Cooling
The MSI SPATIUM M580 Frozr 4TB also has an adhesive seal that can only be removed with loss. If you want to keep your warranty, you should leave the sticker on and not unscrew your gaming rig. There is no need to do a pad mod, as the two pads are both in good condition. The heatsink is made of the usual light metal with pressed-in and ground DHT heatpipes (Direct Heat Touch). This is perfectly acceptable. Otherwise, the coolers of the M570 and the M580 are absolutely identical, which means I can save myself some redundancy in the comparison today.
The heatsink is also made of simple aluminum with a natural oxide coating. There is no need for nickel here
The three pressed-in heat pipes are made of nickel-plated copper, the nickel layer is approx. 1.5 µm thick, which is perfectly adequate.
Interestingly, the cooling fins are nickel-plated, with a layer thickness of between 3.5 and 4 µm, which is slightly thicker than the heatpipes.
Not much needs to be said about the pads used. They are ultra-soft and therefore practical and stick quite strongly on one side. The mixture of viscous silicone, corundum and zinc oxide flakes suggests a rather average pad, but it is perfectly adequate for this purpose. I just hope that it doesn’t oil out (“bleed”) over time.
Temperature behavior and power consumption
The measured temperatures in idle and light operation (e.g. while gaming or working in the office) in explicit, passive mode without an extra fan nearby are quite ok. It is important to note that the controller always gets hotter than the NAND, especially under real load. This is less likely to occur when gaming. Either you use workstation applications or you clone an SSD. Then it gets a little warmer, especially when used as a target medium (2 or 4 TB write). The temperatures in the case (Fractal Meshify) are even slightly lower than in the open setup, because the fans in the case still produce sufficient airflow even when idle without CPU load. In the open setup, it sometimes went up to around 64 °C after more than an hour of continuous load, but you have to provoke that first. This time, we logged with HWInfo64 and our own K-sensors.
The temperatures naturally result from the power consumption, which remains moderate in idle and only rises to up to 9 watts under medium load and just under 12 watts under full load. Peaks during cloning can reach almost 12.7 watts, but this is rather an exception.
35 Antworten
Kommentar
Lade neue Kommentare
Veteran
Mitglied
1
Mitglied
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Mitglied
Urgestein
Veteran
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →