Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Indiana_Krom
Jun 18, 2007
Net Slacker

Ak Gara posted:

Since I'm thinking of getting an Aquaero 6 I kinda regret buying my D5 Vario's. Sure I could set-and-forget but if I do need to change speed, (such as filling the loop) it'll be a pain, as the bottom of the pump would be quite close to a panel leaving only a small gap to try and turn the dial.

Plus, the EK D5 G2's cables are sleeved.
sleeved



Which is clearly better than hotdog cables. I say the EK D5 because I hear the Aquabus D5's can't be found any more?

I have the ekwb D5, it isn't the only PWM controlled D5 out there, but it is competent and as reasonably priced as anything ekwb sells. It is silent at 1800 RPM (~35% throttle) and has plenty of flow at that speed for a CPU+GPU+single radiator loop. 100% throttle is about 4500 RPM and is audible but not terrible as long as you don't vary the throttle any (its sound is most noticeable when it is changing speeds). If you are optimizing for noise it is definitely preferable to slow it down to <2000 RPM and just keep it there, also the pump will last longer at those lower throttles anyway. The temperature difference between 100% and 35% is about 1C on the GPU or CPU at power virus load levels, it has no impact on the coolant temp at any loading or the block temps at lower loads.

Adbot
ADBOT LOVES YOU

Indiana_Krom
Jun 18, 2007
Net Slacker

In your situation I wouldn't spend any money on a 7700k, if you are looking for more performance you won't find anything worthwhile in another quad core. Just dig in your heels and hold on with your 6600 until you can get more cores.

I had a 7700k and it was choking badly in newer games with just a vanilla GTX 1080.

Indiana_Krom
Jun 18, 2007
Net Slacker

Basically freesync (original) has one major advantage: it costs nothing. Which is precisely how much the overwhelming majority of display manufacturers spent on implementing it with predictably horrible results. AMD quickly wised up and launched freesync 2 which while still "free" has an actual certification process with required quality testing before a display can claim to support it. So a freesync 2 certified display is almost guaranteed to work fine with a nvidia gpu in gsync mode and the majority of them are probably passing gsync compatible certification as well.

A friend of mine has a reasonable freesync monitor (40-144 Hz range) and it works on his GTX 1080 card with no issues.

Indiana_Krom
Jun 18, 2007
Net Slacker

There is always a way to disable RGB on GPUs without installing software: find the RGB header under the shroud and physically disconnect it.

Unfortunately motherboards often have the RGB soldered directly to the board so that isn't an option for them, but I haven't encountered one yet that couldn't be turned off in BIOS.

Indiana_Krom
Jun 18, 2007
Net Slacker

Upgrading your PC for games that aren't even out yet is just not a good idea, nobody has any real idea what the game will ultimately need until it is out and even then it usually takes a several months for most games to get patches and gpu driver assists to really shake out the final performance. That being said, there is a safer order to upgrade components based on the expected lifespan of the component and how it impacts the rest of the system. I'd say in almost all cases you should start with the display, it is really a pace setter for basically everything else. It is why one of the most important questions in this thread is "if you are gaming what is your monitor resolution and refresh rate". So if I were you, I would definitely hit up the monitor/display thread and figure out what sort of target you are going to be aiming for once you get to actually upgrading the rest of the system. Then save up and come back when you have the games you want to play on the display you want to play them with. Odds are it will be a lot easier to make the decision then because you will have all the information you will need.

Indiana_Krom
Jun 18, 2007
Net Slacker

Rolo posted:

-I used to build PC’s but it’s been almost 20 years. For reference, my last one was a bleeding edge Athlon XP with a Radeon 9800. Has much changed with respect to assembly?

PC assembly has changed very little in the last 20 years, mainly fewer sharp edges and modular PSU cables. The mechanics of doing it haven't changed, but the products are more polished and the quality of life is higher.

Indiana_Krom
Jun 18, 2007
Net Slacker

PirateBob posted:

Thanks. I might go for an IronWolf.

How do you interpret the info from CrystalDiskInfo?

It says "reallocated sector count: Current 100, Worst 100, Raw data 0000000000"
That basically means 100% for that parameter (others can work in different ways). If your drive did have reallocated sectors, I would think about replacing it, but perhaps more important than the number itself is if they increase. If a drive keeps reallocating sectors over time it is a clear symptom of impending failure. But if a drive has some reallocated sectors and that number never changes over weeks and months of use, it is possible some part of the media went bad for whatever reason and it isn't going to spread anywhere else. But a healthy drive should always have zero.

If it shows "Good" then the drive's built in SMART monitor has not detected any anomalies, but if it says caution or warning you should immediately back up your data and replace the disk. If you mouse hover over a caution or warning indicator it will tell you what parameters are out of the threshold and show their values. To really get full usefulness out of it though, make it auto start with windows, set it to auto-refresh all drives every 1440 minutes (once a day) and leave it running in the background permanently. It will provide early warning via a sound and notification on most impending mechanical drive failures and 90% of the time will allow you to back up all the data then replace the drive yourself if you act quickly. It also tracks all the drive parameters over time and can graph changes, although there will always be cases where a drive just fails abruptly without warning and crystal disk info can't help there.

SAS is for Serial Attached SCSI, this is exclusively an enterprise class feature and does not work on consumer motherboards/controllers.

Indiana_Krom
Jun 18, 2007
Net Slacker

dads friend steve posted:

My understanding as a (bad) software engineer that doesnít typically work on operating system stuff is: the software you run is allowed to kick off as many threads as it wants. For example, web servers will typically maintain a pool of threads for serving incoming requests. The OS is responsible for scheduling those threads to actually be run on a core. So my guess is a chip having X cores but 2X threads allows an OS to schedule 2X threads to run simultaneously instead of X

Since I donít do game or desktop software dev, I couldnít tell you how many threads one of those applications typically spin up, but if itís less than or equal to your number of cores I do not suspect hyper threading will do you any good. Of course, if youíre running more than just one app at a time, then it might push your thread count back up to where itís useful
It isn't so much about running double the threads, at any given time a windows 10 desktop will have a couple thousand threads running regardless of HT, it is about cramming more stuff into the execution units to get the most performance out of them. Instead of a core being able to only execute a single thread at a time and only work on that single threads data at a time, it can now fill the gaps with work and data from a second thread if the first one stalls waiting for something that is too far away or just doesn't have the type of workload that will fill all the execution units. It is just incredibly hard to keep the full width of execution units on modern chips busy with only a single thread, hyperthreading helps to fill out resources that would otherwise go idle (while still running at the full clock speed and voltage of their adjacent busy units). This also means that each individual thread on a HT core actually executes slower because it is sharing those resources with other threads, just the combined output from all threads will be higher because the full width of resources will be utilized more of the time. That is also why some games perform better when HT is disabled (no more sharing of resources; when they run they get it all even if they can't always fill it up entirely).

Indiana_Krom
Jun 18, 2007
Net Slacker

Stickman posted:

Pretty much. If it's primarily for gaming, the difference between loading times on an NVMe SSD and a SATA SSD is usually tiny, and <10% at most, with the exception of Doom 2016's initial load at startup, which involves sequentially processing a very large file. An NVMe drive can be worth it if it's not much more expensive than good SATA drives, but I wouldn't pay much more unless you have a workload that can take advantage of it.

Yeah this, I have a fancy Samsung 970 Pro 1 TB and Doom 2016s ~5 GB sequential startup read is the only ordinary thing where I can tell the difference between it and a SATA drive. Instead of ~500 MB/sec = 10 seconds it goes at 1.8 GB/sec = 3 seconds. The rest of the startup and level loads still take just as much time though, so the overall difference is minor.

Indiana_Krom
Jun 18, 2007
Net Slacker

maniacripper posted:

Quick question as im about to put together a PC and as it stands I'll have a 3700x with a gtx970. As others in this thread have said, my card is going to be my bottleneck, but does the jump to a 2070 Super make sense at the moment?

My whole strategy my entire time as a pc enthusiast I've pretty much held to the rule that I would spend a maximum of $300 to upgrade my graphics card if I could double my performance of my current card. It would usually be every 3-4 years but mining has really put a damper on this upgrade cycle and I'm trying to decide if I should just drop the 500 bucks now and eat it or hold out for something that's a little better price/perfomance ratio?

