• 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 7900 XT(X) Owners Thread.

wookiee87 said:
My RMA for my MBA has been accepted from ocuk, in the process to pack up the card and get it dropped off and then return to ocuk
Click to expand...
Once they received my MBA (DPD were quick too) I got the refund in 5days.
All in all very quick I thought.
They must have a lot of these coming back, so testing etc etc must be a nightmare.
 
NathG79 said:
Once they received my MBA (DPD were quick too) I got the refund in 5days.
All in all very quick I thought.
They must have a lot of these coming back, so testing etc etc must be a nightmare.
Click to expand...

yeah i've been waiting for almost a week, but finally have some closure, i have a red devil xtx anyway, soon as i relised my MBA was running hot i ordered a second xtx aib and put away my sapphire
 
gupsterg said:
The B die I have works well @ 3800MHz, will do tighter timings than Micron E. Only issue is on long usage ie 8hrs+ I can see an error or 2. Usually due to heat, it needs more cooling. Where as the Micron E even when my case is as it should be, will stay fine even in high ambient. So I choose the E die for daily use.
Click to expand...
b-die does seem to be sensitive to heat I had the same issue and it gradually got worse as the silicon degraded I had to replace it in the end
 
LtMatt said:
I never tested the basic preset, I tested slightly lower than that but it should be pretty similar.

Here's what my 4090 GameRock scored (now sold)

Here's what my 7900 XTX MBA card scored (now sold)

I've not rerun it on the Nitro as of yet.
Click to expand...
As these are now sold does that mean you've no longer got a 4090 and have "just" got the Nitro now?
 
sHo0sH said:
Sounds about right for stock performance.
Do you have the oc bios on (it is default I think) and have you overclocked / undervolt the card at all?
Click to expand...
I'm really happy with the set-and-forget performance, it's once I start tweaking anything I get worse results almost consistently.


LtMatt said:
at 3000Mhz and undervolt. Try 1.000mv-1.050mv. Lower more chance of instability but higher score. You want to find the lowest point you can go without crashing. Fans 100% always, m
Click to expand...

Thanks for the advice guys :) I did achieve a new personal high score of 16096 although it took me 60 runs to get there today. I don't know what to do to close the gap between 1600-1700. I wouldn't think there would be such a variance between clocks and scores. Also previously I wasn't really able to get under 1050 so I don't know why I could tonight, it might just be a bit colder. I can boost to 3ghz+ by creeping up the min clocks but It does nothing for performance.
  • 16096: 500-3350, 15% pl, 1035 vcore , vram 2700
  • CPU 5800X @ 4.8

