RAM Overclocking Guide: How (and Why) to Tweak Your Memory

How To Overclock RAM

Playing the Silicon Lottery

There is a lot of chip-to-chip variability when it comes to memory, more so for DDR4 than DDR3. Two identical modules, from the same manufacturing batch, may be able to take vastly different maximum voltages before becoming unstable, but only in certain memory slots. While vendors do test for the performance of each IC, a memory chip is only guaranteed to perform at the advertised specifications. The differences show up when you try to overclock them.

Multiple memory packages can all have the same memory module inside, and there are chip-to-chip differences between ICs, even if the batch-number (when available) is the same.Multiple memory packages can all have the same memory module inside, and there are chip-to-chip differences between ICs, even if the batch-number (when available) is the same.

Given how relatively inexpensive memory can be, serious overclockers generally purchase multiple kits, test each module, and select the best. The testing/selection procedure for this portion involves placing the modules (one at a time) in the same DIMM slot with the same memory parameters, and finding the modules that run benchmarks consistently (SuperPi 32m is recommended here) with the lowest DRAM voltage possible. The modules that pass with the lowest voltage are the best pieces of silicon.

After the best modules are selected, each is rotated into each DIMM slot on the motherboard and tested again to find which specific module performs best in which position. This last check should be performed even for a configuration that is restricted to an already-purchased kit. At this point, each DIMM should be physically labeled (a sticker works well) to identify the slot it belongs in; keeping a log of the lowest voltages for each and the parameters used for this test is very useful if long-term damage is suspected somewhere down the line.

Software Tools For Overclocking Memory

Parameters for memory overclocking can be changed via the motherboard firmware or through vendor-supplied software. Many motherboard vendors offer tuning utilities that incorporate stress testing and parameter manipulation, and there are freeware options like CPU-Z that can provide a quick system report and real-time measurement of each component’s operating frequencies.

CPU-Z Memory Information Tabs, courtesy of user jaquithCPU-Z Memory Information Tabs, courtesy of user jaquith

Intel offers its Extreme Memory Profiles (XMP), which consist of pre-defined and validated overclocking settings that can be loaded via motherboard firmware or a vendor tuning utility. XMP allows the firmware/utility to automatically configure the DRAM voltage and latencies, and it can be a good option for those wishing to work with pre-optimized variables.

A SuperPi Memory Scaling/Timing reportA SuperPi Memory Scaling/Timing report

For stress-testing, the community's current favorite utility seems to be SuperPi, followed closely by Memtest86+. Both tools have extensive configuration options for running tests. Final benchmarking should be carried out using software that most closely mimics the application the system was intended for, like 3DMark for graphics applications and rendering, WinRAR, virtual machine performance, MATLAB’s memtest, etc.

Basic Overclocking

Memory overclocking, like overclocking the processor, requires iterative tuning and patience. The general procedure is:

  1. Confirm stability. You can use Memtest86+, SuperPi 32M, Intel’s Extreme Memory tool, or a motherboard/vendor supplied software suite for this.
  2. Note “good” default parameters (parameters that can be returned to if all hell breaks loose).
  3. Confirm (via the motherboard firmware or software suite) that the memory frequency, timings/latency, and voltage values are the ones advertised by the memory vendor. Repeat Step 1 if any changes are made.
  4. Set the memory multiplier to its maximum allowable value, repeat Step 1.
  5. Increase the BCLK frequency by some small amount (10 Hz or so), repeat Step 1. We veer away from optimizing the memory frequency further if the BCLK being used has been set with regard to processor overclocking considerations. If the memory is unstable with the maximum memory multiplier at this point, it is always possible to reduce BCLK, increase the CPU multiplier, or reduce the memory multiplier (or any combination of these parameters) to achieve system stability. You should adjust the VTT in tandem with the BCLK.
  6. In case of problems (or just to see if it makes a difference), increase the DRAM voltage by a very small increment (0.01V, for example) and repeat Step 1.
  7. The CMD should be set to 1 (the motherboard firmware can call this variable CR1/CR2 or T1/T2).
  8. Tighten the primary memory timings, repeat Step 1. Ideally, the timings would be tightened (decreased) for better performance, but before that it is worth loosening (increasing) them to see if a higher BCLK or memory multiplier can be tolerated by the system with slightly increased latencies. Tightening the memory timings begins with an adjustment of the primary timings—a benchmark/stress test should be carried out after each set of changes—and proceeds in an iterative fashion to lower each number in the primary timings set. Secondary and tertiary timings have a much smaller impact on overall performance, but they can be adjusted in the same manner.

