GPU Buying Guide - (Read first post)

You constantly take one specific instance and try to generalize it, making inductive leaps about an entire brand.
Your counterargument—basically, "Person XY bought an XFX card and it died, therefore XFX sucks"—is flawed. You're suggesting that because a card costs around $1000, we should pay 3600 just because it's MSI and supposedly built better so it won't fail. That logic is insane. Go ahead and buy whatever you want and overpay for refurbished cards just because someone once had a bad experience, but suggesting that to other people on this forum who don't care about brand names? Not cool.

Just like crimsonstag84 said, nobody was attacking MSI here. We were talking about a specific situation where paying that much money made zero sense because a) it simply wasn't in the budget, and b) it's an older generation that pulls more power while offering half the VRAM (which, lo and behold, actually turned out to be useful).

In this case, besides being a great model, it’s also a fair price and a smart value, so it definitely deserves a recommendation. There isn't a better deal out there for that specific card. Regarding that other situation you keep bringing up, NISI right, just like everyone else was telling you. It was obvious, but you refused to see it (or couldn't), and you're still acting salty about it today. 🙂

EDIT: Looks like Timothy Kim60 beat me to it.

Timothy Kim60 said:So just because your brother’s card fried, nobody should ever touch XFX? Is that how this works?

Are you guaranteeing it'll last twice as long? Based on what, exactly? What if I go online and find some random guy whose brother had that exact GPU and it died on him too? Does that make the whole brand garbage?

Now we’re really getting into extreme territory. But honestly, it’s common knowledge that XFX makes cards where the quality is night and day compared to their power supplies. Anyone who knows anything will tell you that... 🙂

I’m sticking to my guns here—there are definitely companies out there that don't cut corners on components and don't try to "offload" cheap junk onto customers like some other brands do. I actually studied electrical engineering, FFS, I think I know a thing or two about electronic components...

Anyway, to wrap that thought up, I used to have an EVGA RTX 3060 for a while, and I was looking at getting an ASUS RTX 3060 instead. I can't even remember what possessed me to go with Gigabyte back then... but the point is, the ASUS hits VRM temperatures of 110°C within fifteen minutes. Anyone with half a brain knows a VRM is going to die after an hour of running at those temps, let alone longer... so yeah, that's why I passed on the ASUS. I've got an ASUS motherboard with the exact same issue—the VRM temps are absolutely insane. -.-

So you're telling me ASUS has a 100% return rate after just one hour of use?

wiredotter14 said:Are you seriously trying to tell me ASUS has a 100% return rate on cards after just an hour of use?

Look, if you’re lucky enough to have a VRM fry itself because the voltage was set way too high, yeah, you'll probably get your money back. From what I've dug up online—and trust me, I've spent months falling down this rabbit hole—ASUS seems to be the one consistently struggling with those overheating VRMs. My old Gigabyte 280 had its own special kind of bullshit, though... the voltage was totally locked down and there weren't even any VRM sensors on the board. It doesn't take a genius to figure out what killed the card when it finally bit the dust. XD I'm telling you, I spent months obsessing over those 280s because I was totally done with them back then $550 ...

Jack Smith4 said:I've got an ASUS motherboard running into that exact same issue—the VRM temperatures are absolutely insane. -.-

Actually, I think those readings are just plain wrong. I’m using an ASUS board myself, and both HWMonitor and HWInfo are reporting temperatures hovering around 110 degrees. It usually happens when there isn't a specific sensor where one is expected, or sometimes the software just tries to aggregate data from multiple sensors incorrectly. There is no way those numbers are actually accurate.

Jonathan Mendoza17 said:That’s just a misreading. I’m running an ASUS motherboard myself and both HWMonitor and HWInfo show me hitting around 110 degrees too. It happens when there isn't a sensor where one is expected, or sometimes the software just aggregates data from multiple sensors incorrectly. There's no way that reading is actually accurate.

I don't know about that, man. I mean, why would XYZ reviews all be saying the exact same thing? I can't say for sure... but yeah, I'm dealing with the same headache. Everything on my motherboard looks totally fine, except for the VRMs, which are freaking blistering hot.

EDIT:

Just so I'm not talking out of my ass here—I remember seeing a Gigabyte motherboard with this exact same VRM issue once. Some guy actually had to strap a dedicated fan directly onto it just to keep those components from melting... 😁

http://s144.photobucket.com/user/aci...f4c5a.jpg.html

There used to be a whole massive thread dedicated to that specific board, but it somehow just vanished into thin air off the internet. 😢

Jack Smith4 said:I’m not entirely sold on that theory—I mean, there's usually a reason why all those XYZ reviews report the exact same thing, right? I'm really not sure... though, to be fair, I'm running into the same issue myself. Everything on my motherboard seems perfectly fine, except for the VRM, which is running incredibly hot.

The best way to figure this out is pretty simple: check your data. If your minimum, maximum, and average temperatures are all identical, something is off. There is just no way it would instantly jump to 110 degrees—you'd at least see a lower reading for the first few seconds while it climbed.
I'm fairly confident you're looking at a faulty reading here. Honestly, I've even stumbled across similar comments online.
Are you pulling that data from those auxiliary sensors? That's the tricky part—where manufacturers actually place those sensors on different boards. It's a real question of whether your specific board even *has* a dedicated VRM sensor. High-end ROG ASUS boards are usually packed with them, but with the more budget-friendly models, you have to wonder if they even bother including them, how far away they are from the actual components they're supposed to monitor, and what exactly that software is actually picking up.