  • 15084 pts: Default settings
  • 14798 pts: 1050 vCroe 500-2550 PL+15%, 2600 vram fast timings
  • 15343 pts: default settings +15% pl & 100% fan speed
  • 15533: 500-2980, 15% pl, 1125 vcore
  • 15393: 500-2980, 15% pl, 1100 vcore
  • 15491: 500-3200, 15% pl, 1100 vcore
  • 15483: 500-3000, 15% pl, 1070 vcore
  • 15330: 500-3100, 15% pl, 1070 vcore
  • 15444: 500-3200, 15% pl, 1080 vcore
  • 15517: 500-3250, 15% pl, 1090 vcore
  • 15255: 500-3250, 15% pl, 1095 vcore
  • 15534: 500-3250, 15% pl, 1090 vcore
  • 15529: 500-3250, 15% pl, 1085 vcore
  • 15463: 500-3300, 15% pl, 1085 vcore
  • 15589: 500-3250, 15% pl, 1090 vcore
  • 15606: 500-3250, 15% pl, 1090 vcore , vram fast timing 2500
  • 15499: 500-3250, 15% pl, 1090 vcore , vram fast timing 2550
  • 15563: 500-3250, 15% pl, 1090 vcore , vram fast timing 2600
  • 15194: 500-3250, 15% pl, 1090 vcore , vram fast timing 2650
  • 15162: 1000-3250, 15% pl, 1090 vcore , vram fast timing 2600
  • 15393: 1000-3250, 15% pl, 1090 vcore , vram fast timing 2500
  • 15574: 1500-3250, 15% pl, 1095 vcore , vram fast timing 2500
  • 15541: 2000-3250, 15% pl, 1090 vcore , vram fast timing 2500
  • 15784: 2500-3250, 15% pl, 1090 vcore , vram fast timing 2500
  • 15587: 2600-3250, 15% pl, 1090 vcore , vram fast timing 2500
  • 15652: 2600-3250, 15% pl, 1085 vcore , vram fast timing 2500
  • 15188: 2600-3250, 15% pl, 1080 vcore , vram fast timing 2500
  • 15744: 2500-3300, 15% pl, 1080 vcore , vram fast timing 2500
  • 15814: 2500-3300, 15% pl, 1075 vcore , vram fast timing 2500
  • 15138: 2550-3350, 15% pl, 1080 vcore , vram fast timing 2500
  • 15597: 2550-3350, 15% pl, 1080 vcore , vram fast timing 2500
  • 15772: 2550-3300, 15% pl, 1075 vcore , vram fast timing 2500
  • 15690: 2600-3300, 15% pl, 1075 vcore , vram fast timing 2500
  • 15673: 500-3000, 15% pl, 1060 vcore , vram 2600
  • 15017: 500-3000, 15% pl, 1050 vcore , vram 2600
  • 15719: 2500-3000, 15% pl, 1050 vcore , vram 2600
  • 15781: 2500-3000, 15% pl, 1050 vcore , vram 2700
  • 15718: 2500-3100, 15% pl, 1050 vcore , vram 2700
  • 15878: 2500-3300, 15% pl, 1050 vcore , vram 2700
  • 15785: 2500-3300, 15% pl, 1045 vcore , vram 2700
  • :15694 2500-3300, 15% pl, 1045 vcore , vram 2750
  • :15222: 2500-3300, 15% pl, 1040 vcore , vram 2750
  • :15636: 2500-3300, 15% pl, 1040 vcore , vram 2700
  • 15799: 2550-3300, 15% pl, 1035 vcore , vram 2700
  • 15786: 2600-3350, 15% pl, 1030 vcore , vram 2700
  • 15629: 2600-3350, 15% pl, 1025 vcore , vram 2700
  • 15554: 2650-3350, 15% pl, 1025 vcore , vram 2700
  • 15617: 2650-3350, 15% pl, 1020 vcore , vram 2700
  • Crash: 500-3000, 15% pl, 1020 vcore , vram 2700
  • Crash 500-3350, 15% pl, 1025 vcore , vram 2700
  • 16096: 500-3350, 15% pl, 1035 vcore , vram 2700
  • Crash: 500-3400, 15% pl, 1035 vcore , vram 2700
  • 15801: 2700-3350, 15% pl, 1035 vcore , vram 2700
  • 15771: 2700-3350, 15% pl, 1030 vcore , vram 2700
  • 15831: 2700-3250, 15% pl, 1030 vcore , vram 2700
  • 15921: 2700-3150, 15% pl, 1030 vcore , vram 2700
  • Crash: 2800-3150, 15% pl, 1030 vcore , vram 2700
  • Crash: 2750-3350, 15% pl, 1030 vcore , vram 2700
  • 16030: 2750-3300, 15% pl, 1030 vcore , vram 2700
  • 15836: 2750-3250, 15% pl, 1030 vcore , vram 2700
 
Last edited:
Vimes said:
With the size of the cooler would the fans necessarily have to be so loud to be effective when the card is being pushed ..?
Click to expand...
Exactly. I almost seems like they got the cheapest 3 fans they could find and stick it on there. I have a custom curve that never goes over 35%, and even then it's the loudest thing in my case, by a wide margin. I never saw the hot spot temp go over 80C in any heavy load. With max fans (and some ear protection) you could probably get some serious overclocking headroom.
 
sHo0sH said:
ASIC quality of 42.1% on my red Devil - that seems pretty low?
Click to expand...

