back to older kits we have seen in the past.
We also know that early adopters of DDR5 will have to pay a hefty tax for the pleasure, but unlike the last set we saw at nearly $400, Kingston is keeping the cost down. So much so that they are the most affordable kit in their class. Searching through what is currently available as a 32GB kit of 5200 MT/s DDR5 on both Amazon and Newegg, these Fury Beast kits are shown at the lowest price. On Amazon, the list price is $193.51, with Kingston listed as the seller, but Newegg is slightly more affordable, with a list price of $189.07, but does come from a third-party seller. With that in mind, if the performance does not fall on its face, Kingston is on the right path to being in more systems than the rest in this class.
Packaging and Kingston FURY Beast
Packaging for the Kingston Fury Beast is simplistic but no less effective. Clear plastic for the clamshell packaging allows buyers to see what they are buying while protecting the RAM from damage and static. Sealing the packaging is the black and white sticker with the name of the RAM, the site address on the front, and the density, speed, timings, and voltage shown in the white portion. The part number is displayed on the left as the sticker wraps around to the back, and we also see an Intel XMP-certified sticker.
Flipping the packaging on its face, we can see how the branding sticker seals the packaging, and it also offers a view of the product stickers on the RAM itself. These sticks come with a small insert that covers the warranty and installation for those who may be unfamiliar with how they work.
An unimpeded view of these sticks shows Kingston's angles, shapes, and designs on the black heat spreaders. To the left, we see the Kingston name, painted just above the exposed aluminum FURY and etched in "TM." To the right, we see the Beast series name, while at the top, they ensure you know that these are DDR5 sticks.
The reverse of both sticks offers the same design elements but lacks any naming we found on the front. What you will find is the product sticker near the right end. On the sticker is the part number, the serial number, the voltage required to run it at the XMP 3.0 profile, that this is one of two sticks in the kit, and that you will not get warranty support if this is removed.
As we view the top edge of the kit, we can see a two-part heat spreader design, where tabs lock them together at either end. While small, Kingston does paint the FURY name on one half of the spreader, and even though we get why it is done this way, we would like it to have been larger and easier to read from a distance. With the lack of RGB in this set, we would like more to pop here than just the tiny FURY name.
In the previous view of this side of the sticks, some of the shapes and designs were tough to make out, but the lighting and shadows in this angle make them much more apparent. We can also see the lines milled into the exposed FURY portion, which was too bright in the previous images. Lastly, the ventilation near the top is open on both sides and can take advantage of chassis or CPU cooler airflow to help keep them cool while in use.
While in the BIOS, we did see that the Kingston Fury Beast showed up as Micron/SpekTek, but it wasn't until we removed one of the heat spreaders that we were sure that they were indeed Micron and not Micron rejects.
When it came time to look at the PMIC, both were blurry and tough to read, but we believe the model number to be AFW85020AX704. Attempting to look it up leads us nowhere via the internet, but we assume at this time that this is a locked PMIC, as voltage changes while trying to overclock lead us to a BSOD with changes to VDD or VDDQ.
After installing the Kingston Fury Beast into the test rig, we find them to blend well with the build and not stand out against the sea of black. At certain angles, you get a glint from the FURY name, and when the system is powered, the lighting reflects off of this area and shines whatever color is in the rig back into your eyes.
Test System Details
To obtain the AMD CPU-Z screenshots, you will see directly following this image, and this is the system we used to do it, as well as in attaining the results seen in the following pages. Thanks go out to be quiet, ASUS, and NVIDIA, for supporting this venture. Detailed specifications of the system can be found below.
After clearing CMOS before powering the system, we jumped into the BIOS to enable XMP, allowing the Fury Beast to run at their rated speed and timings. Once we opened CPU-Z, we saw that they run at 5200 MHz with 40-40-40-80 2T timings. Voltages were 1.25V for the VDD and VDDQ, while the IMC voltage was set at 1.19V.
While we set the default XMP profile, we noticed there was a second option, which adds a bit more compatibility. You have the option to run the Fury Beast at 4800 MHz, but this time the timings are slightly reduced to 38-38-38-70 2T while using the identical voltages to the main XMP profile.
While we could not add any speed to the kit, we could tinker with the timings a bit. While opting to move back to the default 5200 MHz, we got the timings down to 36-38-38-80 2T, with the voltages listed above, as any changes to either the VDD or the VDDQ would render the system inoperable with a BSOD every time we changed them. We were able to address the IMC voltage, but it did not affect what we could attain.
Chad's AMD DDR5 Dual-Channel Test System Specifications
- Motherboard: ASUS Maximus Z690 APEX - Buy from Amazon
- CPU: Intel Core i9 12900K - Buy from Amazon
- Cooler: be quiet! PURE LOOP 360mm - Buy from Amazon
- Video Card: NVIDIA GeForce RTX 3070 Founders Edition - Buy from Amazon
- Storage: Phison B47R Fortis 1600 2TB M.2
- Case: Thermaltake Core P3 TG - Buy from Amazon
- Power Supply: be quiet! DARK POWER PRO 12 1200W - Buy from Amazon
- OS: Microsoft Windows 11 Home - Buy from Amazon
Just as a heads up, we added the best runs from our old DDR4 charts. We did so as a temporary measure which allows one to reference our old setup results compared to what this newer CPU architecture and DDR5 offer over what you are used to seeing from us. As the charts populate with more DDR5 examples, the "Best of Old Charts - DDR4" results will be removed.
