• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Ryzen 7 9800X3D

opethdisciple said:
Undervolting is the easy way to do it.
Click to expand...
Curve Optimiser on the 9800x3d is easier. However on the 7950x3d testing stability for each core in core cycler and then other tests is far from easy. In order to get full stability in everything including SHA3 it takes a long time.

Doing a fast all core 9800x3d -20 for example would be much easier.
 
Last edited:
McCarthy said:
Has there been any negative reports regarding overclocking the 9800x3d?

Think I’m going to stick to just undervolting my 9950x3d on release to be safe.
Click to expand...

McCarthy said:
Curve Optimiser on the 9800x3d is easier. However on the 7950x3d testing stability for each core in core cycler and then other tests is far from easy. In order to get full stability in everything including SHA3 it takes a long time.

Doing a fast all core 9800x3d -20 for example would be much easier.
Click to expand...

Aye, agree with @McCarthy here, tend to find that the binning on both CCDs isn't quite the same and the cache enabled one will be a slightly better bin so per core curve settings takes a while to figure out
 
Cenuij said:
Aye, agree with @McCarthy here, tend to find that the binning on both CCDs isn't quite the same and the cache enabled one will be a slightly better bin so per core curve settings takes a while to figure out
Click to expand...
It was an eye opener to say the least. I had to back some cores off to -4. I know what I’m in for when the 9950x3d arrives.

It’s worth it in the end. That system was rock solid.
 
McCarthy said:
Curve Optimiser on the 9800x3d is easier. However on the 7950x3d testing stability for each core in core cycler and then other tests is far from easy. In order to get full stability in everything including SHA3 it takes a long time.

Doing a fast all core 9800x3d -20 for example would be much easier.
Click to expand...
I've done several AM4 dual CCD chips using my method of CO per core and it's been very easy to get things right. I've tested also 2x R5 9600X, 1x R7 9600X and R7 9800X3D using my method and TBH things like AIDA64 SHA was walk in the park.

One such CPU test data in OP of my thread on OCN, below quote is from section CO profile stability testing.
Next ZIP of stability testing, sort files by time to which test I ran when. I didn't test a single test for lots of hours, as wanted to test multiple loads.

1. CoreCycler P95 PASS 12hrs.
2. CoreCycler Y-Cruncher PASS 6hrs.
3. Standard Y-Cruncher stress test 15min, 80C temp, cTDP 105W.

Screenshot of standard Y-Cruncher stress test, CO profile cTDP 105W temperature limit 95C (HWINFO CSV in ZIP). BBP really hammers CPU on all cores load.

I should have continued on UEFI 3202, but as 3203 had been released I wanted to use it. UEFI 3203 is same AGESA (1.2.0.3a) and same SMU FW as UEFI 3202. I can only assume 3203 only has fixes on ASUS side of FW.

I also started using a 6000C28 RAM profile with CO profile.

4. CoreCycler Y-Cruncher, cTDP 105W, PASS 3.5hrs.
5. CoreCycler AIDA64 CPU, CACHE, FPU, cTDP 105W, PASS 10.5hrs.
6. Standard Y-Cruncher stress test, 80C temp, cTDP 105W, PASS 6hrs.
7. RealBench stress mode, 16GB, 80C temp, cTDP 105W, PASS 1hr.
8. Kahru RAM Test, Cache Enabled, 80C temp, cTDP 105W, PASS 5k%.
9. AIDA64 CPU SHA3 80C temp, cTDP 105W, PASS 15min (seen mention of this in threads as a thing to do, significance no idea, had no issues running it).
10. Kahru RAM Test CACHE + FPU, 80C temp, cTDP 105W, PASS 21k% (link to ZIP)

Realbench was run as it hits CPU/RAM/GPU and I wanted to test more of the system in use.
Click to expand...
Here's some test benching on a CO profile which is now being stability tested on R7 9800X3D.
 
Last edited:
philo-sofa said:
Could I ask what you used to confirm stability?
Click to expand...
I used Core Cycler including Y cruncher and P95 all different sizes and AVX 2 etc. I also ran SHA3 too many times to
Count because this was showing instability for which I had to use “Who Crashed” to find the specific core that was causing the issue. I also ran OCCT and the normal YCruncher as well as realbench.

It took a long time to get everything tuned properly and @gupsterg has posted a great detailed way he did things, which would definitely save time over my method.
 
Last edited:
gupsterg said:
I've done several AM4 dual CCD chips using my method of CO per core and it's been very easy to get things right. I've tested also 2x R5 9600X, 1x R7 9600X and R7 9800X3D using my method and TBH things like AIDA64 SHA was walk in the park.

