6:09 time has come 9700x 260 pounds vs 7800x3d out of stock or some crazy 460 pounds

I'm running an 8700k, 32GB 3200mhz ddr4, and an RTX 3090 and even I average 40 to 50 fps. GPU hits around 80% utilization and CPU is running around 90% utilization.

It's funny how accurate you predicted the 9800x3d performance :D

can you test tuned settings in other games?

but you throughput showed your settings lag of more then just timing. copy and read is below average.

I'm hesitating between buying the 9700x for 296 EUR and the 7800x3d for 420 EUR for a 4080 super 1440p, which would you choose? thanks

If I am not play Star Citizen, does this still apply to games like cod?

Nice just ordered 9700x yesterday 😨

The game isn't optimized, but you inspired me to take a deeper look into my RAM. Turns out I'm rocking awful RAM CL42 and doesn't even have XMP. Just in time to pick up a black Friday dead

Frame chaser did a video about %lows and 3d chips, he basically said the same. Get a 9950x instead of a 3d chip, the % lows will be much better and basically similar level as 14900k

Nice comparison video

How to OC or tweak my R7 7700 RX 7900XT and Patriot CL34 7600 MHZ ram

so do you overall recommend the 6000 or 6400 more?

I've been telling people that as long as you have a fast memory, X3D chips are almost redundant. They add average fps in games and scenarios where you'd be achieving high framerate anyway and although higher average framerate may look good on paper, you don't feel it. I'd appreciate, if you could incorporate the i7-13700K, maybe with E-cores disabled, in your future test. I've seen people getting very good results with it recently, also by memory tuning.

how i tune my ram best?

Have you already tried on 9700x?: Disable Windows memory integrity Undervolt CPU using a negative all-core curve offset Increase CPU power limits Increase max CPU boost clock Increase Infinity fabric frequency

Amazing video! I was just curious how the 9700x Tuned perform vs 7800x3D, so I'm happy I found exactly what I was looking for, I'm also really shocked at the results. If I didn't buy 7800x3D 4 months ago I'd but 9700x for sure. It's 200 euro less where I'm living. Could you please do more tests with different games? Maybe some RTS?

I have the 9700x and wondering if 9800x3d will be an upgrade playing on 4k, mostly playing modern AAA games and world of warcraft.

Think the 9900x3d will finally beat the old 13900KF?

some details in memory tuning? :D

Cannot understand if more than 1.38v is in risk, why use Vcore offset+ to work over this voltage, i use offset to all cores +2 instead vcore offset+ and reach 41k in CB r23 with 1.298 Vcore, max temp 85º. MSI MPG mobo +13900k + ddr5 6400 And a question, where you found that from 1.30v to 1.38v is in safe for an architecture 10nm?

@gamedaytoday1 coud u make a built for the 9700x with mainboard ram thats best for tuning?

OMG basing superb overall CPU performance of the 9800x3D simply on "ShitCitizen". Memory tuning has way less impact on x3D chips. It has been proven over and over again.

If Star Citizen is such memory dependent, why not test AMD with higher than 6400, new boards are allowing 8000+ , given that it will not be 1:1, but maybe Star Citizen being so memory speed dependent, could offset the 1:1 recommendation.

which bios u using? the latest makes a difference im not buying it these results dont make sense i know SC loves fast memort and volumes with more cores but above all..SC is not yet optimized it will only happen after 4.2 onwards by the time the singleplayer actually comes out there will be a 4.5 as the roadmap goes

I dont get why oh why the tests always gets done at 1920x1080, who still run this as their daily settings??? I run at 4k on my 32” Samsung monitor, I really cannot understand why all testers still use this resolution, can you advise on this please?

Everyone also note: he has it set to Vulkan not DX11 in the Settings screen - but in the recording it’s set to DX11 (??)

I have not understood all the negative reviews about the 9700X. I have a 9700x paired with a MSI X870 Tomahawk and Kingston Fury Renegade Beast DDR5 6400mhz CL30 and it outperforms my 7800X3D system equipped with an Asrock B650E Taichi Lite and G skill Neo 5 Expo DDR5 6000mhz CL30. In gaming it trades blows with the 7800X3D and outright outperforms the 7800X3D in all other tasks. With a few tweaks, I believe it is the better CPU. I have not compared it to the 9800X3D.

Leave your BIOS settings on Default on your own risk

yeah, why not synchronized 8000mt/s??? ..... 6400 is bad.. your video is bad and testing aswell...

Btw pretty sure Hydra can bypass the +200Mhz PBO.