Using CPU-Z to get an angle on how the memory plays into single-core CPU performance testing, we find that the Fury Beast does rather well. The XMP1 results are very similar to the much faster XPG Lancer RGB, and we can see that opting for XMP2 did not lose much performance compared to XMP1. However, the best results were obtained by lowering the timings a touch more, which takes the Fury Beast past the 6000 MHz XPG results.
Read bandwidth in AIDA64 is a slightly different story. While the Kingston Fury RGB performed as we expected for each of the three options, we would have assumed it to keep up with the non-descript 5200 C38 run. Even so, we can see that XMP1 is tuned quite well with the 79340 MB/s result. XMP2 took quite a hit with a loss of 5919 MB/s compared to XMP1, and with a bit of tuning, we were able to gain 1361 MB/s over XMP1 with very little effort.
The results of the write performance test are a bit more in line with what we would have expected to see. XMP1 lands us at 73480 MB/s, just behind the 5200 C38 run. The XMP2 results were slightly less than we expected, with a drop in performance of 4808 MB/s compared to the XMP1 results. We are pleased to see that at 5200 MHz using 36-38-38-80 2T timings, we surpassed the 5200 C38 results and gained another 754 MB/s over XMP1.
With AIDA64, copy performance is said to be the metric most similar to real-life experience, and in that vein, we can see the Fury Beat did reasonably well. Starting with the XMP1 run, we landed at 71340 MB/s, which isn't horrible. However, opting for XMP2 takes a huge hit in performance, to the tune of a loss of 3845 MB/s. To9 brighten things up, we then look to the reduced timings run, where we were able to gain another 1062 MB/s, but were still surpassed by the 5200 C38 kit results in the chart.
When it came to latency reported in AIDA64, you can see that Kingston did not tune these kits as well as some of the others but was still able to hang close in the previous results. While our reduced timings settings netted the best result, XMP1 isn't a slouch either., However, opting for XMP2 lands these at the bottom of the chart.
Super Pi shows the Fury Beast in a much better light, with the XMP1 results topping the chart with the shortest time to completion. Even the XMP2 results are impressive, considering they are less than five seconds slower than the XPG Lancer RGB. Tinkering with the kit is something that Super Pi does not like, as it falls to the lowest of the three results.
The physics score portion of 3DMArk Fire Strike scales as one would expect. The XMP2 results were the lowest of the three but still in line with the speed and timings of all the kits listed. XMP1 did better than expected, and while there wasn't a huge gain, lowering the timings did deliver the best results for this RAM.
PCMark 10 may not be the end-all for testing, but it does offer a wide variety of testing types to show how well RAM is suited for various tasks. To say we are a bit shocked at these results is a bit of an understatement. Both XMP1 and XMP2 results top all the other kits in the chart, and if you like to tinker, you have gained some points.
File compression is something many users will be doing with their systems, and using 7-Zip and a near 8GB file, we can see just how well the Fury Beast does here. We fully expected the XPG Lancer RGB to tp the chart versus the XMP1 and XMP2 results, but the gap is smaller than we assumed. We also never thought that reducing the timings would sail to the top of the chart, but they have.
Cinebench R23 is something we see many using to test one CPU to another and how well the RAM is tuned. Again, we see that Kingston tuned this kit well for such a test, with the XMP1 and XMP2 results landing at the top of the chart. Tinkering with the kit did lose us some performance, but only slightly, and may not be the best option for those looking for the best in Cinebench.
For those who transcode video files, we look to Handbrake to sort out which is the best in this type of environment. To our utter shock, we find that the Kingston Fury Beast lands at the top of our chart once again. While the difference in the three runs shows a three-second difference from best to XMP2, there is a wider gap in time to the next best kit, which is shockingly the XPG Lancer RGB.
Regarding initial impressions of the Kingston Fury Beast DDR5, we will say there is a familiarity with many of the past kits we have seen. The styling, the color, and the exposed aluminum FURY on the sides remind us of other Fury kits, even down to the white painted names also found on the heat spreaders. This early in the DDR5 game, we do not expect much room to tinker, but even so, since Kingston did not bin these within an inch of stability, we did have a bit of room to lower the timings and have a little more fun than just running the XMP options. On top of that, even without the speed or timings of some of the other entries, the Fury Beast was able to hold its own and even top a few of the charts, which says a lot about what you are getting for the money involved.
We never saw the temperature go beyond 46°C in all our tests on an open-air system without fans pointing at the RAM. For many, who will be using these kits inside a chassis, thermals can be considerably lower. All around, we do feel that there is a nice mix of performance and styling, and even with the lack of RGB to accompany this version of the Fury Beast DDR5, we didn't miss it. With this much performance, enough to take on much faster kits in the charts, and the styling that is designed into the heat spreaders, we feel that Kingston is onto something with the Fury Beast and is a kit that offers quite a bit of bang for the buck.
Considering in many instances that the Fury Beast could run near or surpass the nearly $400 set of XPG Lancer RGB sticks at half the cost says just about all you need to know. Even if RGB is a must, the premium is not worth it in our minds.
Even with soaring prices around the globe, the Kingston Fury Beast is the most affordable of all of the 5200 MHz CAS40 32GB kits says a lot and is a major factor in the mass acceptance of such a product. While $193.51 seems a bit steep, many DDR4 32GB kits will cost you nearly the same. The fact that you can enjoy the benefits of DDR5 at a similar cost says a ton. It is a lot of the reason we will be recommending Kingston to anyone in the market for affordable, well-performing, yet stylish DDR5 for their new Z690 build or for future AM5 thoughts.
The Bottom Line
Depending on the test, you may find better performance elsewhere, but you will have to pay for that advantage. All things considered, we feel that you cannot go wrong opting for Kingston's Fury Beast DDR5.