Expo vs manual RAM OC

[KrzysiuM]

Is there a difference beetween EXPO and manual one? For instance i had enable expo on 6000mhz cl 30 36 36 80 and change it manually withou EXPO enabled by memory try it on msi mobo to 6000mhz cl 30 36 36 76 and it seems stable. Will it be better than EXPO one or it has some other thing inside that technology that boosts ram speed and performance?

[_I_]

expo should enable xmp for the ram profile
its not overclocking the ram, only the cpus memory controller

[KrzysiuM]

So which one is better for gaming performance?

[_I_]

beyond freq/cl it doesnt make enough difference

[r.linder]

XMP/DOCP/EXPO exist to standardise memory profiles and simplify the whole process, RAM overclocking is extremely time consuming when you're trying to get the best results.

Just enable EXPO and leave it.

[KrzysiuM]

But with expo i had some problems with booting. Orange/yellow led lighted up and couldnt boot when iam using manual OC it seems gine but benchmarks on 3d mark are little bit lower

[r.linder]

Then your RAM is either too much for the CPU to handle, has compatibility issues with the CPU and motherboard, or it's faulty RAM

In any of these cases you should be returning your RAM and getting a different kit

[A&A]

Of course it will be faster.

[Lixire]

Very depends on the configuration itself. You can very much take the EXPO profile and tweak it for yourself like changing the sub timings or changing voltages
EXPO is simply pulling out the configuration on the SPD chip on the RAM and applies it for your system like clocks, voltages, primary timings and subtimings

If your system is more stable with your tweaked profile of DDR5@6000 CL30 then use it first. stability is far more important in your case
I would also make sure to stress test the RAM to check for any crashes or other issues

[Faiyez]

Literally no reason to mess with that when expo is a thing.

[r.linder]

Originally posted by Faiyez:

Literally no reason to mess with that when expo is a thing.

https://steamcommunity.com/discussions/forum/11/4691154153116796092/?tscn=1729376656#c4691154153116804056

It's not stable

[Bad 💀 Motha]

Never do manual, always use XMP or EXPO as these profiles in embedded in RAM onboard ROM and thus when selected will also apply the correct timings and voltage, not just frequency. You can still manually change and/or OC the CPU even with XMP/EXPO applied to DRAM

[GOD RAYS ON ULTRA™]

Why don't you just run the normal speed? Not sure what this xmp/expo fetish is all about, but every time I turn that stuff on I notice absolutely nothing. Nothing at all unless I open up task manager and it shows me the speed, then I feel so much better (NOT).

People overthink these things and pay way too much attention to this. You just need a good kit that runs at the supported speed in four dimms and that's literally it. I do this on my 12600k just running 4800 and it stomps the living daylights out of video games.

Ive had it as high as 5600 and as low as 4000, no difference at all. All I did was waste time. Not anymore, No more wasting time for me with things that are meaningless.

Man the marketing has everyone feeling the fomo, xmp, expo, ray tracing, dlss, bla bla hoo hah worthless nonsense. I'm cruising at 120fps and just laughing my ass off.

[_I_]

the default speed is very low with high timings
its throwing away free performance that the ram can run at

if its unstable at xmp/docp/expo, just lower the freq/speed one tick and try again til it works

[Tonepoet]

I wouldn't exactly call it free performance. Faster rated R.A.M. tends to be more expensive than slower rated R.A.M. It is a bit of a perf. boost though.

Buying 32 gigs of Corsair Vengence 48000 mhz C.L. $40 R.A.M. costs $88[www.amazon.com] on Amazon currently, whereas corsair 6400 CL32 R.A.M. is $115[www.amazon.com]. Granted, the cost differential is only $27, which is cheap, but last I checked, $27 isn't free.

You can see on Techspot's test results[www.techspot.com] that with a 12700k and an RX 4090 that 4800mhz C.L. 40 R.A.M. runs at 127 F.P.S. on average, whereas 6400 C.L. 32 runs at 141 F.P.S.

The differential is 14 frames, so you're basically saying that the slower R.A.M. makes the system approximately 10% weaker than it otherwise would be on average. It's something you do to eek out the last bit of performance out of a system if you're min-maxxing your perf.

But also if you're building a roughly $2000 top of the line rig to play with God Rays on Ultra settings $27 is less than 2% of the system cost.

Tom's Hardware's testing shows that the 12900k runs at 178 F.P.S. on average, whereas the 12700k runs at 169 F.P.S. on average at 1080p.[www.tomshardware.com] That's only 9 frames faster. The 12700k is about 5% weaker.

Going from Amazon pricing a 12700k is $232, whereas a 12900k is $278. That's $46 more. By upgrading the processor instead of the R.A.M. you're spending 20% more to only get half of the performance uplift if I have all of my math right

Something else that needs to be considered is that you're more likely to carry forward your R.A.M. from system to system than you are to carry forward your processor, so not only is the uplift in R.A.M. speed more cost effective, but the uplift you get is likely to last you longer in the grand scheme of things.

Plus by going with a weaker C.P.U. rather than weaker R.A.M., you'll either get a larger perf. uplift when you do finally decide to upgrade the C.P.U. again or the same perf with a weaker C.P.U. that saves you more money.

Now don't get me wrong here. I think R.A.M. speed is a rather min-maxxy thing to concern yourself with in the grand scheme of things. I don't normally expect it to make a 10% perf. difference. I'm just saying it makes more sense to concern yourself with R.A.M. speed than it does to concern yourself with the processor sometimes, especially if you're higher up the processor stack already, where it costs more to upgrade the processor than the R.A.M.

[r.linder]

It's also worse if you buy RAM but don't even run it at its rated specs and choose to run it lower, defeats the whole point. You may as well have just bought the cheapest RAM at that point.