The Ryzen 9 3900X is AMD’s newest CPU, and it’s a beast. It has 12 cores and 24 threads, with a 3.8GHz base clock speed and a 4.6GHz boost clock speed. If you’re planning on building your own PC or upgrading your current system, make sure to check out this article for the best RAM for the job!
The amd ryzen ram compatibility list 2020 will help you find the best RAM for your AMD Ryzen 9 3900X CPU. 3 of these RAMs are recommended.
Choosing the Best DDR4 RAM for the AMD Ryzen 9 3900X Processor
On the issue of memory compatibility and RAM overclocking, Ryzen did not create a favorable first impression when it was initially released in March 2017. Many memory kits could only run at modest clock rates, and the BIOS had limited choices for getting Ryzen to operate with fast memory. Ryzen now works considerably better with many DDR4 RAM kits since AGESA release 188.8.131.52, and users have access to a far wider variety of subtimings and higher RAM multipliers up to DDR4-4000, thus Ryzen users experiencing RAM issues should search for updated BIOS versions.
We tested a Ryzen system in 2017 to see which settings work best for various RAM kits with Micron, Samsung, and SK Hynix processors, as well as how Ryzen runs when fully loaded and with 16-gigabyte modules. With the release of new 2018 hardware (Ryzen 2700X and X470), we revisited the subject of RAM tweaking in issue 09/2018: We demonstrate how successfully two and four sticks with various memory chips may be overclocked, which choices might assist, and what latency adjustment, including subtimings, can provide. You’ll also find tips for turbo-optimizing Ryzen CPUs in the PCGH-Plus page.
Clock speeds above DDR4-3600 may be reached quite easily with Ryzen 3000 (such the Ryzen 9 3900X), however this is automatically converted to asynchronous mode, in which RAM (controller) and Infinitiy Fabric are no longer run in a 1:1 ratio. In reality, this results in performance losses that are difficult to compensate for with RAM tweaking, as we’ll demonstrate in this PCGH-Plus article. When utilizing larger RAM multipliers, it’s recommended to stick to DDR4-3600 or make sure that synchronous operation is enabled.
In this test, we looked at three RAM modules that work well with AMD Ryzen 9 3900X processors.
Best DDR4 RAM for AMD Ryzen 9 3900X CPU Results
Corsair Vengeance RGB Pro comes in first place.
2 Dual-Rank-DIMMs, 64GB DDR4-3600 RAM
- RAM that is very quick
- 64GB of storage is extremely future-proof.
- RGB was used in this project.
- Exceptional performance
Meanwhile, the first desktop kits with 32 GB per module are on the market. The spearhead of them are the hardly accessible machines with a DDR4-3600 release. The Corsair kit, which is approved for 18-22-22-42 at 1.35 volts and includes one of the most luminous RGB designs on the market, can be pre-ordered for under $400 and offers a lot in return: in addition to a whopping 64 GiByte, the Vengeance RGB Pro Duo, which is approved for 18-22-22-42 at 1.35 volts, includes one of the most luminous RGB designs on the market. In daily usage, the dual-rank architecture also provides better performance than single-rank modules, which must be clocked higher or placed in double versions per memory channel. Furthermore, Corsair modules feature significant reserves, which may be used for more strict timing or overclocking to DDR4-3866 at 1.35 Volts. However, don’t anticipate a tuning monster; the kit can’t handle precise timing. And don’t forget: 64 GB of RAM is a little excessive only for gaming, but there’s nothing you can’t accomplish with it.
Conclusion: The best RAM for Ryzen 9 3900X.
Corsair Dominator Platinum RGB comes in second.
2 Dual-Rank-DIMMs, 32GB DDR4-4000 RAM
- RAM at a speed of 3200MHz
- Any usecase will benefit from 32GB.
- RGB lighting is a kind of lighting that uses three
- The best value-for-money
The two 16-GiByte bars for DDR4-4000 are available at timings of 19-23-23-45 and 1.35 Volt, which, along with the dual-rank architecture, allows for excellent performance in all benchmarks. It’s no surprise that Samsung’s K4A8G085WB-BCPB chips, often known as B-Die, are still the best option for high-end RAM with excellent tuning characteristics, with 10% undervolting reserves and strong tuning characteristics across the full frequency range evaluated by us. But Corsair provides more than just well-chosen components: due to the screw connection and thin thermal pads, a beautifully built, sturdy heatsink effectively distributes the waste heat produced by the chips on both sides of the board. In addition, you’ll receive brighter-than-average RGB LEDs that can be individually addressed, as well as Capellix technology, which allows for a more compact design and higher light output per watt. But there are shadows when there is a lot of light: Corsair is well worth the price of this strong overall bundle.
Conclusion: The RAM for Ryzen 9 3900X has the best price-performance ratio.
Hyper X Predator RGB is ranked third.
2 Single-Rank-DIMMs, 16GB DDR4-3200 RAM
- RAM at 3200 MHz
- For gaming, 16GB is ideal.
- Heat spreader made of abrasive aluminum
- Profiles that are compatible with XMP
The timings 16-18-18-36 at the clock stage DDR4-3200 and 1.35 Volt are given for the two 8-GiByte bands. They shocked us in the test with frequency-friendly SK-Hynix chips based on 18 nm manufacturing (C-Die), which also achieved the highest DDR4-3866 clock rate tested at 1.35 Volt stably at the timings 18-20-20-60. The 15 percent undervolting potential heard out undercuts the implanted chips’ reserves, which are only limitedly useful for latency adjustment. Nonetheless, given the suggested price, the whole bundle is especially appealing to RGB LED enthusiasts. The lighting is one of the most consistent options on the market, and it mainly sends light upwards. An unappealing detail: the heat conduction pads on both sticks of our kit were too short, not covering the corresponding outside chips, preventing direct contact with the heat sink. Despite the fact that the temperatures we recorded were quite modest,
Conclusion: For Ryzen 9 3900X, 16GB RAM is a great performer.
RAM stands for Random Access Memory.
RAM stands for Random Access Memory and refers to a kind of memory that is utilized as the computer’s primary memory. Individual memory cells are accessed directly through addresses in this memory type. Without a constant power source, data saved in common RAM modules is lost.
The main memory of a computer stores all presently executing applications, related data, and at least portions of the operating system. DRAM memory chips (Dynamic Random Access Memory) are either placed on exchangeable memory modules or soldered directly into the computer’s mainboard and make up the physical memory of today’s PCs, servers, and notebooks. The term “dynamic” refers to the need to refresh the contents of memory cells row by row at cyclic intervals. DRAM memories are logically linked in a matrix. A row and column address are used to address each memory cell. In addition to the memory’s clock frequency, the time it takes to access a memory cell is a determinant in performance.
DDR3-SDRAM or DDR4-SDRAM are the memory types now utilized in laptops and PCs. JEDEC has standardized them. SDRAM stands for Synchronous Dynamic Random Access Memory, which is DRAM with an external memory controller that controls the clock rate. Data is transmitted on the rising and falling edges of the clock signal in DDR-SDRAM (Double Data Rate), which speeds up communication between the memory chip and the processor’s memory controller. The I/O clock has been doubled in DDR3 SDRAM, whereas the clock frequency of the memory chips has been raised in DDR4 SDRAM compared to DDR3 SDRAM.
In today’s laptops, DDR3-SDRAM is mainly utilized in the low-power version. The memory types GDDR4 and GDDR5 utilized in today’s graphics cards are both DDR3-SDRAM enhancements. JEDEC has standardized the various memory types. The primary memory in desktop PCs is typically installed as DIMMs (Dual Inline Memory Modules), while in laptops, SO-DIMMs (Small Outline Dual Inline Memory Modules) are utilized. The main memory is soldered to the mainboard on extremely flat devices and cannot be changed or updated.
Buyer’s Guide to RAM
In order to offer you with the most up-to-date buying advice, we constantly update the test findings mentioned below, including RAM recommendations. Furthermore, we discuss price reductions for specific RAM kits as well as news that is relevant to the purchasing choice.