Jonathan Mendoza17 said:...

I don't know specifically when it comes to motherboards, but back when I picked up my Gigabyte card, some guy on the Bug forums was tearing down an ASUS RTX 3060, and naturally, he posted screenshots from HWMonitor. The first thing that jumps out at you? A freaking 110-degree VRM temp. Just insane...

Then there's that other thing I linked earlier—some thread about a Gigabyte board where the user basically went full DIY, setting up a manual 120mm fan just to keep the VRMs from melting. And he actually posted that online too. Honestly, I don't think any of this is a coincidence. There’s gotta be some kind of pattern here... something's definitely fishy about how they design these things.

Timothy Kim60 said:You can grab the MSI Gaming 8GB at a few local shops like Best Buy for $717. It’s an incredible deal for the US market and easily one of the best models out there.

That's exactly how I compared the price.

I actually used to run an ASUS Strix GTX 1080 myself, and honestly, I never ran into a single issue with it... 😵 so, yeah—that’s my experience with ASUS.

I'm living in Beach now, so Amazon is definitely my easiest bet. If I get lucky, it'll be here by tomorrow, and honestly, having 30 days to return it is a huge lifesaver.

Using this for World of Warcraft is probably Overkill Sent 😁

Sent from my CUBOT DINOSAUR via Reddit

Jack Smith4 said:Now we’re really pushing it.

Well, you're essentially pulling from that exact same extreme.

Jack Smith4 said:It’s common knowledge that when it comes to XFX, the build quality of their GPUs is worlds apart from the quality of their power supplies. Anyone who knows anything about them will tell you the exact same thing. 🙂

That’s a bit of a stretch. You're generalizing again. What about the XFX GTR?

Don't just be a corporate drone for the brands. Look at every product on its own merits—and obviously, keep an eye on that price tag.

Jack Smith4 said:It’s worth noting that the ASUS hits VRM temperatures of 110°C within just 15 minutes of use. Listen, anyone with even a shred of technical knowledge knows that once a VRM hits those kinds of temperatures, it’s game over. You might get an hour of stability out of it, but after that, the whole thing is going to fry. Anything beyond an hour is just asking for disaster..

That’s incorrect! 😁
A VRM temperature around 100°C is actually perfectly normal; most of them are rated to handle up to 125°C without any issues.

Zachary Nguyen2 said:I compared it against that price point.

To be fair, you’d probably save some cash that way too. Honestly, it just comes down to preference—both options have their own set of pros and cons. For a lot of people, having that third year of warranty coverage is the real clincher; it makes the whole thing much easier to flip later on.

silentfalcon622 said:I live in Chicago, so Amazon is definitely my go-to. With any luck, it'll arrive by tomorrow, and having that 30-day return window really gives me some peace of mind.

Using this setup just to play World of Warcraft is probably overkill. 😁

Sent from my CUBOT DINOSAUR via Reddit

That’s a different story if you aren't here in the States.

Timothy Kim60 said:Yeah, no. You're generalizing again. What about the XFX GTR?

Exactly this. When the XFX GTR first dropped, it was cheaper than an MSI Gaming card, lasted three years easily, and had the exact same performance. People loved it.

Timothy Kim60 said:You're just taking the opposite extreme here.

Not really. You're generalizing again. What about the XFX GTR?

Don't be a brand loyalist. Look at each product individually, especially when you consider the price.

Wrong. 😁
Around 100C is actually a normal temperature for VRMs. Most of them are rated up to 125°C...

To be fair, you’d be trying to save some money too. Both options have their pros and cons. For a lot of people, having that third year of warranty is the deciding factor, and it makes the card easier to resell later.

That's a different story if you aren't living in the States.

You'd save about $100, once you factor in shipping, but an extra year of warranty is worth more than $100.

Quick Update:

It wasn't the GPU after all—turns out it was the motherboard.
I swapped it over to the other PCIe slot or whatever, and so far, no more freezing.
Before that, I was getting zero signal on my monitor, so I had to tear the whole case apart just to reseat the card.
Naturally, Windows refused to boot right after that mess, but hey, I managed to get it working again.

Sent from my CUBOT DINOSAUR via Reddit

So where was it saved originally?

wiredotter14 said:So where was it plugged in originally?

The motherboard has two of those PCIe slots. I just yanked the card out of one and shoved it into the other.
Right now, it's running without any major drama.

Basically, you slot it into the top one (closer to the CPU)

That's because the first slot gives you full x16 speed, while the second one is likely just running at x8—which basically means your graphics card won't be able to hit its full potential...

It isn't even about the x8 bandwidth itself—since SLI and Crossfire function just fine at x8/x8—it's the fact that most motherboards offering a second PCIe x16 slot actually bottleneck you with only x4 throughput.

For instance, looking at current Z170 boards, only the high-end models around the $400 mark actually support CF and SLI properly by providing true x8/x8 lanes.