VehiclesFashionRecipesBlogsHuntTravelsSportFunHandmadeITEducation
Mini-Games
x

x
zakruti.com » IT - Software » Gamers Nexus
Worse Than We Thought: Fixing the XFX THICC 5700 XT

Worse Than We Thought: Fixing the XFX THICC 5700 XT

FBTwitterReddit

video description

Rating: 4.0; Vote: 1
The XFX THICC RX 5700 XT was the worst 5700-series we looked at, primarily at fault for poor thermal performance and poorly configured fan curves. Ad: Buy Thermaltake's C360 Kit on Amazon Original review here: Buy a GN toolkit here: The XFX THICC, ironically, had the biggest issue with its curves. Fan curves, that is. In further testing and attempting to fix the XFX RX 5700 XT THICC (with AMD Navi, we discovered that the plastic embellishments along the PCB also caused issues, as did mounting pressure, thermal paste application, and backplate. Basically everything was causing thermal transfer problems, and that's something we were able to explore in this video of attempted fixes. Not all improved the card, but we did find the important ones that can.
Date: 2020-05-06

Comments and reviews: 10


The note you had on the reduction in workers available in Shenzen is fascinating. I recently was in the market for a 2080ti, and as an owner of the Gigabyte Aorus 1080ti (not the Xtreme, just the 11G, being overall very happy with it, I first looked at the direct 2080ti SKU that was available. The price was pretty awful to begin with, but aside from that, the entire design was a very clear step down from the preceding Aorus model. The heatsink design, backplate, and shroud were all much lower tier. From the extremely robust 1080ti design, the 2080ti Aorus looks about like the more mid-range 1070/1080 AIB variants in overall cooling design. PCB wise it still seems like a fairly premium product, but everything outward from there was a decliine. Ended up going with Zotac this time around, as the performance vs was better with what I had to choose from.
reply

I bought an XFX RX 590 Fatboy, had to return it to Amazon. After two minutes of gaming my whole system would crash. I adjusted the fan curve to have the fans run at 100% always, couldn't even have a conversation around my PC with it like that, and it still crashed! The coil whine was so bad too, and my PC sounded like a 18 wheelers engine when it was cooling down. Anyways I wanted a red system but after I got my refund I went and bought a 2060 Super, felt a bit sketch about AMD GPUs, still some issues but I think I have them all fixed now, driver related. I was going to get a 5700 but in Canada the aftermarket cards still aren't available last time I checked, happy with my 2060 regardless. Point is blows my mind how they sold that shit to me when it wouldn't function at all, and I was worried I was melting my motherboard lol
reply

Jsz Fkn x, How can they use STAINLESS STEEL as material for a component that is supposed to conduct heat? Don't they know the first thing about materials and their respective thermal conductivity? ( stainless has about a 15th of the condictivity of aluminium, they would have been 3xbetter off (from heat-perspective) using regular steel ) But of course they know all this, which makes the choice even more egregious since they made the bad choice knowingly. And I must confess that I still can't see how this even makes economic sense, as stainless is both generally speaking both more expensive, and harder to machine. So I can only conclude that there must be manufacturing conciderations that still makes stainless the preferable choice, for the PRODUCER.
reply

Why polish a turd is my question. I've spent nearly 3 months with the RX 5700 XT and I honestly believe that AMD's RDNA architecture is inherently flawed. Either that, or PCI-E 4. 0 is. I haven't had a single day of stability since installing the GPU, regardless of driver or Windows version. I returned it for a full refund and replaced it with a PCI-E 3. 0 graphics card (the RTX 2070 Super) and suddenly all of my devices are working as intended and I haven't had a single hard-lock, browser crash, game crash or anything of the sort.
reply

Dear Steve, I have a question about my Sapphire Vega 64 Nitro+ in regard to undervolting stability: I undervolted the p7 stage to 1. 075V 1662 MHz, the p6 is 1. 060 V 1537 MHz. HBM2 is 1. 000V 1075 MHz. Power limit was set to +25%. Now I am trying different benchmarks at the moment. I use Fire Strike, Time Spy, Valley, Heaven and Superposition. All benchmarks are running really good but Superposition. It always crashes at the end of the benchmark. Do you have any explanation for such behavior? Thank you in advance and kind regards
reply

I'm really glad you guys didn't give them an inch, especially after that garbage email the company sent you. Not only is that just bad PR, but that's just not something you do to a person who's literally marketing your card for you. And in all fairness tho, It's just a bad card. It seems XFX tried for a quick cash grab rather than an actual product, the fact that it's a custom card but preforms equal to a reference card is shameful at best. Sorry XFX, but just like your RMA rates, this card is GARBAGE.
reply

gamersnexus Hey Steve, does mounting pressure have direct correlation/relationship with torque on the screws when tightening? Meaning can we get the same amount of pressure between the each die and it s heatsink by using a torque wrench to tighten the screw down to a manufacturer specified torque? Or does that not apply because of the variances you mentioned (not all machined screws are equal, die fragility, springs differences, etc) and therefore must be done by feel?
reply

One argument For testing noise normalized is that it will be equal to real world conditions. If someone thinks the card is too damn noisy, they might reduce the fan speeds, and if they aren't bothered by noise, they might increase fan speeds to improve the thermals. This is the simplest way for a user to tweak their cards, and probably the first thing they will try, unless they go directly into overclocking, etc. All in all, this is a highly relevant test.
reply

Ok so I have been using this card for about 2 months now and it is amazing. I am wondering if there was some update they did because I look at a ton of negative reviews from months ago and its like I am in some alternate universe lol. I am seriously not understand what the issue is with this card. I honestly haven't had a single performance issue with this card and it runs at least 8 hours a day with 95 percent of my games maxed out. No overclock either.
reply

Steve, because of you and your team modifying and improving the design of the Coolermaster case to mesh, and Coolermaster's implementation response, I bought that case. I am proof that consumers watch and listen to your criticisms of these kinds of products. XFX you could stand to pay attention to the staff at Gamers Nexus, instead of giving flippant comments towards a community that cares and wants to see you improve. Kudos to you and your team Steve!
reply
Add a review, comment






Other channel videos