What is the difference between GB and GiB?
Many online stores and even manufacturers advertise “16 GB DDR4 RAM,” however this is not entirely accurate. As a result, instead of using the capacity standard “GB” for “Gigabyte,” PCGH utilizes “GiB” for “Gibibyte.” The explanation is straightforward: Multiples based on powers of ten are denoted by prefixes like “Kilo,” “Mega,” or “Giga” (10 to the power of x). The word “kilo” originates from Greek and means “thousand” (or 103), thus a kilobyte is 1,000 bytes. However, since a bit with 0 or 1 can only represent two distinct values and is a binary unit, powers of two are often employed in data processing. As a result, what is usually referred to as “1 GB RAM” is really 1,024 “megabytes” rather than 1,000 (103). (correct: mebibyte). The following is true about RAM: A gibibibyte is made up of 1,024 mebibytes, which are divided into 1,024 kibibibytes. These odd-sounding names are made up of the prefixes “Giga,” “Mega,” and “Kilo,” as well as the word “Bi,” which denotes a “binary” connotation.
DDR4 RAM: How much RAM do I need (8 GB, 16 GB, or more)?
The mother of all queries is, “How much RAM do I really require?” We suggest that gamers have at least 16 GB of RAM, since having less memory is becoming more common. In traditional Fps testing, the changes are typically minor, however frametime benchmarks often show substantially fewer outliers when utilizing 16 GiB rather than 8 GiB. As a result, having more RAM guarantees a more consistent picture output – keep an eye out for outliers in the Frametime benchmark below this text section. The minimal needs for ARK Park, for example, are already at 16 GiBytes, indicating a definite trend. According to our study, 32 GiByte vs. 16 GiByte in games provides significant benefits, although not always decisive advantages. If you utilize open-world game modifications or sophisticated multimedia applications for image/video editing, run virtualization, operate with a big number of memory-hungry apps at the same time, or just don’t want to make any sacrifices, 32 GiByte is definitely the best option (or more). With little setup work, unused memory may also be utilized as a RAM disk.
Another benefit of purchasing a kit with a large memory capacity is that you are more likely to get sticks with an internal dual-rank structure, which are somewhat quicker than single-rank modules running at the same clock rate. It’s usually single-rank RAM if chips are only placed on one side of the module. DDR4-bars with 4 GiByte are always single-ranked, as are DDR4-bars with 8 GiByte. DDR3 sticks with 8 gigabytes and DDR4 sticks with 16 gigabytes are always dual-ranked.
What are the advantages of DDR4 RAM’s high clock rates and tight timing?
“Buying RAM that is faster than DDR3-1600/DDR4-2133 is a complete waste of money!” – Unfortunately, comments of this kind are often heard and, as a result, are sweepingly stated and plain incorrect. What is true is that whether higher RAM clock frequencies contribute to increased performance in practice varies by case. There are certain applications where RAM has a significant impact on computation performance. One of these uses is compressing and encrypting an archive using the 7-Zip program. DDR4-2133 RAM may take up to 50% longer to process a big archive than DDR4-3600 RAM — and this is with the same amount of memory and CPU performance.
In games, whether the graphics card serves as a brake pad is crucial. Those who like to play at high resolutions with maximum details and picture embellishment choices such as edge smoothing are reliant on their graphics card’s performance. If the CPU and RAM performance does not need to be fully used, then acceleration is of little or no value. If you wish to obtain the maximum possible frame rate in order to properly drive a display with at least 120 Hz, the situation is different. When working with three-digit Fps numbers, the computational burden is typically decreased by reducing the resolution/detail level, allowing the CPU and RAM to shine. To be clear, test results from The Witcher 3: Wild Hunt appear once in 1,280 720 and once in 3,840 2,160. (simply select)
If you have the option between a higher clock frequency and shorter main latencies, you should usually choose the clock rate since the higher clock frequency gives a better outcome in reality with the same mathematical, theoretical benefit. DDR4-2133/CL11 achieves a CAS latency of 10.3 nanoseconds when computed. With 12.0 nanoseconds, DDR4-2666/CL16 is inferior on paper. However, when comparing the results obtained in reality, it is clear that the DDR4-2666 solution has the upper hand (not only in The Witcher 3).
It only acts differently in extreme situations (for example, DDR4-2133 CL10 vs. DDR4-2400 CL18). However, the maximum clock frequency varies per platform and CPU. If you hit this limit, latency tweaking may help you obtain even more performance out of the system. By the way, we looked at how much Coffee Lake (Core i7-8700K/Core i9-9900K), Ryzen (Ryzen 7 1700X/Ryzen 7 2700X), Skylake X (Core i7-7820X), and Threadripper (Ryzen TR 2950X) benefit from channels, clock and timings, as well as the number of modules and ranks in applications and games in our issues 01/2018 and 01/2019.
Tests and prices DDR4 RAM prices will rise somewhat in 2020.
DDR4 RAM is presently available for about $5 per Gigabyte, so budget-conscious users should have no trouble getting it. Many predictions for 2020 anticipated price rises, but the financial crisis has complicated the evaluation. Due to the great degree of automation, memory chip manufacturing is rarely impacted by the impacts of the corona virus, according to chip makers such as Micron, which are taking measures to preserve output. The economic crisis, on the other hand, is impacting demand in different market sectors, such as smartphones (lower) and data centers (higher), causing production capacity to be modified appropriately. Prices may rise somewhat in the near run (due to increasing demand owing to teleworking, etc.) but decrease again in the long term, since demand for DRAMs is expected to decline during a recession.
Detailed information about RAM: RAM cooler, main memory upgrade or extension
When it comes to RAM, there are numerous additional factors to consider, including the possibility of upgrading or expanding RAM. Many people are also unsure about the benefits of a RAM cooler. If you want to increase your RAM, use the following checklist to get started:
- On the motherboard, how many (free) DDR3 RAM / DDR4 RAM slots are there?
- Do components protrude from the slots, restricting the RAM modules’ height?
- What is the maximum memory (in GiBytes/MiBytes) that may be used (total/per module)?
- Is there anything more that has to be considered, such as the memory clock rate and internal structure (memory ranks per module)?
When it comes to RAM coolers, neither DDR3 nor DDR4 RAM for desktop PCs need cooling. If RAM overclocking is to be performed with increased voltages and increased temperatures in the housing are to be expected – for example, due to overclocking of other components and the use of graphics cards with an open heat sink and axial fans – we recommend using modules with a heat sink or installing a RAM cooler from the accessories market.
CPU and RAM collaboration
The human brain’s operation is much more comparable to that of a computer than it seems at first glance. The computer contains mass storage and working memory, which serve the same functions as our long-term and short-term memories. The brain also contains a control unit, which is similar to the CPU in a computer, that can call up information as required. There are also veins and the heart, which correlate to the motherboard’s data pathways and chipset. The CPU is frequently referred to as the “heart” of a computer. However, if you look carefully at how the computer works, you’ll see that it’s the chipset that causes the system to pulse.
The aorta is a highly unique vein on the motherboard. The data route between the CPU and the main memory is known as the Frontside Bus (FSB). This data route houses the memory controller hub. If you’ve ever heard the phrase “Northbridge,” here’s what it means: The memory controller hub, or the chip on the frontside bus, is what this refers to. Memory controllers are housed in the memory controller hub. All lines from the main memory that go to the CPU or the AGP/PCI interfaces are managed by the memory controller. The memory controller also addresses lines to caches that are not in the CPU, but not through the frontside bus, but rather via the backside bus. The memory controller does not need to link to the working memory separately for information that the processor stores in the external cache; instead, it may utilize the information straight from the cache.
Cache Memory, or simply Cache, is the clipboard for the clipboard, in a sense. As a result, it is a level of memory organization perfection. This is also true in reality, since the cache is a tiny memory bank that is as near to the PU as feasible (Parent Unit). There may be two, and potentially even more, caches on the board for the CPU. However, having more than one level on a board is uncommon. It is normal for a PC to have two CPU caches at the same time. How is it possible since the mainboard typically only has one cache? The key is to build a memory unit that simultaneously serves as a cache right in the CPU.
