Board analysis and power supply
The board is based on the PG132 reference design, a 12-layer board with backdrill method, which has already been mentioned several times. If we take a look at the upper side of the board, we immediately recognize the division of the main power supply. NVVDD stands for the conventional core voltage, which is what we like to call GPU voltage. Here, the total number of acct individual phases (one less than the FE) is provided by a uP9511R from UPI Semiconductor on the back. This is a digital PWM controller with VID interface, compatible with NVIDIA’s Open VReg specification. The MP2888A can generate up to 8 phases, which is sufficient. These eight phases are each equipped with slightly cheaper 50-A powerstages AOZ5239QI from Alpha & Omega, which do not provide true MOSFET DCR, but do offer a temperature protection circuit and readout output.
The assembly of the second GPU power supply MSVDD, which I had called “Gedönsspannung” and which stands for “Miscellaneous”, is very similar. We find a total of five individually controlled phases generated by a second uP9511R. The five power days are the same as for NVVDD. So in the end there are 13 phases for the GPU alone and the different voltages for the GPU. This should also make power gating in the GPU a bit easier, although AMD uses a similar, if much smaller, outsourcing with VDDCI. The voltage range of both ranges is between 0.7 and a maximum of 1.2 Volt, whereby the maximum value can never be reached by the end user without special firmware and software. However, the minimum value is important, so there are technical limits for undervolting, but NVIDIA assigns them to everyone.
For the memory NVIDIA uses three phases, which also rely on the well-known power stages of Alpha & Omega. The power stages of the PCI Express voltage PEXVDD and the 1.8 Volt are located on the front panel, together with the corresponding coils. The capacitor six-pack on the bottom of the BGA contains a MLCC combination and 5 SP-CAPs, is sufficient.
For input smoothing, three coils with 1 µF each are used, behind which a shunt is located, over whose voltage drop the flowing currents are measured. On the right, we see the uP9511R PWM controller from UPI, which is used for all larger supply voltage ranges and therefore occurs several times.
The following two pictures show one of the Alpha & Omega AOZ5239QI power stages, which is used in all three major power circuits, and some of the many 220 mH coils.
We see below one of the memory modules from Micron. The total of ten 1 GB modules of the GDDR6X memory come from Micron. The memory runs at only 19Gb/s, although it could actually run at 21Gb/s. The extent to which NVIDIA wanted to prevent thermal problems here can of course only be speculated. On the right in the picture the monitoring chip for voltage and current monitoring from UPI.
Cooler and disassembly
Above the heatsink and the partially open structure behind the short board are a total of three PWM-controlled 9.2 cm fans with a rather aggressive rotor blade geometry. But also the maximum speed of the separately controlled fans of up to 3200 rpm is not exactly an expression of noble reserve. Let’s hope that it stays quieter than the card could if it had to or if the user could provoke it with a poorly ventilated case.
The large continuous heatsink cools the GPU, memory, coils of the large voltage converter series and one more MOSFET. A total of six large 8 mm heatpipes, soldered to the back of the nickel-plated heat sink, ensure that the waste heat is removed.
In addition, there is a further, independently operating cooler screwed to the slot panel for the six voltage converters in the direction of the panel. Here, air is used for cooling from above.
The backplate is also made of light metal, anodized matt black, and the manufacturer does not even thermally connect the voltage converters to the backplate, so that it remains purely optical and at least has a stabilizing effect on the PCB.
Kommentieren