• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

XFX Radeon RX 5600 XT THICC II Pro

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
28,745 (3.75/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
The XFX RX 5600 XT THICC II is priced very reasonably at $290, just $10 above MSRP, yet comes with a decent overclock out of the box and a solid dual-slot, dual-fan cooler that delivers whisper-quiet noise levels. XFX has also listened to feedback and made various improvements to their heatsink.

Show full review
 
not even on there web page!
 
Just letting you know, W1z. The Gears 5 page has an error with the Vega 56. The 1440p is the 1080p and the 4k is the 1440p.
 
So you can't see VRM temp even with HWINFO64 ? How is good OnSemi NCP81022 controller ? other brands use IR35217 controller.
 
Last edited:
"Note how the memory cooling pads sit on a separate metal piece, which leads to higher, but perfectly safe memory temperatures. "

This assessment is incorrect. It actually provides lowers memory temps - as there is not direct heat saturation from the gpu die itself. A solid copper plate would be instantly transfering the heat from the die before the heatpipes would even get a chance move it. (I'm a mech eng)
 
"Note how the memory cooling pads sit on a separate metal piece, which leads to higher, but perfectly safe memory temperatures. "

This assessment is incorrect. It actually provides lowers memory temps - as there is not direct heat saturation from the gpu die itself. A solid copper plate would be instantly transfering the heat from the die before the heatpipes would even get a chance move it. (I'm a mech eng)
Check the temperature page, memory chips are 10°C hotter than the GPU itself
 
Check the temperature page, memory chips are 10°C hotter than the GPU itself

The memory temp diode reading could very well be from one of the two chip that do not have a heatpipe contacting directly underneath them. As the other 6 do.

It is not an average reading for all 8 of them, hence unless you are willing to place a probe on each and test, this reading should be taken with a grain of salt.

I can assure you that a solid plate would directly transfer a lot more thermal energy to all 8 of the chips, than the solution employed by XFX.
 
It is not an average reading for all 8 of them, hence unless you are willing to place a probe on each and test, this reading should be taken with a grain of salt.
Correct, it's the highest of all the memory sensors

I can assure you
Maybe you can help me understand.
1) Heat flows from warm to cold
2) How much heat depends on the thermal resistivity
3) Ambient air is the lowest temperature of the whole system
4) Due to airflow over heatsink, heatpipes are at lower temperature than GPU & memory
5) Reducing R between memory plate and the rest of the cooler lets more heat flow either way (depending on what is at lower temperature
6) My argument: tGPU<tMemory -> at lower R = even lower memory temperature, your argument tGPU>tMemory -> at lower R = even higher memory temperature

?
 
The thermal output of the GPU is in the region of about 140watts - compare that to the ~25watts for all of the GDDR6 modules combined - place them on the same copper plate and it is easy to see which way the excess thermal energy will flow.

By decoupling the gpu heat spreader from the memory plate - XFX allows the heatpipes to be the only way the thermal energy from either can flow. Even as is - the gpu is heating up the memory. If you allowed for even a more direct heat transfer(solid plate contacting both the gpu and memory) - those modules would be running even hotter.
 
So you can't see VRM temp even with HWINFO64 ? How is good OnSemi NCP81022 controller ? other brands use IR35217 controller.

XFX is known for kind of, chimping out on components. Usually these cards are locked, meaning you cant do alot with them. I remember i had a 270 XFX card; it woud'nt do the things any proper brand like Asus did, do. I remember also having a friend that coud'nt ramp up the fanspeed for reason with XFX cards.

I mean if i buy a card from Brand A i am supposed to flash it with Bios from Brand B. With XFX, you cant, i think.
 
The thermal output of the GPU is in the region of about 140watts - compare that to the ~25watts for all of the GDDR6 modules combined - place them on the same copper plate and it is easy to see which way the excess thermal energy will flow.
it will flow towards the lowest temperature, which is the heatpipe/heatsink ?

So you can't see VRM temp even with HWINFO64 ?
Correct

How is good OnSemi NCP81022 controller ? other brands use IR35217 controller.
I'd say it's perfectly fine for this purpose (5600 XT is not a high-end card), and all data points in the review confirm that. Actually idle power draw seems reduced by a few watts.
 
Use Afterburner or Igorslab.com Radeon Tweakers group for overclocking/card tuning.
 
Is this a ultra or a pro? I don’t see it for sale anywhere. Why 20.1.1 drivers?
 
Is this a ultra or a pro? I don’t see it for sale anywhere
bah .. I failed .. it is the Pro version, updated the review title and text

Why 20.1.1 drivers?
Currently rebenching all cards with new test system, new drivers, new games, almost halfway through the 2200 benchmark runs needed, one more week I guess
 
XFX is known for kind of, chimping out on components. Usually these cards are locked, meaning you cant do alot with them. I remember i had a 270 XFX card; it woud'nt do the things any proper brand like Asus did, do. I remember also having a friend that coud'nt ramp up the fanspeed for reason with XFX cards.

I mean if i buy a card from Brand A i am supposed to flash it with Bios from Brand B. With XFX, you cant, i think.

I was struggling to decide between the XFX Thicc and the PowerColor Red Dragon because of the nice modern and clean design and ended up picking the Dragon, I'm really glad I did that now seeing this review, mine is extremely quiet and with very decent temperatures.
 
bah .. I failed .. it is the Pro version, updated the review title and text


Currently rebenching all cards with new test system, new drivers, new games, almost halfway through the 2200 benchmark runs needed, one more week I guess
is this vbios different than what is posted on xfx website
 
is this vbios different than what is posted on xfx website
It should be the same, unless there's a newer version on the XFX site.
 
Back
Top