The RMA team got back to me today and it seems I've been approved to send it back. I'll probably have to end up paying the $50 service charge for a new socket but we'll see. They could be rather blind (hopefully). I said above that I'm almost positive that installing my Hyper 212 evo is what did it. I had trouble getting it on when I made it originally but now I seem to have found a sweet spot where it takes very little effort!
The 212EVO is known for being a bitch to install, but it's worth it to have the best bang for buck cooler on the market.
Seriously, it sucked. I've overclocked to 4.9GHz once for testing and it keeps it at a nice and warm 85-90 (at about 85% load). Just putt-ing around/playing Beam/not stress testing. it didn't go over 70.
I've had the four beeps before, it's a problem with internal storage. My old computer had to be sent away about five times in two months and it wasn't getting fixed. My computer still had warranty at that stage so dell sent us a replacement for free. To put it simply those beeps are THE FOUR BEEPS OF YOUR COMPUTER'S DEATH.
Get Prime95 and run a SmallFTT, that will run your CPU over the edge with more use than it will ever see in real scenarios. During a normal blend test I get about 67°c and during a SmallFTT 85°c. The EVO's don't do well over 4.5GHz. That's generally the limit for that cooler unless you have a CPU that will take a high overclock at low voltages and has really good thermal conduction between the die and the heat spreader.
We've figured out the issue. You're a bit late to the party --- Post updated --- I used it during the testing. I didn't do a full-on stress test above 4.6GHz because we were hitting 90+
DONT run PRIME95 on these newer intels, you'll damage the damn chip. Doesn't need to be overclocked either. The fpu/ipc is so fast it heats up and melts the inside of the chip. Good way to burn your socket and chip. Do it with your PC all day, your not doing it on mine. DO NOT USE PRIME95 ON NEWER INTEL CPUS (3xxx series and up!). NO EXCEPTION! sheesh (of course, if you're so inclined, go ahead, prove me wrong, I hope you have good insurance on your house, and i'll see you when you've replaced everything)... Recent Intel CPU's have cheapo thermal paste underneath the IHS (metal lid covering cpu die), it doesn't transfer the heat away fast enough. They're not welded on like Sandy-Bridge and previous chips. It already creates an issue for even paltry clocks of mid 4ghz range, let alone using prime on it. There are many other system benchmarks, even the intel-burn-in test, that you can use. Even the best air coolers money can buy can't keep an i7 4790k @ 4.4ghz cool enough not to throttle (or in some cases, power off entirely) when running prime. Gets hotter, hotter, hotter, and POOF off goes the PC. Of course, you could de-lid the chip, but modern tech guys all know (or atleast most do) how to do this, but nor do they run prime on the PC. --Back on track, I hope you get your board fixed, even 60$ socket replacement fee isn't bad. Make sure to check all your ports and slots when you get it back - write your serial # down before you ship it out to make sure you get YOUR board back (this way, if not, you can check it out better!). Let us know how the RMA goes, too. --Cheers!
(offtopic) Nope, not being so at all, just making sure I got my point across quite sharply so people understand they could damage their PC doing so. I am very blunt (but still friendly) by nature, but on that note above, it's all over places like Guru3d and TomsHardware and such if you'd like to look about. That'd also be a good place to check out other benchmarking software (Guru3d has downloads sections, their hardware reviews also name a number of programs). It isn't just my opinion, or stuff floating about on forums... Hop into local PC shops and speak with the owner, or some of the lead techie guys in there and ask them about Prime95 and how it overheats recent Intel CPUs, especially if they're into gaming and benching stuff, they'll know. --Cheers P.S. Running and 8-car or 10~12-car AI enabled test for an hour or two on Beamng.drive is more than enough of a stress-test for most things.
In that case I'm sorry I misinterpreted you. I'll read up about that, I haven't really kept up to date with tech news since I quit my job. I love your maps by the way.
I used it during the testing. I didn't So, what about throttling? That is and should be in place to lower the speed if the chip is dangerously hot. Therefore, the danger is minimal, right?