Well, I don't see the price/performance ratio improving significantly in the next year, and there is nothing officially on the radar for a 2020 release from nvidia but that could change at any time.

If you have games that would benefit from a 2070 Super now, I don't think it is a particularly bad time to get one. I think it is highly unlikely nvidia is going to do nothing with 7nm in 2020, but I don't expect whatever they do with it to be a particularly compelling upgrade for an existing 2070 Super owner when it does happen. Just like as an owner of a 1080 I found myself pretty much completely disinterested in the 2000 series, not enough performance uplift for the price, especially since almost everything I play is already pushing triple digit FPS at the highest presets.

Indiana_Krom
Jun 18, 2007
Net Slacker

I have some huge tube of arctic silver that is like 20 years old and still obviously works and keeps CPUs cool. It probably helps that I keep the cap on the tube and the whole tube is in a sealed bag, but on the rare occasions when I pull an old system apart it is still greasy and flowing even after 3-6 years of being installed on heavily used systems. Its arctic silver III, which I don't know if you can even buy anymore.

Indiana_Krom
Jun 18, 2007
Net Slacker

Hyperlynx posted:

Hullo thread.

I have an Intel Core i5 6600K that I'm perfectly happy with, sitting in a Gigabyte B150M-D3H that I was perfectly happy with until the system started powering off for no reason.

I wanted to just replace the motherboard, but they don't make motherboards compatible with my CPU any more. So I guess I need to buy a new CPU too.

What's out there these days that would be about as good as my 6600K? I honestly don't give a rat's arse about higher performance than that, since that CPU was serving perfectly well.

I'm in Australia. My CPU cost me about $350AUD at the time, so I guess that's about my budget this time (I mean, ideally I'd like to pay less for the same performance as a four-year-old CPU, but hey).

I mean, I guess if someone happens to have a link to where I can buy a mobo that'll work with my CPU and doesn't cost >$400AUD then I wouldn't turn that down either, but I'm realistic about my chances.

Are you sure it is the motherboard and not the PSU? Random shutdowns can be attributed to either one, but the PSU is cheaper and easier to replace. And even if it is the motherboard, a 6600k era PSU is getting up there in years so if you are swapping the motherboard you would probably be wise to swap the PSU with it anyway.

Indiana_Krom
Jun 18, 2007
Net Slacker

Hyperlynx posted:

So, I ended up finding a motherboard I liked that's compatible with my CPU! I realised you can search partpicker by chipset. So, I got the list of compatible chipsets off of Intel's site, searched for only those, and it came up with https://www.itspot.com.au/asus-prim...rd-p750513.html for only $112!

Partpicker is incredibly good!


Yeah, I'd already replaced the PSU. I had an issue where two of my drives just suddenly died, decided the PSU was likely to blame and got a new Corsair. Then, about a month later, this issue. Which only leaves the motherboard as the culprit. I posted here about it, mostly as a "am I missing anything?", and then here when I decided to just go for it and get a new mobo.

Ah, looks like you covered all your bases then. Perhaps examine the old motherboard for bulging capacitors or some other sign of failure in its voltage regulation circuits, either that or there is a failing connection or developing short somewhere that is occasionally tripping the protection mechanisms of the board/PSU.

Actually that train of thought does bring up something worth investigating: Check your house wiring, make sure everything is grounded properly and the hots and neutrals are all in order. A short circuit on the motherboard or in the PSU would normally not damage drives because it would just take the "shortest" path to ground, but if your house wiring isn't grounding properly then you could have weird stuff like stray voltage in the motherboard/case grounding plane working its way through SATA cables and in to all sorts of places that would normally be protected. If your house wiring doesn't check out, call an electrician immediately and unplug any affected devices because it is a fire hazard.

Indiana_Krom
Jun 18, 2007
Net Slacker

MarsellusWallace posted:

Easy free option: check if your surge protector has a ground fault indicator. It might just straight up tell you if there's a problem. Easy enough to check with a multimeter jammed into an outlet as well, if you've got one available anyway.

