Crucial brand RAM is good

Crucial Ballistix 3200 MHz CL16 test - Can Micron’s house brand keep up with the B-Die Top-Dogs?

A good 5 years ago, the fourth revision of the double data rate main memory became the standard for the masses of PCs, and with the first information about the market launch of DDR5 next year, the DDR4 standard will soon reach the end of its life cycle. The manufacturing techniques and products of the DDR4 manufacturers are correspondingly sophisticated and right now it is worth paying particular attention to the price-performance ratio when buying in order not to miss one or the other insider tip and not to be dazzled by high clock rates and bright LEDs .

As a manufacturer of memory chips, Crucial has long been a well-known name that stands for a good compromise between price and performance. Without anticipating too much, the product of today's review is also precisely geared towards this compromise. Crucial Ballistix DDR4 kits are available in various designs, with XMP profiles from 2400 to 3600 MHz, in colors black, white and red and with and without RGB. So there should really be something for every taste and budget. Over 3600 up to 5100 MHz, the product portfolio is then expanded by the Ballistix Max series, which is based on the same basic properties and builds on them with small improvements. The kit of today's review with a clock rate of 3200 MHz and a price of 75 euros is located in the middle of Crucial's DDR4 portfolio and should give us a good insight into the Ballistix series.

It is also important to mention that Crucial is a brand of the memory manufacturer Micron. Even if Crucial, like many brands in the memory and SSD market, does not manufacture the memory chips itself, but only buys them and then uses them on their own products, only memory chips from the manufacturer Micron are used in the Ballistix kits. Particularly when looking at the performance potential beyond the XMP profile, the manufacturer and type of memory chips are particularly important. So if you don't want to dig your way through various product variants from other manufacturers in search of a certain type of memory chip, Crucial can be sure that you will get Micron memory chips. Which type of Micron memory chip exactly and which properties it has will be highlighted in the course of the review.

In terms of performance, the Micron memory chips generally rank behind Samsung’s B-Die and are characterized by a relatively high clock frequency with a relatively low load on the memory controller of the CPU. However, there are shortcomings in the relatively high tRCD, tRP and high tRFC compared to tCL, all of which have a correspondingly negative impact on performance. However, the performance is not nearly as much lower than the price compared to Samsung’s B-Die. Thus, in theory, Micron’s memory products represent an interesting compromise between price and performance, which we will test in practice in the following.

I was able to get hold of this special kit in the Amazon Prime Day Sale for a mere 46 euros. So it was a "no-brainer" for me to get the kit and put it through its paces. The black 2x 8GB kit in today's review listens to the part number BL2K8G32C16U4B and has the following specifications according to the Crucial homepage https://www.crucial.de/memory/ddr4/bl2k8g32c16u4b:

The XMP specs of 3200 MHz clock, 16-18-18-36 timings and 1.35V voltage clearly indicate a product in the DDR4 midfield and are appropriate to the retail price. If we read out the SPD with the Typhoon Burner program, i.e. the memory built into the bars in which the XMP information is stored, we get a lot more information:

In the “PART NUMBER” field we find both the part number of the kit “BL2K8G32C16U4B” and the suffix “M8FE1”, which designates the type of Ram bar.

Under "DIE DENSITY / COUNT" we find the first information on the type of memory chip and thus also on the rank layout of the bars. The size of the memory chips is 8Gb (Gigabit), from which it can be concluded that 8 chips have to be installed in order to achieve the capacity of 8GB (Gigabyte) of a bar. Furthermore, the type of chip is specified here as "E-die". Typhoon Burner's formulation is a bit unfortunate, as Micron does not name the chips that way, but that is the first indication that these are Micron RevE memory chips.

At "CAPACITY" we find the confirmation of acceptance from above, namely that the capacity is made up of the following number of chips: "8GB (8 components)". Further interesting information is "MANUFACTURING DATE" and "MANUFACTURING LOCATION", that is the date and place of manufacture: "March 16-20 / Week 12, 2020", "Boise, USA", that is, March 16-20 in the 12th week of the Year 2020 in Boise, USA.

Last but not least, we find the already known XMP specs under XMP-CERTIFIED: "1600 MHz / 16-18-18-36-72 / 1.35V". The obligatory question of every RAM kit review arises, why 1600 MHz and not 3200? For the sake of completeness, it should be explained that with DDR, i.e. double data rate, memories are executed two ticks with each clock (unit Hz). So the kit actually runs at 1600 MHz and, strictly speaking, the name 3200 MHz for the kit is incorrect. But it has just established itself in the industry, similar to gigabytes and gibibytes for hard drive capacities, but that's a discussion for another time, IT specialists will be able to tell a song about it. 😉 But back to our main memory, in addition to the 4 primary timings tCAS, tRCD, tRP and tRAS, the secondary timing tRC is also listed.

Further down in the table we find even more secondary timings, tFAW, tRRDS, tRRDR, tWR and tWTRs, both for the XMP specification and for the JDEC specifications, which every DDR4 kit must bring. The application of the secondary timings from the XMP profile differs from platform to platform and sometimes even from one motherboard manufacturer to another. For example, with Intel systems, tRC is not displayed in the BIOS at all, but only managed automatically by the memory controller, while the setting is accessible with AMD, and some motherboard manufacturers, such as Asrock and Gigabyte, take over the other secondary timings from the XMP profile, while Asus or MSI ignore these and put their own. To cut a long story short, the secondary timings from the table are effectively irrelevant to the user.

It should also be mentioned that the modules do not have a temperature sensor. This shouldn't be of any importance to the average consumer, overclockers may need to know how to use an external temperature sensor.

Appearance and feel

The simple design of the heat spreader combined with the dark gray-black coloring appears neutral and at the same time high-quality on the first impression. The various know-how and indentations of industrial design catch enough light to emphasize the texture of the surface and not appear boring. However, extravagant design elements are consistently avoided.

A simple, light gray "BALLISTIX" lettering adorns both sides of the heatspreader in the middle.

Matching the color of the heat spreader, there is a black sticker on one side of the module on its lower edge with the most important data such as product number and XMP specification in white letters.

On the other side, a simple “crucial by micron” logo can be seen in the same place.

If we peek from below between the PCB and the heat spreader, we can see the 8 memory chips on one side of the board, to which the heat spreader is attached with an adhesive thermal pad. Due to the size of the memory chips and the relatively large gaps between them, the “RevE” chip type from Micron can again be inferred. On the other side of the there are no memory chips, the circuit board is completely bare there. This is to be expected because - as already explained above - the 8GB capacity of a bar is composed of 8 8Gbit memory chips and a single rank layout follows from this.

The heat spreader is only attached directly to the PCB with double-sided adhesive tape. However, attention was paid to the height of the adhesive tape so that the heatspreader is the same distance from the circuit board on both sides and the symmetrical appearance is retained even at second glance. Should it be of interest to someone, the Crucial logo on the heatspreader is on the board side with the memory chips and the sticker on the heatspreader is on the bare side of the board.

The simple, industrial design of the heatspreader is also continued on the top of the heatspreader, which should get the most attention when it is installed, paired with a third, smaller "BALLISTIX" logo. Minimalist branding is capitalized here, literally.

When installed next to each other, the upper sides of the two modules form a solid visible surface. Of course, it should be noted that on most mainboards with 4 DIMM slots there is still some space in between. But the sides of the heatspreader do not have to hide under any circumstances and should contribute to a harmonious overall picture with their light-catching depressions even in a 2-bar setup.

The feel is generally surprisingly solid and of high quality. You don't get the impression that it is a PCB with two pieces of sheet metal glued on, but it feels as if the PCB and heatspreader have merged into one solid module. Not least for the heat absorption and removal from the memory chips via the heat spreader, this already looks very promising. The anodized aluminum of the heat spreader is pleasant to the touch and even after several installations and removals you don't see any fingerprints like on the high-gloss surfaces of some other RAM modules.

The design is almost perfect for friends of minimalism or RGB heretics and should fit seamlessly on every primarily black motherboard. And if you really can't do without the colorful play of light, you can also get an RGB version for a few euros extra.

Test system

The kit was tested together with an Intel Core i9 10900KF as CPU, overclocked to 5.4 GHz core and 5.2 GHz cache, and an Asus Maximus XII Apex as motherboard. Basically, it should be noted that the architecture of the Intel platform used here has a relatively large advantage in terms of latencies compared to equivalent AMD platforms, and the high cache OC of the CPU reinforces this again. The benchmark results should therefore only be compared within the same platform.

This is what the kit looks like in the system. Please ignore the RAM water block next door, as I did not want to expand my daily set from the loop for the review. The Crucial Ballistix Kit fits very well into the overall picture of the system. The industrial design and the black anodized aluminum go well with the heat spreaders of the Maximus XII Apex and without RGB lighting, the overall structure looks simple but high-quality.

Here is an overview of the entire specs of the test system:

 

The AIDA64 Memory and Cache Benchmark, which provides synthetic measured values ​​for data throughput and latency, was used as a benchmark tool. On the other hand, Geekbench 3 was used to compare the memory score for the overall performance of the main memory and thus to have a reference for “real-world” performance. Testmem5 v0.12 with the profile "Ollie" was used to test the stability, as this tool also generates a high thermal load and thus the stability can be checked after heating.

All performance levels were tested for stability using TM5 Ollie and passed the test without errors. Of course, several and longer tests should always be used for 100% validation of a memory OC, but this would have exceeded the time frame of the review. A flawless TM5 Ollie loop is still a very solid indicator of stability.