While the throttling is in place, it can get so hot so fast that the temp readings don't respond fast enough. Also, sometimes board readings of temperatures may be slightly less perfect - sometimes by as much as 10~12C!!! This is also why there are temperature discrepancies in Windows VS bios. So while you think your below the limit, and the software may say so too, you could actually be sustaining damaging temps. It heats up the chip so fast running Prime 95, that users have actually damaged the system because of this. It's not a realistic test anyways, use more comprehensive tests on your computer recommended by hardware review sites. Anything over 80~85C is not good for the chips (on anything, actually, though at 80~85C you could sustain this for some time on some parts), it reduces the lifespan of the hardware. Some video cards used to be OK running 90+C for example, but if you compare the hours logged on those until death, VS the ones that don't go over 80C, there's a clear difference in premature failure rates with the hotter cards. As process nodes get smaller (wafer and trace sizes), the tolerance for higher voltages and heat goes DOWN, so it's only getting more picky as time goes on. So, the quickness of the 4xxx and up cpu's built-in FPU's, + the 3xxx series and above IHS (not your heatsink, the metal lid of the chip = IHS), in addition to the discrepancies in temperature reporting between bios and windows, can result in early chip and/or mobo death. 85+ C for your CPU is *VERY* bad for it. You really knock years off doing this. Read the product datasheets, as per-Intel. Throttling works, however, don't rely on it 100%. If your system is throttling at all, back off the clock speeds or volts some. Volts = heat. Cycles = heat (not quite as much but still). You should be able to still have speedstep enabled for energy savings so your machine doesn't cost 50~100$ a month to run! Intel burn test is a pretty good system stability tester, along with many other pieces of software. To cut to the chase, keep your cpu below 80C. Your cpu and the components on your motherboard in the immediate vicinity will live longer.
Good to hear!, Ty for the comment, for being disabled, I sure can sit in this chair and read-read-read*** all day (especially because my back is basically disintegrating and it hurts to heck in a handbag to move much). I've dug through almost every PC hardware forum (PC hardware would be my 'specialty' if you had to name one, that's it), over a good 20+ years. Overclocked my computer for the 1st time around 96~97, whenever Cyrix 6x86 was the new thing just to get quake running a tad smoother. Temps in the 70~76C range aren't bad, you could do those everyday for 10 years* and not have an issue. Sustaining 80~85C and above will actually be technically even more lethal for the nearby caps and vrm's to the cpu as they don't always get cooled as well with a tower-style cpu cooler [what you and i have] vs a downdraft orb-style cooler [stock]. This is why folks OC'ing often blow up those 3+1 and 4+1 AMD boards all the time, especially if the VRM's aren't heat-sinked (your on less-power-hungry intel, your board has heatsinks over those cpu-power delivery parts). Better boards have more VRMS/Chokes/Coils etc for power delivery, more hands for the job makes light (and cool) work. So yeah, there's a lot more to keeping the cpu cool than just the CPU itself. Atleast you didn't have one of the cheaper MSI boards that cause serious fires. Keep the fire limited to Beamng.drive cars @ the bottoms of cliffs and poorly made jumps masquerading as mountains/terrain I could go on all day, but it'd far outweigh the scope of this motherboard-ram-problem thread. So for sanity's sake and not typing a book I'll go onto other posts now! --Good day! *I've made PC's for family that frequently last 7~10 years or more, though I don't have/want my own business of it. ***Cue the Johnny-5 from short circut going INPUT INPUT INPUT IIIINNNNPPUUUUUUTTTTT at the book store! P.S. Hope you can keep your machine going and whatnot b/c we need more mappers here that make good stuff!
Always like these posts as I am emerging in the industry. My classes still push, and I mean shove it down your throat push, the older technology (IDE drives, AGP cards, learning how to install dial-up for God's sake) without really touching on the current stuff. Idk, maybe they think we already know it (most of us do) but just doesn't make sense when they haven't sold these components for quite some time. I watched LGR build a 486 DOS PC from scratch on YouTube and it looked awesome. Idk but I so weirdly love the satisfaction of typing in something and having it just do it, exactly like you wanted it. Even though a mouse does exactly this; could just be that I've grown up on GUIs. Back to boards, I'm thinking about going liquid cooled soon to get that extra push out of my chip. Ideally, I'm shooting for 4.8, but I won't be afraid of 5 if I can keep my temps under 80 on load but we'll see I guess. My board is of pretty good quality I feel. Things are solid. My Mobo from my older PC (older pre-built HP pos that I'll have to use in the interim on the RMA) is just so cheap. I'm really surprised that thing still runs after it ran at 100% load for like 6 years straight. Especially for being an AMD system. The thing is seriously a space heater lol. Since we've already hit several topics, I hope you find some happiness among your disability. Seems to be running rampant in my older family and friends; I see it bring people down real hard. My mom has some serious back issues as well, degenerative discs or something of the matter. Stay positive! Alpha coming soon on Deering
Update: Got a new motherboard in the mail today and everything is 100% working! 16GB of RAM, woo hoo! View attachment 166696
Hi Kyle. For future reference, you generally want to blur the Product ID. Actually, you should do that immediately.