Search results

b die is not the best RAM for high frequency and depending on the sticks PCB design being A1, A2, A0 can have an effect on the max frequency. ~4400 is a good goal for frequency with b die and aim for tight timings.

51c is high enough that it can create instability with tighter timings so try keep them under 50c if you can which should be doable by improving airflow. My b die runs 4000 16-16-16 with ~1.47v you should be able to run them at a similar speed\v. 4266-4400 should run the same timings with more...

If Micron it will likely top out below 6000 and still have high latency no where near what buildzoid recommends for hynix.

Is it Samsung b die or something else? Exact model of the RAM please. If it is b die you can safely run ~1.55v with a fan blowing directly over it or up to 1.5v just with good case airflow. Some run 1.6-1.7v 24\7 but this can require tweaking of other settings to make it safe.

You can still buy decent kits of samsung b die but it's rather expensive per GB now and just doesn't make good sense to buy...

Some of the first DDR4 made Samsung b die was the fastest. Newer models that were made on smaller processes which made them cheaper per chip and with higher density's which made them cheaper per GB were unfortunately never able to match the performance of the more expensive b die.

If you play one of the few games that drops below 60FPS due to a CPU bottleneck even with the fastest CPU and sees decent gains from high RAM speeds then it can make a bigger difference than upgrading from 12900K to 14900K. But there is a lot of programs that see little to no difference and for...

Memory speeds over 6400 can bring decent gains to CPU performance for Intel in some programs but it will make no difference to others which fit within the CPU cache. AMD however may see a drop in performance especially in latency sensitive tasks when going over 6000\6400. Also keep in mind you...

You can get desktop RAM with higher JEDEC speeds but usually with really loose timings as they are limited by the vdimm. Most XMP kits start at a low speed for stability and allow you to increase V and performance with XMP. A lot of laptops don't support XMP so I guess you just see more of one...

It helps to stabilize the memory controller so it applies to all MB. Up to 1.4Vdimm and a boost to SOC as suggested above is the easiest wway to improve stability with 4 sticks.

It's ok for finding if you have a dead stick but useless for memory controller instability and not great for slightly unstable RAM. OCCT RAM test and large CPU stress test are good for checking RAM and memory controller stability. https://www.ocbase.com/download By tweaking BIOS settings it...

If its 16-19-19 you may be looking at ~0-3% if its 16-16-16 maybe 0-5% in some CPU limited sections of gamepaly. So if you think that is worth the $ sure but there is other ways to get much bigger gains like 5800X3D or 7800X3D. As mentioned tweaking the current kit will get you bigger gains than...

What are the timings on the current RAM? for example 3200c14 to 3600c18 would be slower than what you have and 3200c14 can OC to 3600-3800c14. Going over 3200 will also change something within the CPU that will make it use ~4w more power idle and load which takes a smidge away from the power...

As mentioned 4 sticks of RAM can be hard for Ryzen 3000 but Ryzen 5000 handles it very well if there single rank and can typically do 4000+ without a problem but the IF usually tops out around 3733-3800 1:1.

Background processes have a huge impact on AIDA64 latency so it is best to run in safe mode to improve consistency.

Keep in mind that is a very fast kit of DDR4 which is getting expensive and harder to find these days vs what is considered a cheap kit of DDR5 now costing at least half as much. As RAM speed effects file decompression speed of the CPU it can add a little extra zip to load times with a lot of...

In some programs a lot in others it will make no difference at all. Also depends greatly on the speed of DDR4 or DDR5 you are talking. So in what software are you wondering if there will be a difference?

This was with older BIOS as the divider would have to change after 6000 to maintain stability which incurred a small latency hit. Newer BIOS do fairly well up to 6400 as they have made some tweaks to improve stability of the FCLK up to 2200 is possible with some CPU\MB now. But more important...

It is a rough calculation to give an idea of how frequency vs cl comes out as far as the sticks latency goes and yes that is correct the 2400 kit with the same cl will have lower latency. 8.3ns vs 10.7ns

If they are both free or near free I would go with the 4x8GB 2400. 1600 to 2400 with some tuning can improve CPU performance up to 15% in a CPU limited game like FS. The MB is about as good as it gets for running 4 high speed sticks of RAM and the 4770K can run well over 2400 while 2600K was...

This will drop performance unless your already running it around 4800MHz as it is much much hard to hit higher frequency with 4 stick of DDR5. You are better of with a leaser amount of RAM around 7200-7600 on Intel and around 6400 for AMD.

64GB will bring a small boost to CPU performance if the 32GB kit is single rank.

Temps will be fine probably only a few c above the air blowing over them. The issue may be how the HS are attached as when they are glued on with thermal tape sometimes the memory chips come off the stick with the tape. Make sure you get the sticks nice and warm before removing them to reduce...

Anything 3200 c14-14-14 will be Samsung b die and thus dual rank with 16gb sticks.

If it is slower in synthetics for both latency and bandwidth then it will be slower in the real world but depending on the program it just may not be a noticeable or even measurable difference. If latency or bandwidth was better for one setting vs the other then in the real world programs would...

He shows 6400 28-37-37-28-1T auto subs performing ~58-59ns in AIDA64 and ~ scoring about 10k lower in write\copy vs 6400 40-40-40-84-2T tuned subs which managed ~54-58ns Also PYPrime scored 9.862 vs 8.894

Also available in 64gb

This SODIMM kit from Kingston will run 5600 c40 without enabling XMP. https://www.kingston.com/datasheets/KF556S40IBK2-32.pdf

Not sure which DDR5 clocks best with low V but the power use will only be a few watt so I doubt the VRM will have any trouble. One of the things I love about overclocking RAM you get all that performance for bugger all change in power use.

maybe run the OCCT power test and check the smart data for your SSD before the reinstall just to be sure.

I am bettering the dying RAM stick has corrupted some install files. Try a reinstall of video drivers after using DDU to remove the old ones and verify the game files but you may need format the SSD for a compete reinstall of everything.

Sounds like your CPU\RAM is now stable so one problem fixed now to find the other. Run the OCCT GPU stress test.

Well it seems like you have found that you have a stick of RAM that is dying and now you just need to find out if there is anything else or if it has just corrupted some files and you need to reinstall.

Test the working kit at XMP 3600 with OCCT large extreme CPU stress test and Memory stress test. https://www.ocbase.com/download

Could also be Micron e die which can often hit ~4800-5200+ ~1.4-1.45v and while you will lose CPU performance by running outside 1:1 I believe you will still gain IGP performance which is what matters most for you so it may be worth a test with it decoupled and seeing just how far you can go...

The difference between the MB you have tested that work and the one that doesnt could be nothing more that the settings pendragon suggested changing. The BIOS may be setting the V higher on one board than another or it may have more Vdrop than the other. It not really overclocking it is just...

In adition to the above make sure you have the latest BIOS.

Intel gets ~80GB/s multi threaded 40GB/s single threaded with that kit of RAM vs AMD who gets ~65GB/s multi threaded 60GB/s single threaded. So the website is just saying that anything that is much below 80GB/s multi threaded is under performing and mostly ignoring the latency and single...

👍 Looks like 5800X3D large cache adds a fair bit of memory latency so getting under 57ns is fairly good I would say and may be comparable to 5800X getting around 51ns. https://hothardware.com/reviews/amd-ryzen-7-5800x3d-review-and-benchmarks?page=2

Probably more to do with MB makers like GB tweaking the sub timings beyond what the kit is rated for. Gives a nice performance bump when it works but some sticks need a little extra juice to stabilize.