ALSO very imporant: bios 1.2.0.2A is currently bugged it increases latency by 7-9ns vs 1.2.0.2 (non-A) on every board. No matter how good you are no way of going under 60ns on the A version of the bios.

Not having a good time in SC on my Laptop, 14900HX 96gb DDR5 5600mhz 4090, CPU is absolutely pegged at 100% I get around 20-40 fps in NB and even on stations, graphic settings make zero difference, Nvidia overlay says my CPU is 40-55% usage but if I open task manager it says 90-100% on SC.exe 720p or 4k makes no difference. This is only since Patch 3.24. Getting about the same performance on this system as I did on my ROG ALLY in 720p but with no stuttering. . Meanwhile in Arma reforger I can get over 100FPS in 3840x1600p with Discord/OBS scenes, Nvidia Broadcast background removal on webcam and HQ software encoding.

I hope my computer won't explode if I set such hardcore timings. 😉

"The Game isn't unoptimized, your PC is" TELL EM <3

You should test actual games besides scam citizen...it has become a meme at this point and testing other games would show that in many cases the 9800x3d has a huge advantage over everything including the 13900k by a wide margin

i still wonder what morons pay you to see memory timings when you have people like buildzoid that litterally give it all away for free including hints and tips on how to achieve better timings...

Same spot as you but with 9700x + 6800xt 8000mt/s cl36 (~58ns latency) 1.52V 2200fclk (all auto settings) +200pbo (currently top2 firestrike and top1 firestrike extreme in 3dmark for 9700x) 75.4avg / 45.1(1%-low) / 35.4(0.1%-low) - im gpu limited thats why lower avg. Also my capframe x is way smoother with only 2 spikes from 15ms(basE) to 30ms no huge spikes like you have from 20ms to 80ms. You should try 8000 my friend tmr. im trying 8400/2100/2100 and 8400/2200/2100 and 8000/2266/2000

Great video, my man. I'm almost two years late to the party, but it's still relevant! Are you by chance a Swede? I detect an ever so slight Swedish tinge to your accent. Possibly Norwegian.

Which windows did you use for testing?

This is awesome, would love to see more data like this! Can you do these same tests with more games, like Cyberpunk, Baldur's Gate 3, and maybe some older less demanding games like Mech Warrior 5, Valorant ect. Love the 0.1% data! I'd be curious to see data like percentage of frame times over 30ms, or frame rate under 40 fps (or some appropriate thresholds, could be like 50% of aggregate) Because really want we care about is mitigating the lows and reducing lag spikes; the super high highs don't matter AT ALL and thus the average, median would be much better.

1. Higher FCLK is not always better performance. While MCLK:UCLK may be 1:1 and FCLK asynchronous you gain a little latency benefit by going 2:3:3 (FCLK:MCLK:UCLK). So for 6400 an FCLK of 2133 is ideal in most cases. If you go for 6000 or 6200 a FCLK of 2200 can be beneficial. Reason being that the benefit of way higher FCLK compared to the latency penalty of leaving 2:3:3 is usually bigger. Thereforce if you can go 100+ more FCLK you usually should do it. But to go from 2133 to 2200 is only 66 more FCLK and by that losing 2:3:3. It can be better but certainly not always. This especially holds true when you go into Gear 2. Then you can synchronize FCLK with UCLK and leaving that 1:1 synchronization by having more FCLK is a HUGE penatly in latency. 2. What is also missing is a DDR5-8000 comparison. 3. A latency of 65.4 in AIDA seems kinda high. Espcially on 3D chips where AIDA dips into the 3D Cache you should have less than 57ns. MSI released a latency killer feature in their newest BIOS that reduces AIDA by around 10ns. Seems like the latest AGESA increased latency heavily for no reason and that MSI feature just restored what was lost before.

one game tested?hahaha,test please in all games AAA bro

Get rid of the clicking sound lol

This is exactly why Jufus from Framechasers is not a fan of x3D. He's been going on about this for ages.

6400@2133 VS 6400@2200 Not correct, bec 6400@2133 its in synh, but 6400@2200 - not synh. To be correct IF@2200 you need be 6600mhz RAM. latency on x3d in good tune <=60ns. Good tune 6400cl30@2133 can give you ~57ns.(Same on 8000cl36@2000) In general this a good video, thx for this! P.s. Sorry for bad engl. Hope you understand what i mean.

thx good teszt

What do you think about 7600x3d versus 7900x? I’m wondering between the two for gaming

I seem to be unable to get my 6400 mhz ram kit stable with it's xmp.. i tried everything i can think of :( running it at 6000 mhz with my 9800x3d. eventhough its on the QVL of my motherboard.