Following the Sapphire RX 460 Nitro 4GB graphics card review that we posted, we decided to send the card through a tear-down, as we did with the RX 470, RX 480, GTX 1060, and GTX 1080 (links go to disassembly articles).
The RX 460 Nitro uses a custom PCB and shroud. This is a step away from the reference coolers provided by AMD for the RX 470 and RX 480 cards. The Nitro is easily dismantled, done by removing a handful of rear-side screws to release the shroud & cooler, then four more screws to release the heatsink.
Let's run through some photos and discussion of the PCB. Here's a video of the process, for more perspective:
AMD's fanfare surrounding CrossFire with the RX 480s demanded a test of the configuration, and we decided to run the architecturally similar RX 470s through the same ringer. We only have two RX 470s presently in the lab, and they're not the same card – but we'll talk about how that impacts testing in a moment. The cards used are the Sapphire RX 470 Platinum Edition ($200) and the MSI RX 470 Gaming X, tested mostly in DirectX 11 and OpenGL titles, with some DirectX 12 Explicit Multi-GPU testing toward the end.
The benchmark runs a performance analysis of two CrossFire RX 470s versus a single RX 470, single RX 480, CrossFire RX 480s, and the latest GTX cards (1070, 1060). We're looking at framerate and CrossFire power draw here, with no thermal testing. Read our RX 470 review for in-depth thermal and frequency stability analysis (and overclocking).
We liked the RX 470 well enough, which, for our site, is certainly considerable praise; we tend to stick just with the numbers and leave most of the decision-making to the reader, but the RX 470 did receive some additional analysis. As we stated in the review, the RX 470 makes good sense as a card priced around $180, but not more than that. That's the key point: Our entire analysis was written on the assumption of a $180 video card, presently fielded only by PowerColor and its Red Devil RX 470. Exceeding the $180 mark on a 4GB 470 immediately invalidates the card, as it enter competition with AMD's own RX 480 4GB model (see: 4GB vs. 8GB VRAM benchmark). Granted, it's still far enough away from the RX 480 8GB & GTX 1060 that the 470 may exist in some isolation. For now, anyway.
But as seems to be the trend with both nVidia and AMD for this generation of graphics cards, the RX 470 has some pricing that at times seems almost silly. Take, for instance, the $220 XFX RX 470 RS Black Edition True OC card: it's $20 more than a 4GB RX 480, it's clocked to where we overclocked on our RX 470, and it will perform about 3-5% slower in AVG FPS than the RX 480 4GB reference card. And let's not start on the seemingly irrelevant $240 8GB RX 470 Nitro+, effectively an RX 480 8GB card (even in clock-rate) with four fewer CUs, fewer TMUs (from 144 to 128), and slower memory – though it does have a better cooling solution, to Sapphire's point.
The EVGA Hybrid closed-loop liquid cooler has been the center of attention for a few of our DIY “Hybrid” mods, and for good reason: It was the best-performing CLC for GPUs that we tested last year, largely due to the protrusion in the coldplate and the dense microfins. The EVGA Hybrid cooler ($100) uses Asetek's supply, so we resurrected our old Corsair H100 (torn apart years ago) to compare a Generation 3 Asetek cooler to the new generation on EVGA's unit. You'll see that the setup is largely the same, though.
The EVGA Hybrid cooler we're tearing down today accompanies the GTX 980 Ti Hybrid, 970 Hybrid, and the forthcoming 1080 Hybrid. We managed to drop thermals of the GTX 1080 down to 18C (delta T over ambient) from ~57C by mounting an EVGA Hybrid cooler to the card, and even got an extra 100MHz out of our OC. Let's look at why that was the case.
We recently had a chance to speak with AMD's Robert Jameson about the Radeon Pro SSG, or “solid-state graphics,” that was announced earlier this week. This isn't a technical deep-dive by any means, but we did get some additional top-level information as to how the Radeon Pro SSG works. As a reminder, the SSG is targeted at professional production users and is not a gaming card; that said, the technology is interesting and new, and worth exploring for potential future implications.
Here's a quick run down of how this thing works.