Yeah this, or go to the nearest hardware store and there are probably outlet testers for dirt cheap that you can plug in to any outlet which will tell you if its all wired correctly. In the US these devices cost like $1 and require zero skill or understanding of electricity to use, just plug it in and compare the lights on it to what the sticker on the side says.

Indiana_Krom
Jun 18, 2007
Net Slacker

Moola posted:

hey guys please call me a dumbass

I bought an RTX 2070 and forgot to check if my PSU would be enough for it

It isn't, requires a 550 and I have 500

Please roast me and my dumb gay rear end

If you have enough PCIe plugs out of the PSU to wire up the GPU, you're fine, plug it and go. The kind of overclocking, tuning, and expensive cooling setup required to make a single RTX 2070 system draw close to 500W is well beyond the ability of a person concerned about a 10% difference in PSU specs.

Indiana_Krom
Jun 18, 2007
Net Slacker

I've had 3 PSU failures within my own systems over the last ~10 years, and in all 3 cases the rest of the machine survived without any lasting ill effects. One a always on PC (acting as a router) stopped responding and then failed to reboot/power on after being manually shut down, replaced the PSU and it was live again and is still operating today. The other two required multiple attempts to get powered on in the first place but seemed to operate with normally once they were actually booted, and again, one of them is still in service (other was an old C2D system that was retired in favor of an i7). Granted in all three of these machines I replaced the PSU within hours or even minutes of symptoms appearing because I always keep at least one new-in-original-packaging spare in the "tech supplies closet". But just the fact that I keep at a spare unused PSU in my closet should be a good indicator of my opinion about PSU reliability. At least I've only ever paid for one spare, all the other iterations have been free replacements on the original 10 year warranties. When one burns up, I put in the spare, then call in the warranty on the failed one and eventually a replacement comes in and becomes the new spare. The current spare is an EVGA Supernova G2 850w, its original stopped powering on the machine early last year about 3 years into its 10 year warranty and EVGA sent me a brand new still in the shrink wrap replacement. I've never had trouble with getting a PSU RMA approved either, there is never any resistance when you tell them "I swapped it for a spare PSU and the computer started working again.", the next email you get is the RMA approval.

I have seen OEM PSUs from the likes of gateway/dell/etc pretty much explode a motherboard or two before though, like nearly to the point of smoke coming from the chipset.

Indiana_Krom
Jun 18, 2007
Net Slacker

GruntyThrst posted:

Oh. That's kind of a bummer. Does explain why I didn't feel like g-sync made any difference whatsoever, though.

I've been running a high Hz gsync monitor right next to a plain old fixed 60 Hz monitor for like 5 years and have never encountered the issues those two are describing. Through multiple combinations of video cards, monitors and connections, and it has always just worked. No idea what they are doing wrong, because I've tried a lot of combinations and generally failed to get it to misbehave.

Like the rare occasions where I've gotten gsync to quit working were when I cloned an output of one of the monitors like it was going to a projector (so three monitors, 1 and 2+3). And it can also sort of stop working if you play a video or some other hardware accelerated thing on the 60 Hz display while trying to play a high Hz game on the primary, and even then it still works but is capped to 60 Hz.

Indiana_Krom
Jun 18, 2007
Net Slacker

Klyith posted:

The issues with multiple monitors are strictly with different refresh rates. Also the root of these problems tends to be with the DWM, the part of windows that draws all the 2d windows and desktop. DWM is perfectly fine with different resolutions, refresh rate is what it wasn't really designed to handle. This is why the weirdness is on both nvidia and AMD, though different flavors of weirdness with each.

Having different refresh rates can trip up DWM in various ways, mostly when playing video or playing a game, and especially doing both at the same time. If you like to play a game on screen 1 and a video on screen 2, it's best to just set both screens to 60hz and forget about VRR.

Also multiple refresh rates had a bug in the nvidia driver where it locker into high-power mode rather than the idle low-power mode it normally uses in 2d windows. This was annoying, but afaik has been fixed.

Variable Refresh Rate (aka G/FreeSync) compounds these problems because now it's not just out-of-sync refresh but variable. Using VRR really wants a game to be in exclusive fullscreen mode, not windowed or borderless windowed. If you like to alt-tab and have the game stay up, just don't use VRR. The reason for this is when in windowed mode the DWM is still involved, and DWM was not designed for VRR. Gsync allows games to use VRR while not fullscreened but it's very much a kludge.

Incorrect: in Windows 8+ there is no escape from DWM. It is always running, even when you have an app in "exclusive fullscreen" it is still running inside DWM. When an application requests exclusive fullscreen Windows says "Okay, here you go." and DWM promptly serves it up a nice borderless fullscreen just like everyone else gets. It is all faked/emulated in software, it has been this way for like a decade already:

https://docs.microsoft.com/en-us/wi...er-is-always-on

Indiana_Krom
Jun 18, 2007
Net Slacker

I have a 9900k with a GTX 1080 and I have a friend who has a 3700x but also a GTX 1080, we ran a relatively CPU intensive game benchmark on comically low resolution (800x600) in order to isolate the differences between our CPUs. Turns out at best I could pull about 160 FPS average in a given benchmark while he was pulling about 150. Our minimums were practically the same (110 vs 107), so literally unless you are pushing a low resolution across a very high Hz display with some very specific lightly threaded games there is no point at all to spending extra on a 9700k/9900k. In the average game the difference between these CPUs doesn't start to show up till well past 100 FPS even in the most CPU intensive titles, for the most part any GPU is going to be the limit once you switch to 4k and even if the GPU isn't the limit a 60 Hz display will be.

I'd say the only real advantage I had with the 9900k over my friends 3700x is it took him about two weekends worth of tweaking with a memory calculator to get his performance to measure up where mine was 1 minute of "set power limits to infinity and toggle XMP on".

Indiana_Krom
Jun 18, 2007
Net Slacker

I had a 7700k (4.5 all core delidded liquid metal on custom water cooling) with a GTX 1080 and in newer games it was choking hard on that CPU. 4/8 just doesn't cut it in well threaded titles which are growing more common. I switched to a 9900k and other than ubisoft quality games it gave me a significant bump in minimums and smoothed out the frame time variance in several newer titles. (Off the top of my head shadow of the tomb raider doubled its minimum framerate to the point where the minimum on the 9900k was higher than the average on the 7700k.)

Indiana_Krom
Jun 18, 2007
Net Slacker

HerniaFlange posted:

Okay, I now have a stupid beginner question for the modular power supply: My video card says I need to have both the 8 pin and 6 pin power inputs filled, which I can do, but I noticed the cable has connections for two separate 6+2 plugs. I have one going in the 8 pin jack, could I use the other for the 6 pin jack too, or would I be better to have a separate cable going in there from the supply? My main concern is that I don't know if the wattage for the output is split between the two outputs or if they share the same output section of the supply. I'm assuming they share it but I've also made mistakes with assumptions and power supplies in the past so I want to be sure.
One cable with both plugs should work fine. Odds are your video card doesn't even break 200 watts regularly and that configuration is good to 300.

Indiana_Krom
Jun 18, 2007
Net Slacker

On the newegg subject: One way to avoid some of their bad policies is to buy from them over ebay.

Indiana_Krom
Jun 18, 2007
Net Slacker

National Parks posted:

If I have a CPU cooler that is 158.5 mm high, and my case spec says it can fit a max height of 159 mm, is that playing with fire? Or is there a buffer space to make it work?
*checks math* 158.5 is less than 159, should work.

I mean I've stuck a cooler that was something like 165mm into a case with a 160mm max and it worked, I just had to bend the side panel out carefully when closing everything up. I replaced it with a cooler that actually fit within the confines of that case a couple days later, but it didn't bother anything other than looking stupid with a bulge in the side panel.

Indiana_Krom
Jun 18, 2007
Net Slacker

IuniusBrutus posted:

I'm building a PC inside of Sliger Cerberus, and I wouldn't mind using a 240mm AIO due to layout concerns.

The problem is...I hate running some lovely proprietary software to make it work. I've used NZXT and Corsair's solutions and wasn't a fan of either of them. Are there any companies that produce good AIO units that are purely motherboard controlled?

I ran a corsair AIO for years without ever plugging in the USB, just plugged the PWM fans straight into the CPU header. It works fine, even if it isn't quite as optimal as controlling the fans based on the coolant temperature which may not even be a feature of most AIOs anyway.

