EVGA GeForce GTX 1080 FTW2 With iCX
The iCX cooler is not completely new. Rather, it has been compared to EVGA’s existing ACX thermal solution, found on the FTW and SC models. But there’s one innovation we’re sure that no other manufacturer shares: built-in temperature sensors with a matching micro-controller.
Unfortunately, we received the iCX-equipped card just on Wednesday afternoon, limiting the time we could spend time with it. Do we dig into the tech or run loads of gaming benchmarks? We went with the former, and are glad we did.
EVGA aimed its latest at enthusiasts who want more control over individual temperatures. Thus, the company’s focus is more on the technology used to enable this, along with its Precision XOC software, and less on the heat sink itself. EVGA also implemented some incremental performance improvements, such as more (and thicker) thermal pads. But we aren’t expecting any miracles compared to the ACX design.
Meet the GeForce GTX 1080 FTW2 with iCX
The new card weighs in at a hefty 37 ounces (1,050g), yet it’s still lighter than some competing 1080s. Measurements of 10.95 inches long, 4.92 inches tall, and 1.4 inches wide are about average for a two-slot card.
EVGA’s fan shroud consists of anthracite-colored plastic with aluminum highlights, illuminated by LEDs. A look at the bottom shows that EVGA again opts for vertically-oriented fins.
The top of the card is dominated by two eight-pin power connectors, a large panel with various LED indicators, and a back-lit logo.
At the end of the card, we see four 6mm heat pipes and two larger 8mm heat pipes. There is another, shorter 8mm heat pipe toward the front. The capillary action of these nickel-plated composite pipes works well in any orientation.
EVGA implemented a familiar array of display outputs, including one DVI-I, one HDMI, and three DisplayPort connectors.
Layout and Features
Initially, we didn’t see any differences between this card’s PCB and the ACX cooler-equipped version, but the design does diverge in places. To start, the newer model sports a two-part backplate with plenty of heat-conducting pads. This solves issues we uncovered in our Nvidia GeForce GTX 1080 Graphics Card Roundup, helping the plate actively contribute to the card’s cooling performance.
The GDDR5X modules come from Micron and are sold together with the GP104 processor to Nvidia’s board partners. Eight of these memory chips, operating at 1,251MHz, are connected to a 256-bit aggregate bus, yielding a theoretical maximum bandwidth of 320GB/s.
EVGA again employed a Texas Instruments INA3221 three-channel high-side current and bus voltage monitor. For additional protection, should something go seriously wrong, it also solders a fuse onto the PCB.
Power to the memory is provided by two phases controlled by an 81278 that doesn’t come from ON Semiconductor and is in a different package than we’re used to. This dual-phase synchronous buck controller facilitates phase interleaving and includes two low-dropout regulators. It also integrates gate drivers and the PWM VID interface.
A Siliconix ZF906 dual-channel MOSFET, which unifies the high- and low-side MOSFETs, is used instead of ON Semiconductor’s NTMFD4C85N.
The 5+2-phase implementation we covered when we reviewed EVGA’s GeForce GTX 1080 FTW Gaming ACX 3.0 made its return here, utilizing an ON Semiconductor NPC81274 PWM controller that offers many more control options than the µP9511P on Nvidia’s reference design.
EVGA somewhat deceptively claimed the GPU gets 10 power phases, but there are really only five, each of which is split into two separate converter circuits. (This isn’t a new trick by any means.) It does help improve the distribution of current to create a larger cooling area, though.
Furthermore, the shunt connection reduces the circuit's internal resistance. This is achieved with a NCP81162 current balancing phase doubler, which also contains the gate and power drivers.
For voltage regulation, one highly-integrated DG44E (instead of a NCP81382) is used per converter circuit, combining the high-side and low-side MOSFETs, as well as the Schottky diode, in a single chip.
Thanks to the doubling of converter circuits, the coils are significantly smaller. This can be quite an advantage because the current per circuit is smaller as well. As a result, conductors can be reduced in diameter while retaining the same inductance.
MORE: Best Graphics Cards