Artryx

Thanks for clarifying. I think I'll definitely be fine in that case. As I think I actually discovered something that was infact causing my CPU to run hotter than usual. The mobo I got for this build, I had flashed the latest BIOS to ensure it supported the 3D cpu. For some ungodly reason, MSI decided that upon updating the BIOS, it would automatically enable the CPU game boost preset in the BIOS menu. Having previously been on a gigabyte board for quite some time, I was not aware of this 'feature' so didn't even realize it was on until just now when I was doing a bios settings. I don't know enough about what changes this 'setting' makes, though have seen a lot of negative reviews online regarding bad voltage adjustments and whatnot. Boost mode is now turned off in the BIOS, already noticing a huge difference (before, If I tried running a prime95 stress test, the CPU temps would instantly jump from idle to 90c. Now it jumps to about 70-75c, and gradually rises the longer I run the test for, averaging at about 82-85c. Similar results during CInebench, and from what I can see, my multi-core benchmark actually scored higher with it turned off, so a win-win all across the board for me.

Cheers for the suggestion. I ran another test, after changing my fans from a custom curve I was running, to a flat 1600RPM (1550 on exhaust). CPU basically sat at 80c on average during cinebench that time. So I think maybe I should just consider upping the strength of my fan curve during higher temps (since the general levels were based on my expectations from older CPUs at the time) haha.

Hey there, thanks for that bit of information. I tend to clamp to working parts until they stop being just that, working. So it's definitely safe to say my thoughts/expectations were living a bit in the past, after getting a much more recent CPU. I'm going to take a look at my cooling setup (since my cooling config I typically go for with fan curves etc... was based heavily on my old CPU). But out of curiosity, if 90c is more normal with these kind of chips, what is the 'danger levels' seen as now? Would you say anything higher than 95c would be cause for concern instead?

Lets get some specs out of the way real quick: CPU: Ryzen 7 5800x3D Cooler: Corsair iCUE H150i ELITE CAPELLIX 75 CFM Liquid CPU Cooler Motherboard: MSI MAG X570S TOMAHAWK MAX WIFI GPU: MSI Gaming Z Trio Geforce RTX 3070 LHR 8GB Memory: Corsair Vengeance RGB Pro 32GB 3200MHz (4x8) PSU: Corsair RM850 850W 80+ Gold OS: Windows 10 Home (64 bit) Version 22H2 BIOS: AMI Version 1.40 (7D54v14) I have not done any overclocking to the build, apart from turning on XMP for my RAM. Everything else in the BIOS is the default settings. So I just finished building and setting up this new PC. I wanted to test the cooling of my new CPU, since I heard the 3D chip tends to run a little bit hotter than other AMD processors, and I'm always pretty stressed when building PCs that I'll mess up the thermal paste / cooler installation. For idle temps, all seems relative good/normal, at around 30-35c (with lows of 28c being shown on HWINFO). ICUE coolant sits at an idle temp of 26c, with the pump running at around 2250 RPM (My pump is on the variable speed setting from iCUE) I've ran Cinebench R23 Multi Core test to completion a few times, with hwinfo on my 2nd monitor to check the temps. At the beginning of the tests, the sensors show current temps of around 80. After 2-3 minutes in, the temps have risen above 85c. From here, the temps continue to slowly rise, upto a point of 90c where it remains until the end of the test. My icue sensors show the coolant temp is sat at 34c during this, with the pump RPM climbing to 2650. I have cleaned & reapplied the thermal compound (using Arctic MX-4), on the off chance I just done a bad job the first time around, but no change to any of the above results. I'm worried because I've always read that 90c is the danger zone you want to avoid. And the temp sensors on HWINFO are also turning red, starting from 85c (which only adds to my uncertainty). But at the same time, the CPU isn't thermal throttling at any point during this. So I have no clue if these temps are actually normal (or rather, bearable) for the 5800x3D and I'm just paranoid, or if it actually is too hot, and at risk of getting damaged / degrading. I just finished getting all the parts for this build (over the last 3 months). I don't have the money to go replacing parts now, so really can't afford to just "assume it's fine". I just need to know if this is cause for a concern, and hopefully catch the problem early, when I've just finished building the PC. Or if it's actually perfectly reasonable/fine for this specific CPU, and to be reassured that I'm not going to burn through a brand new CPU once I start actually gaming / running CPU intensive processes for longer periods of time. I'm sure this is probably a really dumb post, but I'd much rather be embarressed, and have a healthy new PC, then proud and have a brick by the end of the week,