One such CPU test data in OP of my thread on OCN, below quote is from section CO profile stability testing.
Here's some test benching on a CO profile which is now being stability tested on R7 9800X3D.
Click to expand...
This is the most detailed and by far the best information I’ve seen. I really wish I had all of this when I was trying to find my way around.

I will be taking from what you have documented. This will save time over the way I did things when I receive my 9850x3d.

Hopefully you continue to post your findings in the future, and many thanks for this :)
 
McCarthy said:
This is the most detailed and by far the best information I’ve seen. I really wish I had all of this when I was trying to find my way around.

I will be taking from what you have documented. This will save time over the way I did things when I receive my 9850x3d.

Hopefully you continue to post your findings in the future, and many thanks for this :)
Click to expand...
No problem :) .

I'm going to do a little section in OP of my thread later today or day or so later. But here is a data ZIP highlighting something I saw comparing R7 9700X vs R7 9800X3D.

The R7 9800X3D seems like great power efficiency. The R7 9700X does need the cTDP 170W as would be riding cTDP 120W in CB23. The thing to also note, the 9800X3D has RAM OC, which adds to power consumption.

On average in CB23 the R7 9700X uses ~32W more when comparing each CPU at FMAX +200MHz. As a percentage ~21% to 27% more power than R7 9800X3D whichever way you wish to calculate it. The CO may seem less deep on 9800X3D, but if we look at voltage, it's using less.

I may redo testing at some point with 9700X at FMAX 5450MHz, and stable PBO CO it holds for that FMAX and both at same RAM setup.

Maybe I will see you on OCN in my thread :) .
 
gupsterg said:
No problem :) .

I'm going to do a little section in OP of my thread later today or day or so later. But here is a data ZIP highlighting something I saw comparing R7 9700X vs R7 9800X3D.

The R7 9800X3D seems like great power efficiency. The R7 9700X does need the cTDP 170W as would be riding cTDP 120W in CB23. The thing to also note, the 9800X3D has RAM OC, which adds to power consumption.

On average in CB23 the R7 9700X uses ~32W more when comparing each CPU at FMAX +200MHz. As a percentage ~21% to 27% more power than R7 9800X3D whichever way you wish to calculate it. The CO may seem less deep on 9800X3D, but if we look at voltage, it's using less.

I may redo testing at some point with 9700X at FMAX 5450MHz, and stable PBO CO it holds for that FMAX and both at same RAM setup.

Maybe I will see you on OCN in my thread :) .
Click to expand...
The efficiency was great on my 7950x3d, so I would hope the 9950xd will be able to perform better with similar efficiency.

I will be using your data for reference as soon as my new system is complete. Never found anything so detailed before like I say. You’ve pretty much logged everything which is a great tool for someone like me. :D

In terms of memory I’m not up to speed with timings etc and always require help on that side. It’s a rabbit hole I dare not go down. I prefer to take advice from the pro’s.

I will be using OCN surely when my system is up and running, some very knowledgeable people on there. If I can remember my log in, it’s been so long.
 
@McCarthy

No worries. I never owned 7000 series, just didn't attract me. There is also a DDR5 Tuning section, pretty much OP has become a culmination of all I'm seeing as tweaks from own testing and other owners from various threads/forums.
 
ntemples1 said:
How much faster will this be than my 5800x3d at 4k on a 4090?
Click to expand...

Depends on the title and more specifically the scenario within the game but I have seen up to a 30% increase in average perf and up to a 60% in 1% lows going by various benchmarks like this one.

4K section at 11:19 -

 
Last edited:
Well at 1080p the 9800x3d is 40-50% faster than the 5800x3d.

At 4k this mostly just translates to higher 1% and 0.1% lows for a smoother experience.

For me a lot of the new titles have a recommended CPU of a 7800x3d so I upgraded mine.
 
cputemps.jpg
cputemps1.jpg


So these was my CPU temps before and after.
So I managed to get my CPU temps down from doing a -20 curve opt, limiting the throtting and changing the PPT, these temps are from running Cinebench, by at least 20deg, left picture is what it was, right is latest
PC hasnt crashed yet, not sure if I should still mess about with the curve opt or not.
This is one a Noctua NH-D15
 
Burtoner said:
So these was my CPU temps before and after.
So I managed to get my CPU temps down from doing a -20 curve opt, limiting the throtting and changing the PPT, these temps are from running Cinebench, by at least 20deg, left picture is what it was, right is latest
PC hasnt crashed yet, not sure if I should still mess about with the curve opt or not.
This is one a Noctua NH-D15
Click to expand...
Big improvement but seems a bit warm for 105W - have you checked mount is good?
 
You must log in or register to reply here.
Back
Top Bottom