Indiana_Krom
Jun 18, 2007
Net Slacker

Klyith posted:

The games that currently demonstrate this are held back by 4 cores CPUs without hyperthreading. 4 core 8 thread CPUs are still fine for all games. The 7700k is good for above 100fps average in most games, and the ones that it can't max out 144 are difficult for any CPU to get 144 in. Check these two reviews, which include a 7700k and show it still doing just fine.

Whether that will change in the near future with new consoles is up in the air, but buying a hardware in advance because maybe the 4c/8t won't be good enough anymore is dumb.


Pairing a 2080ti with a 7700k is a bit unbalanced, but if the idea is big video card now, CPU & system upgrade later, it's defensible. (Though if anyone thinks that non-used 2080tis are gonna have big price drops any time soon, they'll probably be disappointed. That's not how things generally go these days, the OEMs tail off production so they don't have dead stock to firesale.)

Yeah, generally this. The 7700k remains competent in most titles but won't hold up forever, I had one personally and I can only name one game that really bottlenecked on it: shadow of the tomb raider. Upgrading to a 9900k brought my minimums up about 50% in that game, and pretty much only that game. That is a nearly stock 9900k (unlimited turbo time) vs a 4.5 all core turbo (also unlimited time) on the 7700k. Literally if I cranked it down to CPU benchmark levels to get rid of the GPU bottlenecks (800x600 with no AA) the 9900k straight up doubled the performance and was still waiting on the GTX 1080 occasionally. That game gobbles up threads like mad and does real work with them, but it is currently more the exception than the rule.

Indiana_Krom
Jun 18, 2007
Net Slacker

Mu Zeta posted:

I wouldn't be surprised if Windows goes free soon.
It pretty much is already if you think about it since it will run forever without being given a key or being activated and only disables some appearance related options.

Indiana_Krom
Jun 18, 2007
Net Slacker

apropos man posted:

Does anyone else use Argus Monitor for a no-frills experience at controlling your PC fans?

I really like the simplistic approach and it seems to have no limitations, in that you can control any fan and have a poo poo-ton of temperature sources to look at.

I've only used it about 3 or 4 weeks but I think I'm willing to pay the 9 bucks a year or whatever.

If I go for the 3 year value subscription, can I then register the same key on a different installation of Windows, when Windows inevitably fucks up to the point where reinstalling it is easier than trying to troubleshoot?
You do not need the 3 year subscription, all it covers is updates. If the current version works on your PC, buy the 1 year sub and it will remain fully functional even after your subscription ends, the only difference is you won't be able to update to new versions that come out after that.

The 3 year subscription only makes sense if you change your computer every year to brand new hardware that requires a new version to control.

I would have gotten Argus Monitor myself after Speedfan stopped updating, but instead I found out my motherboard has two thermristor headers, so I stuck a thermristor plug in my water cooling loop and slaved the fans/pump speed to the coolant temperature. Most effective fan control ever and it is a stupidly simple linear curve in BIOS.

Indiana_Krom fucked around with this message at 21:07 on Jun 30, 2020

Indiana_Krom
Jun 18, 2007
Net Slacker

PageMaster posted:

This is a pretty open-ended question but hoping I might get some ideas from some folks' favorites here: are there any good recommendations for well built and functional cases that don't look too "gamer?" Starting a new build and will be planning the rest of my components around the case, but I've found a lot of the recommendations from google top 10 lists are big on either glass panels LED lighting everywhere, or oddly shaped fronts/tops (meshify and some phanteks/coolermaster fronts/tops come to mind). My last case was a silverstone case that was just a silver steel box with tons of ventilation and ports and worked great. I'm not worried about price (not necessarily looking for lowest budget), but would like something sturdy and well-built with good air cooling potential that looks sharp.

edit: since it seems like almost every case has it now, clear panels/windows aren't a deal breaker, though I don't intend on going all-out with RGP/LEDs or color coordinated components.

Corsair Obsidian 750D is a giant monolith with a nice brushed metal front, very clean design with no unicorn vomit modes. But it is a full tower, so its pretty bulky, but that also makes it really easy and nice to work in.

