The launch of the Intel 12th generation microprocessors comes with the accompanied introduction of the DDR5 memory in the market. This new memory interface standard aims to improve power consumption, while doubling the bandwidth compared to the outgoing generation. Like the ones before it, DDR5 is not your typical interface update just to keep the numbers going.
Now what we actually have here is of course no other than the Kingston FURY Beast DDR5 in a dual channel 32GB configuration. Apart from that mouthful, it features a 5200 MHz frequency and a CL40 CAS latency. For reference, our review unit comes with the KF552C40BBK2-32 part number. KF stands for Kingston FURY, 552 stands for DDR5 5200, C40 is CL40 CAS latency, BBK2 is most likely Black in kit of 2 and 32 is the memory capacity. CMIIW.
Disclosure: Kingston sent this as a review sample for the purpose of this review. The company did not ask me to say anything particular about it.
- Product Page: Kingston FURY™ Beast DDR5 Memory 32GB/5200MHz
- Price: ₱ 14,999 PHP (MSRP)
- Release Date: Q4 2021
|Capacity||32GB (2x 16GB)|
Packaging and Accessories
The FURY Beast comes in a simple blister pack.
The product comes with the following items inside:
- 2x Kingston FURY Beast DDR5-5200 16GB
- Warranty guide
- FURY case badge
Design, Layout and Connectivity
The FURY Beast DDR5-5200 is a much simpler looking DRAM over its predecessors – despite the many edges and perforations it comes with. At a glance, it’s no different compared to the outgoing DDR4 models but considering that DDR5 modules comes with their own 12V power management ICs, I’ll take this as a design advantage over the previous generation.
Height is pretty standard at around 35mm even with the added arch for the stamped metal heatsinks to attach to each other. Width is about 6.6mm which is most likely due to the sheet metal used. With decent compatibility in mind when it comes to tower coolers, I still suggest checking out the specification sheet for clearance.
Each FURY Beast DDR5-5200 stick features a single rank module. Kingston, like the proverbial manufacturers that they are, of course made their own chips for this one. Sticks defaults to the JEDEC standard DDR5-4800 with timing of 40-39-39 at 1.1V.
Test Setup and Methodology
Our test setup relies on the measurements taken from industry standard benchmark tools and real-world applications. It is important to note that we are testing the review sample after burn-in, with at least 24-hours of uptime. This is done so to negate the FOTB (fresh out the box) state of the DUT (device under test), yielding better benchmarking consistency.
|Test System Specifications|
|CPU||Intel Core-i9 12900K|
|Motherboard||ASUS ROG Strix Z690-I Gaming Wi-Fi|
|Cooler||ASUS ROG Strix LC II 360|
|Memory||Kingston FURY Beast DDR5 5200MHz 32GB|
|GPU||GALAX RTX 2060 EX White 6GB|
|Storage||Kingston FURY Renegade 2TB|
|Case||Mechanical Library JXK-K3|
|PSU||Thermaltake Toughpower GF1 650W|
|OS||Microsoft Windows 10 Pro 64-bit|
The DUT is tested with the following configuration from our test system:
- UEFI configuration: Default
- Windows Power Plan: Balanced
Note: We’ve tested the Kignston FURY Beast DDR5-5200 at its rated JEDEC specification (4800) and XMP 3.0 profile #1 (5200). XMP 3.0 profile #2 is supposed to be a tighter JEDEC frequency based profile but I cannot get it working properly on this particular test system. Additionally, we do not have comparisons yet since this is the first DDR5 module we got.
Throughput and Latency
Memory throughput is measured in Megabytes per Second (MB/s). This is done via AIDA64 Extreme and its Cache and Memory Benchmark. The test aims to check the combined read and write performance of the memory.
Memory latency is measured in Nanoseconds (ns). This is done via AIDA64 Extreme and its Cache and Memory Benchmark.
Throughput is excellent and so does the latency. JEDEC profile is definitely slower here on the synthetics.
Pi calculation speed is measured in Seconds (s). This is done via SuperPI. The test aims to check how fast a system calculates Pi.
Roots calculation speed is measured in Seconds (s). This is done via WPrime. The test aims to check how fast a system calculates square roots.
DDR5-5200 is actually slower here at the arithmetic compared to the DDR5-4800 profile – at least on the single-threaded SuperPI benchmark.
Cinema 4D score is measured in Points (pts). This is done via Cinebench R20. The test aims to check the Cinema 4D performance of the system.
V-Ray 5 score is measured in V-Ray samples (vsamples). This is done via V-Ray 5 Benchmark. The test aims to check the V-Ray 5 performance of the system.
Rendering scores are too close to each other to tell any clear winner. Technically though, it goes to the DDR5-5200.
Digital Content Creation
Image editing speed is measured in Seconds (s). This is done via RealBench and its built-in GIMP benchmark.
Video encoding speed is measured in Seconds (s).This is done via RealBench and its built-in HandBrake benchmark.
Same story could be said at the content creation benchmarks.
Web browsing speed is measured in Seconds (s). This is done via PCMark 10 and its built-in Chromium benchmark.
Web browsing speed is measured in Seconds (s). This is done via PCMark 10 and its built-in Firefox benchmark.
The web browsing benchmark results are also neck to neck.
Productivity speed is measured in Seconds (s). This is done via PCMark 10 and its built-in LibreOffice Writer benchmark. The test aims to check the speed of loading documents.
Productivity speed is measured in Seconds (s). This is done via PCMark 10 and its built-in LibreOffice Calc benchmark. The test aims to check the speed of copying data and compute.
Pretty much within the margin of error here.
Compression speed is measured in Kilobytes per Second (KB/s). This is done via WinRAR and its built-in benchmark.
Instruction speed is measured in Giga-Instructions per Second (GIPS). This is done via 7-Zip and its built-in benchmark.
Compression benchmark suggests that DDR5-5200 is the way to go – as it should be.
Frame rate is measured in Frames per Second (FPS). This is done via Final Fantasy XVI: Endwalker and its official benchmark. The test aims to check the 1% low FPS performance.
Frame rate is measured in Frames per Second (FPS). This is done via Sid Meier’s Civilization VI and its built-in benchmark. The test aims to check the 1% low FPS performance.
Gaming also saw a boost at Civilization VI – as far as the 1% percentile FPS benchmark is concerned. I would really love to use another CPU and memory benchmark here (Total War: Three Kingdoms) but the 12th generation Intel processors are having a fit with the game’s DRM.
Software, Lighting and Special Features
The Kingston FURY Beast doesn’t come with RGB lighting and other gimmicks. It comes with two XMP 3.0 profiles though on top of the JEDEC specified speed.
The FURY Beast is our first DDR5 memory kit. It comes with a throughput of 5200 MT/s at 32GB of capacity which is a combination that is more than enough for the tasks tested. This is a fast and high capacity kit with no major nitpicks aside from the price (actually decent) and availability. Once the global suppliers sorted this out, then everything will definitely fall into a much better place.
Now since we have no comparisons yet, I’ve eye-balled the performance difference compared to my AMD Ryzen based DDR4 test platform and this absolutely ate it for breakfast. I’ve expected it to happen though (Ryzen 5 3600 vs i9-12900K) so my impression could change once I’ve tested DDR4 vs DDR5 with the same CPU – in the near future.
For now, this particular kit gets my absolute approval.
Kingston FURY Beast DDR5-5200 Memory Kit
The Kingston FURY Beast DDR5-5200 is a fast memory kit with tons of capacity in a stick of two.