A SuperPi Memory Latency comparison of a single memory IC at various frequenciesA SuperPi Memory Latency comparison of a single memory IC at various frequencies

Given the number of variables and minute changes that can have large effects, a stopping point for the optimization can be any of the following:

  • Maximum safe DRAM voltage reached
  • Maximum BCLK reached, or an incrementally higher BCLK is unstable despite all other parameters being optimized.
  • A maximum RAM frequency is reached (for DDR3; DDR4 doesn’t have a ceiling as of yet)
  • The memory ICs refuse to boot because of thermal or overvoltage damage

We can also go in the opposite direction. You can underclock the memory for improved power savings; lower voltage will lower power consumption a bit, although the clock rate has a far bigger impact.

RTL And IOL: Raw RAM Performance Metrics

Advanced overclocking begins with looking at the memory configuration's real latency. RTL and IOL values can be used as fitness scores to optimize memory performance; the lower the better. Also, these figures are given in clock cycles, but they really should be converted to real time values to get a good idea of the actual latency (the clock frequencies may change).

IOL, the Input Output Latency (also referred to as I/O Latency) is the time it takes the chip to send a response after the query comes in.

RTL, the Round Trip Latency, is the length of time it takes for a signal to be sent to the memory, plus the length of time it takes for an acknowledgment from memory of that signal to be received (the complete round-trip time taken for a signal to be transmitted from point A to point B and then back to point A). RTL values are not always directly (manually) accessible, but they are a function of the IMC frequency, tCL and clock skew (mismatch) between the IMC and DRAM clock frequency. Our sister site Anandtech extrapolated a formula for correctly predicting the RTL values. Of course, their formula requires some motherboard-specific modification, mostly based on the board layout, specifically the distance from the DIMM slots to the CPU.

These values may be directly accessible on some motherboards, in which case it makes sense to set them to some static “good” value to minimize instability issues, which have the potential to increase as more parameters—in addition to the BCLK/Multiplier/Primary timings—are tweaked. Where the RTL initial value and IO latency Offset are manually accessible, the easiest way to tighten the RTL/IO numbers is to set the RTL Initial Value to the lowest number that allows the memory to POST, then, after the system has booted successfully, to set the IO Latency Offset value higher one cycle number at a time (iteratively) until the lowest RTL/IO values are found after a reboot.

Memory Training

There are multiple clocks in a system (CPU, IMC, memory, etc.), with a variable tick/tock initialization from start-up to start-up, a wide variety of signal pathways and variable environmental parameters, all of which combine to create a disparity (a skew) in the real arrival time of various signals at their destination. The pre-boot DDR calibration sequence introduces various delays between signals in order to achieve synchronicity. This is where DDR training kicks in; there are a number of patterns (either preset/provided by vendors, or custom-made) that test various signal/delay sets for the best possible ranges of these values. The accuracy of these delays determines the RTL/IOL, and ultimately influences memory performance. Since RTLs and IOLs are set at boot, training has a very real impact on the CAS latency.

Fast Boot settings either skip the memory training entirely, or use a very rough-and-ready form of training. While this is good enough for normal purposes, the best possible training sequence (determined from literature, or comparing the RTL/IOL values resulting from using each test, or in the absence of additional data, using the sequence that takes the longest time) should be used when fine-tuning memory parameters or benchmarking, because the variable signal/delay accuracy from a sub-par training regime makes parameter comparison questionable. Still, if enthusiasts are looking for moderate increases in memory performance, this step is generally optional.

Create a new thread in the Reviews comments forum about this subject
This thread is closed for comments
30 comments
Comment from the forums
    Your comment
  • Sakkura
    Perfect timing, I was just about to get to work on my DDR3-1600 kit. Gotta squeeze out all the performance possible from my aging Ivy Bridge build. Especially now that I'm running VR.
  • Amdlova
    No i don't want oc the memory if i can't oc the ssd :)
  • BobsKnob
    You know why there's no benchmarks? Because overclocking memory offers such small 'real world' gains that it's not even worth it. You might as well try to overclock your usb ports. This is just a lame attempt to advertise products, in this case Kingston and Arctic heat spreaders and try to pass it off as reporting.
  • derekullo
    They also mentioned Mushkin, Crucial and G.Skill.