Wouldn't worry too much.

The RD XTX I have, has higher than yours, 49.4%. But my MBA XTX is way lower, 36.1%.

The RD XTX on OC VB with +15%, I set same CLKs as a profile on Nitro XTX (54.0%). Which I believe is just shy of max my Nitro can do, stable based on loads I did. I have some bench stable runs which I will compare with RD, etc.

Bench results are very close, haven't worked out % as still doing some runs. I'd say max 1% diff if guessing.

I think if my MBA XTX (Vapor chamber fault) was using same cooler as RD it would be close performance.

As the power used and performance gained is not equal.

I am liking the RD more than Nitro. For starters save £125, £1175 vs £1300. Next I like the visual look of RD more than Nitro.

The build quality of Nitro is the highest I've ever seen. The heatsink frame for ram, etc is so chunky. But don't think temps for VRAM are any better than RD. The Nitro to me looks like a glowing submarine in my case.

Yeah the RD is as big. But dunno if it's because I see more of the heatsink, that it looks more GPU like. I prefer the RGB strip on RD vs bar. The look of the back plate and fan trim is more visually appealing than Nitro, like the brush alu. The eye and black plate logo looks great. The cuts in backplate with red glow again adds visual appeal.

Yeah luving the RD at present.
 
Last edited:
gupsterg said:
Wouldn't worry too much.

The RD XTX I have, has higher than yours, 49.4%. But my MBA XTX is way lower, 36.1%.

The RD XTX on OC VB with +15%, I set same CLKs as a profile on Nitro XTX (54.0%). Which I believe is just shy of max my Nitro can do, stable based on loads I did. I have some bench stable runs which I will compare with RD, etc.

Bench results are very close, haven't worked out % as still doing some runs. I'd say max 1% diff if guessing.

I think if my MBA XTX (Vapor chamber fault) was using same cooler as RD it would be close performance.

As the power used and performance gained is not equal.

I am liking the RD more than Nitro. For starters save £125, £1175 vs £1300. Next I like the visual look of RD more than Nitro.

The build quality of Nitro is the highest I've ever seen. The heatsink frame for ram, etc is so chunky. But don't think temps for VRAM are any better than RD. The Nitro to me looks like a glowing submarine in my case.

Yeah the RD is as big. But dunno if it's because I see more of the heatsink, that it looks more GPU like. I prefer the RGB strip on RD vs bar. The look of the back plate and fan trim is more visually appealing than Nitro, like the brush alu. The eye and black plate logo looks great. The cuts in backplate with red glow again adds visual appeal.

Yeah luving the RD at present.
Click to expand...
How much does the RD sag?
 
Guys I am thinking of maybe changing, now I am trying to decide between the Red Devil, Nitro, and the Tuf. From what Ive seen the Nitro has the highest power limit, followed by the Tuf, and then RD. However I saw that the Tuf pulled more frames on TechPowerUp reviews, so is it one of those cases where it is down to silicon lottery?
 
Nghtmare said:
Guys I am thinking of maybe changing, now I am trying to decide between the Red Devil, Nitro, and the Tuf. From what Ive seen the Nitro has the highest power limit, followed by the Tuf, and then RD. However I saw that the Tuf pulled more frames on TechPowerUp reviews, so is it one of those cases where it is down to silicon lottery?
Click to expand...
Tuf still not here in the US.
 
LtMatt said:
Max power limit, core at 3000Mhz and undervolt. Try 1.000mv-1.050mv. Lower more chance of instability but higher score. You want to find the lowest point you can go without crashing. Fans 100% always, memory try 2700Mhz no fast timings, then try 2800Mhz. At some point memory will cause score to degrade. Note, this won't be stable in games.
Click to expand...
Why no fast timings?
 
You must log in or register to reply here.
Back
Top Bottom