The GTX 1060 Hybrid series has come to a close. This project encountered an unexpected speed bump, whereupon we inserted a copper shim (changing the stack to silicon > TIM > shim > TIM > coldplate) to bridge contact between the CLC and GPU. This obviously sacrifices some efficiency, as we're inserting two layers of ~6W/mK TIM between ~400W/mK copper, but it's still better than air cooling with a finned heatsink.
Our previous Hybrid projects (see: 1080, RX 480) axed the baseplate, thereby losing some VRAM and VRM cooling potential. For this project, we filed down the edges of the GPU socket to accommodate the protruding EVGA coldplate. This allowed us to keep the baseplate, granting better conduction to the VRAM and VRM. The blower fan is also still operating, but by removing the cover from the shroud (“window”), we're losing some pressure and air before it reaches the VRM. After speaking to a few AIB partners, we determined that the cooling was still sufficient for our purposes. An open air bench case fan was positioned to blast air into the “window” hole, keeping things a little cooler on average.
The GTX 1060 Hybrid tear-down went smoothly. We were able to remove all of the components with relative ease, look things over, and make a loose plan for part 2 – the build, which also seemed to go smoothly.
Until it didn't.
We were able to re-secure everything and, despite some very close clearance, even got the shroud back onto the card. Unfortunately, plugging it in revealed high idle temperatures, and a 30-second test led us to nearly 90C almost immediately. We terminated the test and cooled the card down, then re-evaluated the installation.
Our “Hybrid” mods aren't necessarily something we recommend for cards like the RX 480 and GTX 1060 – you're increasing cost of the card by 30% just to add a CLC – but the mods have routinely discovered throttle points. The GTX 1080 was our first Hybrid mod (one which we would actually recommend), and gave us an additional ~100MHz OC with perfectly flat clock-rate stability – something sorely lacking on the FE card. That's what we want, and will help further smooth over the 1% low and 0.1% low performance metrics (explained here).
Today, we embarked upon our journey to build a GTX 1060 “Hybrid” card. This is a DIY approach to liquid cooling the GTX 1060, and aims to stabilize the clock-rate over time to eliminate spurious frametime performance. We also hope to reduce thermals drastically enough that the overall noise levels will be reduced, presumably while maintaining a lower thermal value. This is what happened when we ran the same test on the RX 480 ($240) – it was trivial to run the radiator fan at 30% on the RX 480 “Hybrid” and keep lower thermals than stock.
Honestly, though, this GTX 1060 Hybrid endeavor is mostly within the realm of “because we want to.” It's not something you should necessarily do – that's an extra $50-$100 to throw a cooler on a card that's ~$250 to $300. Poor value. But we're doing it anyway, and hopefully we'll learn something about the performance and clock stability along the way.
This content is basically just a video, since we can't very well convey noise through words. Except maybe by YELLING with CAPS. We produced a similar type of video for the RX 480, basically comparing fan noise levels by recording them (using the same level of input each time on an X/Y H6N mic), then playing them back. Fans were tested at idle, 50%, and 100% for this comparison. The GTX 1060, RX 480, GTX 1070, and MSI GTX 1060 Gaming X are included in this video.
Just to be clear straight-away: This test was largely conducted under the context of “because we can.” For the full, in-depth GTX 1060 review, check this article. Also note that this test does not make use of the Scalable Link Interface, and so we're throwing scare quotes around “SLI” just for clarity. The GTX 1060s do not have SLI fingers and can only communicate via the PCIe bus, without a bridge, thereby demanding that applications support MDA (Multi-Display Adapter) or LDA Explicit (Linked Display Adapter) to actually leverage both cards. NVidia does not officially support dual GTX 1060s. This was just something we wanted to do. We also do not recommend purchasing two GTX 1060s for use in a single gaming system.
All that stated, this test pairs an MSI GTX 1060 Gaming X with the GTX 1060 Founders Edition card, then pits them vs. a single GTX 1060, 1080, 1070, and RX 480s (+ CF). This is mostly a curiosity and an experiment to learn, not a comprehensive benchmark or product review. Again, that's here.
Ashes supports explicit multi-GPU and has been coded by the developers to take advantage of this DirectX 12 functionality, which would also allow cross-brand video cards to be paired. We already tested that with the 970 and 390X. Testing was done at 1080p and 4K at high settings, mostly. The Multi-GPU toggle was checked for Dx12 testing. We've also listed the results as AVG ms frametimes, just for another means to convey information.