The level of each cache belonging to the CPU is used to categorize it. The lower the level, the closer the cache unit is to the CPU. Of course, the lowest level is the cache in the CPU itself, which is 1. This cache is thus referred to as the L1 cache, with L denoting level. L2 cache denotes level 2 cache, while L3 cache denotes level 3 cache, according to this standard. The keyboard codes, for example, are an example of information stored in the L1 cache. Because the codes are the same every time and are used often, employing a near-processor memory unit for storage seems reasonable.
You may be wondering why not just put all the RAM as near to the CPU as feasible. Actually, this is what is being attempted, but the distinction between RAM and cache is important. Unlike memory modules, which utilize DRAM, cache employs SRAM, which is considerably quicker. Because the cache is placed directly on the motherboard, SRAM takes up much more space than DRAM. The memory banks for the DRAM modules are positioned as near to the CPU as feasible to save room for SRAMs.
However, on second thinking, caching all of the RAM isn’t a good idea. The cache was created to store data that was used often. More data could be stored in the cache if the cache size was raised, but it’s debatable whether the frequency of usage justifies storing data in the cache at all. Exaggeration would have the effect of reversing the primary benefit of speed. Because of its small size, the cache benefits from its slimness by requiring less addressing time. The cache, unlike DRAM, does not address columns and instead reads whole pages. Cache lines are the names given to the individual pages in the cache.
Using the desk as an example, the comparison could look like this: On the desk, there are compartments that hold papers that are needed that day. In addition, there is a single piece of paper right next to the keyboard. This piece of paper includes vital information that is needed at nearly every stage of the job. As a result, this representation is a refinement of the example we described before. Each time a work step is performed, the documents from the trays must be removed, but the sheet of paper may be examined without difficulty. This example also demonstrates the need of keeping the cache small. The primary benefit would be lost if five pieces of paper were always on the desk.
Cycle of the system
Before the system cycle became the de facto norm, a component’s speed was determined by its access timings in separate time input ns. The days when each computer component performed its own thing are long gone. With SDRAM, the memory integrated itself into the system cycle, resulting in tremendous advancement. Since then, memory speed has been measured in megahertz (MHz) rather than nanoseconds (ns). The latency of the modules, which is measured in the number of clock cycles elapsed during a certain activity, is the same. This has the benefit of allowing a comparison of two components without the need for conversions. Even EDO RAMs, which were also clocked at 66MHz, could keep up with the modules, thus the initial SDRAM modules at 66MHz did not show the full capability of SDRAMs in reality. However, with the next faster modules, the PC100 modules, significant changes emerged. Without a central control unit, the 10ns access times for these modules would have been unusable.
The chipset, particularly the memory controller, determines the system cycle, not the CPU, as is often believed. The chip’s clock cycles are approximately equivalent to the pendulum of a wall clock. The clock, on the other hand, does not move from left to right, but from top to bottom. The following are some synchronization considerations: It’s conceivable that the memory connection was in vain, or that the system clock is a little too sluggish for the modules’ real performance. On the other hand, if the memory needs to notify the chipset every time fresh data is accessible, the amount of resources wasted is considerably greater. In fact, this is a common concern. A train, for example, stops at every station, regardless of whether passengers wish to board or disembark. Garbage cans are also emptied on a regular basis, even if they are not fully loaded. Bulky garbage, on the other hand, is only collected upon request due to the lesser demand. Because data from the memory is now required on a continuous basis, signaling that data is accessible is no longer feasible.
Please provide feedback…
Only the memory controller cares about the precise addressing of the module’s main memory. All other components are based on a usable capacity. According to this logic, the processor is unable to choose any memory cells and must instead rely on the memory controller to do so. For the processor, this is a big pain in the neck, since the separate organization wastes a lot of time. The FSB’s autonomous organization, on the other hand, is critical for a smooth distribution process and is just as vital as its synchronization. Furthermore, the memory controller does not, by default, reply to a memory request directly. As a result, the CPU will have to wait until the memory controller responds. The memory controller, on the other hand, does not get an instantaneous answer to its memory request and must wait a set amount of time for the result.
