I just use ReLive. AVC codec, 30MB, 98Kbps sound quality.
-
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.
The Radeon RX 6900/6950 XT Owners Thread.
- Thread starter Kaapstad
- Start date
That’s just it - this doesn’t happen in practice mode. Proper weird. Only in actual games mate.
I spent some time playing Solo and Quads this morning. Although I did not experience any significant stuttering whilst actually playing/moving around, I did notice more variance in the frame times whilst playing Quads, particularly in the cut scenes and moments where you are not in control.
Will share the footage later.
I’ll capture some footage too later. I’ll show you the difference between Very Low and Normal. Huge 190+ frames in both, but stutter on Normal
@Duckwave
@LtMatt
Quick update swapping my ram seems to have fixed it on initial testing although not had long enough to say 100% but played 2 games of quads and landed in busy areas no micro stutter on even high detail and close to 10% fps boost I don't have time to fully test today however I'm fairly sure I would have had a few hitches on 2 games of quads
Old ram - 16gb 3600mhz 18 22 22 42
New ram - 32gb 3600mhz 14 16 16 36
Just enabled xmp profile 1 as profile 2 is greyed out for me for some reason
@LtMatt
Quick update swapping my ram seems to have fixed it on initial testing although not had long enough to say 100% but played 2 games of quads and landed in busy areas no micro stutter on even high detail and close to 10% fps boost I don't have time to fully test today however I'm fairly sure I would have had a few hitches on 2 games of quads
Old ram - 16gb 3600mhz 18 22 22 42
New ram - 32gb 3600mhz 14 16 16 36
Just enabled xmp profile 1 as profile 2 is greyed out for me for some reason
That's good news. You should be able to fine tune your secondary timings on the memory for extra performance too.
I saw a reddit post regarding warzone and different ram and had noticed my ram usage was often close to 16gb so thought I would give it a try. Thought it was abit of a longshot but fingers crossed its worked
Glad to hear it mate.
It's starting to seem like 32gb RAM is required for this game if you want to be playing this game on the higher settings. I may need to get myself another 16gb.
@LtMatt Any idea how I can get this to 3800?
Try changing these.Glad to hear it mate.
It's starting to seem like 32gb RAM is required for this game if you want to be playing this game on the higher settings. I may need to get myself another 16gb.
@LtMatt Any idea how I can get this to 3800?
FCLK 1900Mhz
DRAM Freq 3800Mhz
DRAM Voltage to 1.575v (you can lower this later once stability is confirmed, I have mine at 1.55v but my B Die is a bad bin so requires higher voltage you may get away with 1.5v-1.55v)
tRRDS - 4
tRRDL - 4
tFAW - 16
tWTRS - 4
tWTRL - 8
tRFC set all to 256
tRDRDSCL - 4
tWRWRSCL - 4
tRTP - 8 (try 6 later is you confirm all these are stable)
tRDRDSD - 4
tRDRDDD - 5
tWRWRSD - 6
tWRWRDD - 7
Here is what I am am currently using.
I recently dropped tRCDWR and tRP to 12, just testing stability now to see if they are stable.
Try changing these.
FCLK 1900Mhz
DRAM Freq 3800Mhz
DRAM Voltage to 1.575v (you can lower this later once stability is confirmed, I have mine at 1.55v but my B Die is a bad bin so requires higher voltage you may get away with 1.5v-1.55v)
tRRDS - 4
tRRDL - 4
tFAW - 16
tWTRS - 4
tWTRL - 8
tRFC set all to 256
tRDRDSCL - 4
tWRWRSCL - 4
tRTP - 8 (try 6 later is you confirm all these are stable)
tRDRDSD - 4
tRDRDDD - 5
tWRWRSD - 6
tWRWRDD - 7
Here is what I am am currently using.
I recently dropped tRCDWR and tRP to 12, just testing stability now to see if they are stable.
Thanks.
Aida64 has been running a memory stress test for 30 mins without failure. How long should I run it for?
I would use HCI Memtest, run 32 instances (one for each cpu thread) until you reach 1000%. If it passes that you should be stable. Divide your total memory up so each instance uses a chunk. Leave 1.5gb spare for the os.
Example
Last edited:
And here it is.
2160P Competitive Settings (shown at video start)
Solo
Quads
If your main purpose is using that particular app, might not be a bad idea.One area I miss my 3070 is video rendering, amd gpus don't seem to be as widely supported. Using Filmora and the cpu seems to be doing the grunt work.
Considering flogging it on as pc games lately are the drizzling *****
According to their recommended minimum hardware requirements, you need a Intel or Nvidia GPU for hardware acceleration on the GPU side.
One area I miss my 3070 is video rendering, amd gpus don't seem to be as widely supported. Using Filmora and the cpu seems to be doing the grunt work.
Considering flogging it on as pc games lately are the drizzling *****
Nvidia are always the cards to go for with Video work. There is the odd AMD card that is compatible (strangely enough Adobe only list a 5600xt as being GPU acceleration capable for their software of the few newer AMD cards listed)
And here it is.
2160P Competitive Settings (shown at video start)
Solo
Quads
If your main purpose is using that particular app, might not be a bad idea.
According to their recommended minimum hardware requirements, you need a Intel or Nvidia GPU for hardware acceleration on the GPU side.
Hey mate.
I tried an all core OC yesterday to get a comparison. Now in Cinebench R23 at 4.2hgz at a Voltage of 1.125 my temps get to 80c pretty quickly but never really exceeds 81-82c for the 10 minute duration. That I'm comfortable with.
But in Warzone, with the all core OC, my temps never exceed 65c and my performance is noticeably less smooth.
As a comparison, PBO temps are actually better in Cinebench compared to all core, but in WZ my PBO temps shoot up to 75-82c.
I don't get it?
Soldato
- Joined
- 8 Jan 2003
- Posts
- 3,951
- Location
- Scotland
Well Cinebench benchmarks your CPU whereas WZ will use both the CPU and GPU. Do you have any OC applied to your GPU?
Yes - I pretty much copied most of Matt's Powertools and Radeon software configuration, but the Radeon OC features only kick in for Warzone specifically.
TBH, it's the temps that I'm confused by. Why is an all core never getting to 70c within WZ but PBO is getting over 80? And yet the all core gives me a higher cinebench score?
For the most part, games only put a few cores (2-8) under any sort of load. In this scenario the CPU will automatically extract the best possible performance by increasing CPU voltage anywhere from 1.35v-1.475v which allows the highest possible CPU frequency, which results in better gaming performance. PBO/Curve optimiser can be enabled and tuned which can improve on this default behaviour further, depending on thermal and power headroom available.
An all core overclock might require less voltage based on your example, 1.125v for 4.2ghz, but it is also running at a much lower CPU frequency. This will give you worse gaming performance as those few cores mentioned above are now not running at their highest possible frequency.
The temperature metric will be reporting the highest CPU core temperature (out of all available cores), so under a gaming load with a higher voltage/frequency you should expect to see a higher peak temperature on one or more CPU cores. Remember, that metric is not reporting the temperature for all of the CPU cores, it just reports the highest. If you use HWINFO64, you will be able to monitor this yourself. Example below:
For the most part, games only put a few cores (2-8) under any sort of load. In this scenario the CPU will automatically extract the best possible performance by increasing CPU voltage anywhere from 1.35v-1.475v which allows the highest possible CPU frequency, which results in better gaming performance. PBO/Curve optimiser can be enabled and tuned which can improve on this default behaviour further, depending on thermal and power headroom available.
An all core overclock might require less voltage based on your example, 1.125v for 4.2ghz, but it is also running at a much lower CPU frequency. This will give you worse gaming performance as those few cores mentioned above are now not running at their highest possible frequency.
The temperature metric will be reporting the highest CPU core temperature (out of all available cores), so under a gaming load with a higher voltage/frequency you should expect to see a higher peak temperature on one or more CPU cores. Remember, that metric is not reporting the temperature for all of the CPU cores, it just reports the highest. If you use HWINFO64, you will be able to monitor this yourself. Example below:
Thanks for the explanation pal. So PBO is the way to go in your opinion?
Also, I have 32gb of RAM coming tomorrow. Let’s see if that solves the stutters.
If you want to overclock for gaming workloads, PBO and Curve Optimiser is the way to go.