    Heaven forbid they mention the manufacturer of the ram in an article about overclocking ram.

    Now if this would have been an article about Farcry 5 or Elder Scrolls 6 and they had a picture of an Arctic Heat Spreader then you would have been on to something.
  • BobsKnob
    Thank you for proving my point. It's just advertising masquerading as reporting.

    You know, there used to be a time when Tom's did real, unbiased reviews and reporting. When something sucked...it sucked and the review wasn't candy coated.
  • damric
    Well done. Will there be a Part 2 that dares to tweak secondary timings and beyond?
  • dish_moose
    I'm with Bob on this one - until you show real wold benchmarks showing real improvement from OC RAM, I won't waste my time.
    -Bruce
  • Maxxtraxx
    Check out the youtube channel Digital Foundry, this video specifically: https://www.youtube.com/watch?v=Er_Fuz54U0Y
  • Maxxtraxx
    Proof of increased performance from Ram Overclocking from real world benchmarks can be found here:
    https://www.youtube.com/watch?v=Er_Fuz54U0Y

    Am I not allowed to edit my own posts? or am i really missing something here? I wanted to say more after I posted but could not find any way to edit my own post.
  • BobsKnob
    What GPU is being used for that test?
  • BobsKnob
    Also, when I skip through that video, I see a lot of the FPS's are similar regardless of ram speed.
  • Maxxtraxx
    Watch the whole GTA5 section of the video, also here is a second video from DF on the same subject:

    https://www.youtube.com/watch?v=qksXthUcbiQ

    You'll see the same FPS results there the gains are anywhere from a few FPS more under light CPU loads to 20FPS more with heavier loads.

    The amount of FPS gained by RAM overclocking is much less consistent when compared to GPU overclocking due to it being bound to the CPU job of keeping the GPU fully fed without bottle-necking.

    The CPU can go from doing very little to supply the GPU to doing a great deal at different times during the game. The GPU, unless it hits a frame limiter like v-sync will be running at 100% all the time unless it is bottle-necked by the CPU.
  • blackmagnum
    Did you know that overclocking the memory might corrupt the data? That's why you should only overclock ECC memory.
  • inmyrav
    How about some measurements to show the alleged benefits of overclocking memory? I have never seen a significant impact on gaming performance even from significant overclocks / reduced timings in memory. I'd be interested to see how much the benefits are and where exactly they exist, if they do?
  • Sabishii Hito
    Memory OC is mainly for fun and competitions, and I thoroughly enjoy it.
  • daglesj
    Buy the ram that looks the coolest in your motherboard. That's the only valid advice you need these days.
  • g-unit1111
    That's very interesting. I had no idea you could identify that much from a RAM module. Kind of like how you can identify everything about a car from the VIN number.
  • JackNaylorPE
    Quote:
    But current-generation heat spreaders from reputable vendors do make a difference in operating temperatures, and when the same kit is provided in two different packages—one with a heat spreader and one without—going for the heat spreader makes technical sense.


    What is the difference ? Why is it significant ? Here I see the aesthetics as the primary driver here. Yes, of course, two sets of modules, one with and one w/o the HS, take the one with ... but what is the thermal issue here ?. I have not heard of an issue with RAM temps since DDR2. As an analogy, while I saw the reasoning behind putting a hybrid cooler on a FuryX, what is the point on a AIB GTX 1070 ? The card operates well below (10C) it's throttling point on air so, exactly, what is the hybrid doing for me ? What do I get out of the extra $100 ?

    The only cooling effect of these big tall toothy coolers is that they "look cool". While they served a purpose (when they were effective) w/ DDR2, they were absolutely useless on DDR3. And DDR4 runs even cooler.

    http://benchmarkreviews.com/index.php?option=com_content&task=view&id=773&Itemid=67&limit=1&limitstart=1
    Quote:
    At more than 2" tall in certain areas the Corsair Vengeance could pose a problem for users like me who use large coolers such as the Scythe Mugen 2. I was able to use the Corsair Vengeance only after I mounted the fan on my cooler on the backside. Size is definitely a concern with heat spreaders of this size and therefore I encourage users to check that they will have enough space under their heatsinks before purchasing the Corsair Vengeance kit.


    http://benchmarkreviews.com/index.php?option=com_content&task=view&id=773&Itemid=67&limit=1&limitstart=6
    Quote:
    The problem I have with the Corsair Vengeance is the same I have with many kits of RAM on the market. Companies insist on putting large coolers on their RAM and it limits the choice in CPU heatsinks that can be used within users system. DDR3 does not require these elaborate coolers with its lower voltages which translate to lower temperatures then RAM saw during the DDR, and DDR2 era. Corsair is correcting this with low profile versions of its Vengeance line but ultimately I would like to see the average size of coolers drop instead of having to look for specific low profile versions of a memory line.


    When such a "makes a difference" statement is made, as a reader, I want to see the data behind the conclusion.. 1) what are the differences in temps ? 2) how is this significant ? and 3) what are the performance impacts in something we actually do (besides benchmarks).

    With regard to the "glued on" comment, this is an issue that we never see mentioned in reviews / comparisons. Some RAM HSs can be removed with a screw ... some are glued on, why not mention this in a comparison review ? GFX card reviews typically include, RAM manufacturer, specs, how it is being cooled and yet this is rarely done for system RAM.

    Would love to see time and effort put in to showing more than "same ole same ole".

    For example, gaming average fps is limited primarily by the GFX card performance... tho there is some general acceptance that CPU can be limiting in certain games, especially when in multi-player mode. What is less accepted is that memory speed can affect performance... and when this subject is argued, we oft see links to tests that "prove" the hypothesis via "google something that shows this". Sometimes that results in a link that can prove both opposing arguments

    http://www.tomshardware.com/reviews/32-gb-ddr3-ram,3790-10.html

    Metro 2033 shows no performance gain going from 1600 to 2400 DDR speed, and yet F1 shows an 11% increase in average fps. Like anything else, system performance is impacted by the weakest link in the chain. If RAM is not the weakest link, then the impact is zero. So a "test" that "proves this" is misleading. To make testing useful, the script should involve determining under what circumstances RAM becomes the bottleneck and then seeing if faster RAM or overclocking can have in impact.

    Among the areas of relevance therefore, based upon past articles on the topic.

    -what games / programs can be impacted ? (CAD, video editing ...)
    -what actions within those games / programs can be impacted ?
    -what parameters need to be analyzed ... for example min fps vs avg fps ?
    -Is the test bed adequate to perform the evaluation ... for example, what happens to previous tests when 2nd GFX card is added ? ... 3rd and 4th

    Was never a popular undertaking, my guess cause of the amount of effort involved, but here's some old links which did look into some of these aspects

    http://www.anandtech.com/show/2792/12

    22.3 % (SLI) increase in minimum frame rates w/ C6 instead of C8 in Far Cry 2
    18% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in Dawn of War
    15% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in World in Conflict

    Also see http://www.bit-tech.net/hardware/memory/2011/01/11/the-best-memory-for-sandy-bridge/1

    http://www.anandtech.com/show/7364/memory-scaling-on-haswell/10
    http://www.anandtech.com/show/6372/memory-performance-16gb-ddr31333-to-ddr32400-on-ivy-bridge-igp-with-gskill/14
  • gamebrigada
    You should mention that some games do not like memory overclocks. Ubisoft is pretty famous in this regard some of their games will run very low frame rates with any overclock no matter how stable. FarCry 4 is a good example.
  • RedJaron
    Everyone asking for benchmarks showing the benefits of OC'd RAM, look back at any number of SBMs, particularly any Codemasters racing games. You will see big fps gains when RAM bandwidth is increased. Other games, hardly anything. This is nothing new. Most games don't care about CPU OCing either.

    However, games aren't the only thing computers are used for. Many professional apps, particularly Adobe products like Premiere and After Effects, are very RAM sensitive.

    Anonymous said:
    You should mention that some games do not like memory overclocks. Ubisoft is pretty famous in this regard some of their games will run very low frame rates with any overclock no matter how stable. FarCry 4 is a good example.
    I'd love to see the proof of this. It's not difficult to OC your RAM poorly, resulting in lower performance all around, not just in one game. My suspicion if that is the case you saw here, if indeed you did see it and aren't simply repeating something told you.