A memory module’s architecture is similar to that of a table. The rows of this table are referred to as banks – not to be confused with the motherboard’s plug-in bank – and include additional sub-tables, notably the memory chips, which are referred to as arrays. Each cell in a table may be referred to by an index that contains the column and row numbers. With the exception of Rambus, all modules use a mix of multiplexing and rapid page mode to operate. Multiplexing implies that the address bits are transmitted in a sequential order rather than all at once. Because RAM is configured in Fast Page Mode, the RAS signal is applied first, followed by a series of signal bits that identify the bank and line. The bank is defined by the first two bits of the signal. The concept of grouping chips into banks is very helpful since it enables two or more processes to execute simultaneously inside the module. The line number of this bank is represented by the following bits.
The addressed row is fed into the Sense Amps (signal amplifiers), and specific cells may then be selected. How are the sense amps loaded with the row? There are two parallel CAS lines for each column, each of which is first charged to a specified charge, the so-called reference level. This procedure is known as RAS Precharge, and the time it takes to complete it is known as RAS Precharge Time. The memory cell is linked to one of the CAS lines when both of the column’s CAS lines are completely charged. The Sense Amp of this column’s job is now to calculate the difference between the two CAS lines, save the result, and write it back to the cell. The RAS-to-CAS delay is the time it takes for all sense amps to store their data. This period has a maximum upper limit, which assures the controller that the data is accessible and that CAS may reference the cells in this bank row. When the CAS signal is delivered, the sense amps move the appropriate value to the output latch, where the memory controller may retrieve it. The benefit of this technique is self-evident: If a page hit occurs, the address logic does not need to link to the array for all other cells in this bank row, and may instead access the index in the sense amps directly. Only in the event of a Page Miss must the RAS be recreated.
After reading this, you’ll understand what the three parameters of the PC specification, which appear after the operating frequency, mean. A DDR module with a 133MHz operating frequency and timing parameters of 222 denotes The CAS lines need two clocks to be preloaded, and the sense amps require two more to give data. Two clocks are required to make data accessible following a column request, as specified by the 2 CL standard. You’ve probably observed that the statistics aren’t identical, since in the case of a page hit, just the CL matters. In any event, this remains true until there is a page miss.
However, if you want to be realistic, you can’t presume a page hit, which happens often but not always in reality. Nevertheless: With a 133 MHz operating frequency, 6 latency cycles and the t AC, a query time of 50.4 ns is theoretically possible. This translates to approximately 152MByte/s of data throughput. So, how could such a module achieve claimed transfer speeds of more than 2GByte/s? The bandwidth requirements determine how many bytes per second may potentially travel across the contact strip, not what the module can actually make accessible. The harvest, however, is not as small as it seems at first sight. The transmission may easily be 15–30 times greater with appropriate settings than in the “worst-case” estimate. This, however, requires high-quality modules that support different optimization functions.
The best ram for ryzen 9 3900x reddit is a question that many people are asking. The 3 best DDR4 RAM for AMD Ryzen 9 3900X CPU (2020) are the Corsair Vengeance RGB Pro, G.SKILL Ripjaws V Series, and the Crucial Ballistix Sport LT BLS8G4D240FSA.
Frequently Asked Questions
Which RAM is best for Ryzen 9 3900X?
The best RAM for the Ryzen 9 3900X is G.Skill Trident Z RGB 32GB (4x8GB).
What RAM is best for Ryzen 9 5900x?
8GB DDR4 3000MHz is a good option for Ryzen 9 5900x.
Which RAM is best for AMD processor?
16GB is the best RAM for AMD processors.
- best ram for ryzen 9 3900x 2020
- amd ryzen 9 3900x ram compatibility
- best ram for ryzen 9 3900xt
- best ram for ryzen 3700x
- ryzen 9 3900x 3600mhz ram