Yup, looks like you and a few others hit the nail right on the head here. I lowered it down to 2933MHz and it smashed the memtest and showed no errors at all. My actual problem is (hopefully) now resolved. But just a quick question if you got a second: I've just got to hold out for a month until I get a better CPU. In the meantime, I'm sure this is a dumb question, and answer is simple, but I've made that assumption to many times when it comes to computer specs, and have been wrong way more times than I've been right. Would you recommend I just stick with the slightly lower speed (2933) and keep all 32 gigs, until I get the new CPU and can raise it again? Or would you say removing 2 sticks and utilizing the higher speed (3200) will outvalue the higher memory capacity? I know there are different reasons to go with higher RAM size, or higher RAM speeds, but just in general, like for someone who is probably gonna be playing various games, while having tons of webrowser tabs and software open?

No worries, apologies if I sounded a bit 'ratty' in my reply. It's been a stressful week so far (this ram issue is just the latest of pc troubles I've had to address so far). It's my bad for not clarifying in my post that I was already aware of the risks of 2 sets of 2, and that I was just eliminating other possibilities first. Appreciate the advice, and yea, if the suggestions mentioned on this post don't show any new results, then I think it's safe to say I have been extensive enough, and the next steps would be to try a set of 4 instead^^

Personally the problem is that every single enthusiast has a different opinion on the matter whenever I look this up. Anytime I have considered running 2 sets of 2 I have looked up what other people say, and you constantly find people either swearing by it, swearing against it, or saying "It very likely will work, but might not". So I appreciate that this could be the issue, and have considered it being a possibility, but this post is to check any other possibilies first, before I immediately default to (get rid of the sticks and buy a new set of 4x4)

Hi there, thanks for the speedy replies. I'll hop in the bios now and try lowering them to 2933, and then run another memtest. If it shows no issues, does that confirm it's just my CPU, and that I should be fine to bump them back up to 3200mhz when I upgrade to the 7 5800X? Or is it still an issue with the sticks themselves?

*Update - Problem (hopefully) resolved: Following the helpful advice mentioned from several users here, I have tried running all 4 sticks with XMP enabled, but at a lower speed (2933, instead of 3200) and so far every test I've run has shown no errors at all anymore. So the problem seems to be that I mistakingly assumed that because my CPU could handle 2 sticks at 1 speed, it could automatically handle 4 sticks pushed to the same level. This wasn't something that even crossed my mind in the realm of posibilities, and I feel pretty stupid because it sounds so painfully obvious to consider in retrospect. Will definitely be more cautious around this in future. Thank you everyone for the help! Hi folks, so recently I upgraded some parts in my computer, (more like, upgraded the entire computer, minus the CPU&MOBO which I will be upgrading next month). My current specs are now as follows: CPU: Ryzen 5 2600X Motherboard: Gigabyte AB350-Gaming (Just the original version, not the Gaming 3) GPU: MSI Gaming Trio Z RTX 3070 RAM: Corsair Vengeance RGB Pro 32GB (4x8gb) 3200MHz PSU: Corsair RM850 80+ Gold 850w Modular PSU My previous build used the current CPU, MOBO & half the RAM (Yes, I'm using 2 sets of 2x8gb. I originally had just the 16gb and bought the exact identical set, making sure I didn't just buy a set that looks the same). And I also intend on upgrading to a Ryzen 7 5800x, on a b550 board (currently looking at the MSI MAG B550 Tomahawk), but won't be able to do so until next month when I have some more money. I originally wasn't going to mix and match the old with the new, but my old GPU (GTX 1080) was dying, so using the new GPU right away was a given, the rest of the changes I've made so far were mostly case/fan related, with the only other big changes being the new psu (put it in, over my old EVGA 500W 80+ Bronze, just because adding the new GPU alone put the estimated wattage very close to my old PSUs limit, and didn't want to run any risks with it, and the doubling of my memory (from 16 to 32gb). Anyway, I fomatted my SSD and reset windows, just because I'd prefer to start clean, even if it's only for 30 days and then I'll be resetting again for the new mobo/cpu combo).Finished setting up all my drivers, no issues so far. Then after a few hours in to testing the waters of my new PC (opening up some games, configuring some graphic settings), I got hit with a BSOD. A few hours later, I got a 2nd BSOD, this one stating: "Stop Code: DRIVER_IRQL_NOT_LESS_OR_EQUAL | What Failed: ndis.sys" . After doing some research online over this crash, I was advised it was something to do with drivers, or the driver network. I was following guides and solutions online, and at some point, I was recommended to run a windows memory diagnostic. I done this, and it came back and told me I had mermory errors. Following on from this, I looked online and was advised perform a full memtest86. I set up the flash drive, booted memtest and started the test. Got hit with 130 errors in the first pass (Test 7: 125 errors | Test 8: 3 errors | Test 9: 2 errors). I immediately got to work on trying to diagnose if it was an individual stick / slot which was at fault, but found that all 4 sticks tested fine on all occasions, as did all 4 slots. The only time memtest would turn up any errors, it was when I had all 4 sticks connected (I hear this is not uncommon, but I have found there is slightly more to this in my situation). At some point, I decided to also try doing the tests, but with XMP disabled (sticks running at base 2133MHz), and in doing so, I discovered that errors were only occurring, when all 4 sticks were connected, specifically with xmp enabled. This I'm not sure if is still a common issue or not, I've been having a harder time trying to search up the issue/solutions. It was worth noting, that not long before switching to 4 sticks of RAM, I had recently updated my BIOS version. I have been running the f50d (a 2020 release from Gigabyte) up until now, as I'm aware it's not always wise or necessary to constantly be updating your BIOS if everything is already working. The only reason I decided to update it from f50d to the latest, f51g version, is because while checking for the latest drivers on Gigabytes site, I seen the BIOS list and it had a note next to a more recent update saying "Major vulnerabilities updates, customers are strongly encouraged to update to this release at the earliest.". So I figured, if the manufacturer is strongly recommending updating, it's probably fine to update. However, just to be sure it wasn't related to the BIOS version, I QFLASHed back to f50d out of concern that the new BIOS update was causing the XMP issues. After doing so, Windows memory diagnostics came back and said "no errors detected" but I was sceptical, and decided to go with another memtest to be safe, and yea I'm still detecting errors there. So at this point, I'm really not sure what else to try. Any advice you can give me here would be greatly appreciated. I don't know why XMP is only causing problems with 4 sticks (again, 4 sticks without XMP running fine, and the sticks are completely identical (I even compared of all the readings for each stick on CPU-Z). The XMP profile works perfectly well for all sticks involved when I cut back to just running 2 sticks at a time, but it just isn't playing ball with 4. The tim,ings and voltages are the same as what corsair list on their website for XMP, and the same when running 2 sticks vs 4 with XMP enabled (1.35voltage, timings: 16|18|18|36). It's not a faulty BIOS version, because I've tried both updating and downgrading. I'm pretty sure it's not a individual/set of faulty sticks, as they all individually work a treat, and all slots on the mobo appear to be functioning fine as well.

Will the coffee lake be significantly more expensive than the one I'm looking at? And would that be any better with the b250 mobo I listed or would I still want to get something better for it? Edit - I should also add, I won't be buying this till the end of the week so plenty of time to wait for new releases and rethink current choices

What would you recommend. I wasn't really sure what the deference between the b series and the z series was. Was originally looking at a z270 board but i'll admit, I turned to the b250 cause it looked better visual wise... Guess I might just have to bite the bullet

Alright everyone, been planning for a new high end gaming pc for the past few weeks and thought I'd come here for a 2nd opinion now that I have my final draft idea. As I said in the title, I currently have a budget of about £1500 (I can go up to £1550 but it'd leave me quite tight after), however if I felt there was something I absolutely needed that pushed me above that budget, I could get the rest of the setup and then come october I'll have another £500 to spare that can go towards the item(s). My aim originally was just to upgrade some older parts in my current PC. I wanted to upgrade my memory from 16gb DDR3 to DDR4, which to do so I realised I needed a new mobo to support it. So then I thought I'd throw in a new processor too (currently using an i5 4460) just for the hell of it. Eventually I found myself spending hours looking at pc builds and parts and decided I would just fork out and try and get the top end of each key component and in result, build a new gaming PC that can handle a lot of current and future releases like PUBG, Destiny 2, Overwatch, CS:GO and other games I am a fan of at roughly the highest graphics while getting a smooth 60-80+fps performance, as well as knowing it would consist of high-end parts that will last a while before becoming outdated. I also want to upgrade to a 3 monitor setup, not necessarily for 3 monitor gaming but just for multitasking and reducing screen clutter. I'm currently just looking at 3x 21.5" 60Hz monitors as I already have 1 of them, and would rather spend the bulk of the budget on the PC itself than replacing my current monitor and spending even more for higher refresh rate or size. For Peripherals I'm gonna be keeping my current mouse and keyboard: The Razer Blackwidow X Chroma and the Razer Naga Hex V2. Both are still in good condition and I'm relatively comfortable with both. I'm planning on getting a Logitech g933 7.1 Wireless Hedset, a corsair mm800 RGB polaris mouse mat, and a Logitech c922 webcam for youtube and video calls. I'm going to be sticking with my current copy of Windows 10 for the OS. So here are the specs I'm currently looking at for the new build: Mobo - ASROCK B250 GAMING K4 Cpu - Intel I7 7700K 4.2Ghz QC 8MB Cache CPU Cooler - ARCTIC - Freezer 13 CO CFM Ball Bearing CPU Cooler RAM - 16gb G.Skill Trident Z series RGB DDR4 3000Mhz (2 x 8gb) - Eventually getting another set to total 32GB but that wont be anytime soon GPU - Gigabyte nVidia GeForce GTX 1080 G1 Gaming - Can't decide if it's worth the extra £200-300 for a 1080TI for my build SSD - SanDisk X210 250GB 2.5" - My current SSD with my OS installed on it, with plans to upgrade to the Crucial - MC300 1.1TB 2.5" a few months down the line. HDD - Western Digital Caviar Green 2tb 3.5" (5400RPM) - Seagate BarraCuda 1tb 3.5" (7200RPM) - Both of these are also my current HDDs, not really sure if I want to upgrade them any time soon as they aren't 'too old' and have a lot of my important data on them. PSU - EVGA 500W 80+ Bronze Certified ATX - My current PSU, multiple PSU calculators said it should be fine to cover this build, except if I get the 1080TI, then I need atleast a 600W? That's all the specs. For cases I'm looking at one on amazon called the Game Max Moonstone Full Tower Case which comes with 4x 120mm RGB fans. And should have another 2 slots on the top which I'll fit with 2x Corsair HD120 RGB fans for a total of 6 case fans + cpu cooler. So basically I'm looking for any comments on what people think of this list, whether there is something not compatible with my build that I don't know about, whether there is anything in that list that I can get an even later / better version for little price increase (less than an extra £100), and most importantly, whether the GTX 1080TI should be a major priority to me over the GTX 1080 G1 (and whether I'm right about needing a 600W PSU if I do). Also I'll just throw it out there but bonus points if any of you know of some really nice cases that are white with RGB/LED designs and RGB fans included for a better price (the moonstone case is £110). Really wanted to get a nice white case with a side window and cool look + features like the Cooler Master Stryker case but I felt it was a little to expensive, especially since it only seems to come with 1 default fan included which means it will cost me more upfront to get some immediate fans.