Indiana_Krom
Jun 18, 2007
Net Slacker

LionEyez posted:

My current gaming pc is about 4 years old and I'm getting the feeling it's dying.

Has anyone ever experienced both a (secondary) internal harddrive and dvd drive suddenly disappearing (in Windows AND the bios?). The first time it happened I just rebooted my pc and there it was again. Ran some software checks but it came back completely clean. Note also that when the drives are detected, they work flawlessly.

A few weeks later it happened again. Sometimes when I check in the bios settings upon reboot, I will actually see the name of the hdd, but it will say '0 GB'. This also seems to mean that Windows can't see the drive at all. I assumed this could still be some kind of hdd hardware failure that the software checks can't detect, but I don't understand how this could happen to the DVD drive as well? This leads me to believe the problem is on the motherboard connection side or with the power maybe? Not sure how to diagnose either.

I've since disconnected the DVD drive and wanted to switch around SATA cables, but bundled cables are in my way and I'd rather not mess too much (I only half know what I'm doing whenever I open up my pc). At the moment the drive is working again and I've backed up all my data.

What can I do to determine where the fault could lie? I don't have spare SATA cables or another desktop to test the drive in, unfortunately.

Won't hurt to download a copy of Crystal Disk Info and get a readout of the SMART status of the drive. Load it up and it gives you a quick Good/Caution/Warning indicator for all connected drives.

Indiana_Krom
Jun 18, 2007
Net Slacker

tuyop posted:

What do you all do to vent these sick gaming rigs? The computer is in the bedroom and itís just like a 400w heater at load, right? And the bedroom is obviously hotter because of this.

Air conditioning isnít really an option. Should I build some elaborate duct out from the back of the PC to the window? Is this what liquid cooling is for? Like it sounds like the radiator just has to be somewhere else because ultimately the waste heat will end up in the bedroom, so the liquid cooling would also need some ducts.
400w is probably insufficient to raise the temperature of your bedroom enough to bother doing anything about it, and besides a high end gaming computer only generates 400w of heat when you are actively gaming on it. The rest of the time modern computers idle down to <100w, and many <50w. Crack the door and turn on a ceiling fan and it will probably be a significant challenge to even measure a difference.

Indiana_Krom
Jun 18, 2007
Net Slacker

PageMaster posted:

are there any good trustworthy PC component retailers online besides newegg or amazon to look into? Trying to get a Noctua cooler, but they're out of stock on amazon, and newegg sellers are charging 75-100 dollars for shipping. Microcenter is also out of stock, and I'd like to avoid going down to some of the lower listed options on google just yet..

https://www.bhphotovideo.com has some noctua coolers in stock, don't know if its the specific one you are looking for though. I've ordered some random stuff from them before (have a batch of like 4 cyperpower battery backups that I got on a really good sale for instance). They have been around long enough to earn some trust at least.

Indiana_Krom
Jun 18, 2007
Net Slacker

Subjunctive posted:

I've heard that first avail will be about 10 days later, but that is likely to be "lucky bot operators and scalpers" available, and not "leisurely stroll through amazon and one gets shipped that evening" available.
Yep, the GTX 1080 officially launched on 5/27/2016, I picked up mine on 8/5/2016 and IIRC the stock still hadn't really stabilized at the time. I didn't buy a RTX 20 card but I think it was similar 2-3 months after the official launch before they were readily available. If the 3000 series launches in September, honestly it wouldn't surprise me if its in to January or February before inventory really catches up due to covid complications.

Adbot
ADBOT LOVES YOU

Indiana_Krom
Jun 18, 2007
Net Slacker

albany academy posted:

i want massive amounts of storage in a windows desktop. some of the seagate stuff caught my eye on pcpartpicker https://pcpartpicker.com/product/cc...e-st12000vn0007
but idk anything about using NAS drives in a desktop. Would it just be simpler to buy this https://pcpartpicker.com/product/JW...rive-wd6003fzbx?

I have IronWolf Pro drives in my desktop, the only downside is the motors are a bit noisy. (Reason for going with the Pros was the longer warranty.)

End of the day, its a SATA hard drive that is good for big dumb sequential files.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply