Teardown: The cooler
The backplate is only attached with four screws and no seal. If you remove the screws, then the PCB is exposed. AMD has also dispensed with the obligatory spring cross and only uses four spring screws for the GPU. However, the stoppers and the spring strength should be checked again in a compressed state. There would certainly be better. The two thick pads cool the areas below the memory, which is rather pointless, as the IR analysis will show later. It would have been better to cool the area underneath the package instead of drawing the waste heat outwards to the memory.
The rest is quickly told. A massive copper heat sink is used instead of the usual vapor chambers and a total of three flattened and black coated heat pipes together with the cooler. AVC has outdone itself once again. By the way, the cooler hangs loosely in the cover and thus only adapts to the GPU independently of the rest. Therefore, the note about the spring screws, because in the ideal case with no curvature, the phase change pad could well tear during vibrations or by unhooking. The two 8 cm fans that provide the right wind are surprisingly quiet.
That this can happen I unfortunately noticed myself and had, thank goodness, only after all the gaming benchmarks. During the final tests hanging in the second case, there were suddenly much too high temperatures. Positive: the card immediately goes into emergency mode and shuts down the entire PC or throttles down gracefully, depending on the permanent load. But at least nothing breaks. Of course, you only notice it in the performance, because a 100 watt card hardly delivers any more.
Let’s take a look at the heatsink and we can see that the “paste” has come off. I would conclude that the burn-in is insufficient here, also due to my own many laboratory tests with phase change pads. Note the many “free” areas where bubbling probably never resulted in a true thermal bond. The cured material is extremely brittle, and you really have to guarantee a strong bond to keep it from tearing off. Despite all this, the surface is unacceptable, because everything that looks round here was not torn off during disassembly, but was always like this (bubbles and air pockets).
However, this could be solved relatively easily with shorter spring screws and, above all, better paste. Gain: about 2-3 degrees less GPU temperature and a significantly lower delta to the hotspot, which is unfortunately no longer displayed in the Wattman of the press driver (“transition temperature” is missing). But HWInfo still shows what they probably wanted to hide, for whatever reason. Well, you can’t have everything.
Yes, you can of course permanently complain and look for errors, but PC-Partner’s quality management didn’t cut a good figure here either and it seems as if they didn’t really learn anything from the mistakes with the vapor chamber. The lapse with the 6+2-pin connector and this with paste and fixing are things that can be found out in a maximum of 10 minutes and also have to be if you are on the payroll. The sufferer is then always the customer without own workshop and measuring possibilities.
- 1 - Introduction, technical data and technology
- 2 - Test system and igor'sLAB MIFCOM-PC
- 3 - Teardown: PCB, compnents and 8-Pin issues
- 4 - Teardown: Cooler and thermal grease
- 5 - Gaming performance Full-HD (1920 x 1080)
- 6 - Gaming performance WQHD (2560 x 1440)
- 7 - Latencies, DLSS vs. FSR
- 8 - Details: Power consumption and load balancing
- 9 - Transients, capping and PSU recommendation
- 10 - Temperatures, clock rate and thermal imaging
- 11 - Fan curves, noise and sound sample
- 12 - Summary and conclusion
74 Antworten
Kommentar
Lade neue Kommentare
Urgestein
1
Mitglied
Urgestein
Urgestein
Urgestein
1
1
Urgestein
1
Urgestein
Mitglied
Mitglied
Mitglied
Urgestein
Urgestein
Urgestein
Veteran
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →