|
Look all the other guys are super jealous you're getting all the attention. thank god you went for noctua nt h1
|
#
?
May 13, 2014 23:49
|
|
|
|
| # ? Jun 2, 2024 22:05 |
|
|
Gelid Solutions GC Extreme thermal paste has come in at the top rank of most of the TIM roundups I've seen, only being consistently beaten by the CoolLaboratory metal products.
|
|
#
?
May 13, 2014 23:54
|
|
|
Alereon posted:Gelid Solutions GC Extreme thermal paste has come in at the top rank of most of the TIM roundups I've seen, only being consistently beaten by the CoolLaboratory metal products. Its non conductive to boot.
|
|
#
?
May 14, 2014 00:46
|
|
|
Is there any good reason to get a top tier mobo (like an ROG board) vs a "regular" overclocking board? I'm looking into Z97 options and it's becoming apparent to me again how little I really know. The problem is I don't know how much those extra overclocking options will matter to me. If I have access to vcore, vrin, cache voltage, LLC modes, etc, I think I would happy which basically they all do after a certain price point. As far as features and slots they basically all do what I need (SLI + 3-4 SATA ports). But if there was something real to gain from having access to the other options and learning about them I wouldn't mind doing so. Assuming the rest of the setup is pretty vanilla as far as cooling and stuff like that. The other fuzzy thing is quality of components and power delivery and at what point is it a waste of money. I don't really want to cheap out on power, but I simply don't know what to really look for. 
|
|
#
?
May 14, 2014 18:19
|
|
|
Ignoarints posted:Is there any good reason to get a top tier mobo (like an ROG board) vs a "regular" overclocking board? I'm looking into Z97 options and it's becoming apparent to me again how little I really know. No
|
|
#
?
May 14, 2014 18:23
|
|
|
Hmm, that is the preferable answer, sweet.
|
|
#
?
May 14, 2014 18:30
|
|
|
It mostly comes down to what color you want
|
|
#
?
May 14, 2014 20:34
|
|
|
Ignoarints posted:Is there any good reason to get a top tier mobo (like an ROG board) vs a "regular" overclocking board? I'm looking into Z97 options and it's becoming apparent to me again how little I really know. look at the non-overclocking features, like number of fan headers, etc.
|
|
#
?
May 14, 2014 23:20
|
|
|
Cool. For the most part, the motherboards with the higher tier VRM stuff all have what I need for features including fan headers. I have a pretty standard amount and I don't mind the quiet ones running off the PSU. That's what I needed to know, thanks guys
|
|
#
?
May 14, 2014 23:34
|
|
|
Do goons still prefer Intel LAN Controllers over others on boards? That's a non-overclocking quality I like to check, though I'm not sure it's a real thing to worry about.
|
|
#
?
May 15, 2014 03:25
|
|
|
Intel is preferable, but Realtek is fine for most uses. Qualcomm/KillerNIC/etc, on the other hand, is a dealbreaker.
|
|
#
?
May 15, 2014 03:36
|
|
|
Ignoarints posted:Is there any good reason to get a top tier mobo (like an ROG board) vs a "regular" overclocking board? I'm looking into Z97 options and it's becoming apparent to me again how little I really know. Sidesaddle Cavalry posted:Do goons still prefer Intel LAN Controllers over others on boards? That's a non-overclocking quality I like to check, though I'm not sure it's a real thing to worry about.
|
|
#
?
May 15, 2014 03:42
|
|
|
Alereon posted:If you're not buying a ROG board you're likely making a serious mistake. There's no point in spending $300+ on a crazy enthusiast board, but if you're going with something cheaper than an Asus ROG Maximus VII you need to strongly consider how it could be worth giving up those features for the limited money you're saving. I think it doesn't help that the ROG line itself is also trying to reach down into more mainstream price points like with the Hero and the Ranger. That, and how MSI and Gigabyte are trying to edge in on that marketing segment with GAMING GAMING GAMING GAMING LOOK IT'S A GAMING motherboards.
|
|
#
?
May 15, 2014 04:07
|
|
|
Alereon posted:Yes, thankfully we've finally reached the point where only the cheapest low-end boards don't use Intel Ethernet controllers. And Gigabyte and MSI's high-end boards.
|
|
#
?
May 15, 2014 04:10
|
|
|
I do want to mention that to be fair, there really is a huge gap in price nowadays between a lot of the ROG line now and its competitors. To take an example, in my goofy old Micro ATX segment the Maximus VII Gene is $220. Gigabyte's highest "Gaming" mATX board on Newegg is $135. Eighty-five dollars is not a small amount of money for quite a few posters, I'm sure. There is something to say for nipping and tucking on deals here and there looking for a happy medium in features for the price; when put together, you might find that a $700 system can do everything that you care for instead of a $1200 one on the same platform. And I just finished typing that without remembering that this is the Overclocking Megathread.
|
|
#
?
May 15, 2014 04:27
|
|
|
atomicthumbs posted:And Gigabyte and MSI's high-end boards. Sidesaddle Cavalry posted:I do want to mention that to be fair, there really is a huge gap in price nowadays between a lot of the ROG line now and its competitors. To take an example, in my goofy old Micro ATX segment the Maximus VII Gene is $220. Gigabyte's highest "Gaming" mATX board on Newegg is $135. Eighty-five dollars is not a small amount of money for quite a few posters, I'm sure. But yeah, at the end of the day, if you're posting in this thread I think you're the kind of person who'd probably appreciate the features and quality a ROG board offers. It's not just color and marketing, you are paying for real things that cost money and are totally worth it imho.
|
|
#
?
May 15, 2014 05:02
|
|
|
Do you believe the extra overclocking features of an ROG board is worth the time to learn? My main focus is quality of components really, and power delivery. And I'm fine dropping money on it but I also don't want to be spending money on just those overclocking extras. But if it's more than that, that's cool too edit: missed some stuff above, thanks for the info  price sure is going up sadly Ignoarints fucked around with this message at 22:33 on May 15, 2014 |
|
#
?
May 15, 2014 06:40
|
|
|
^ https://www.youtube.com/watch?v=1d43r6iVgsc Alereon posted:But yeah, at the end of the day, if you're posting in this thread I think you're the kind of person who'd probably appreciate the features and quality a ROG board offers. It's not just color and marketing, you are paying for real things that cost money and are totally worth it imho. Can you comment on the TUF series and how it's evolved over Z87/97 so far? I'm eyeing the Gryphon and it's kind of at an interesting place lodged between the ROG board and entry-level-quality boards, with a price point that matches MSI's strongest offering in that size, for what it's worth (not much).
|
|
#
?
May 15, 2014 07:39
|
|
|
Sidesaddle Cavalry posted:^ I recently built a system with the Z87 Gryphon, I'd say overall build quality is about the same as the Z87 ROG I used on a system next to it but I liked the Gryphon's colour scheme better. Never could get quite as high an overclock as the ROG board with the same chip, but considering the ROG cost about $100 more I'd say the Gryphon is good value for money. Unfortunately I didn't record numbers at the time.
|
|
#
?
May 16, 2014 01:05
|
|
|
Sidesaddle Cavalry posted:Can you comment on the TUF series and how it's evolved over Z87/97 so far? I'm eyeing the Gryphon and it's kind of at an interesting place lodged between the ROG board and entry-level-quality boards, with a price point that matches MSI's strongest offering in that size, for what it's worth (not much).
|
|
#
?
May 16, 2014 08:59
|
|
|
Alereon posted:Their high-end boards use Intel chipsets, but you are right that they switch to using lovely Killer chipsets happens pretty quickly below the $200 mark on MSI/Gigabyte. Whats the deal with DPC latency being a big thing; as far as I know it's almost entirely due to driver implementation and unless someone uses a particularly lovely driver it's not really going to matter who built the board.
|
|
#
?
May 16, 2014 09:10
|
|
|
Malcolm XML posted:Whats the deal with DPC latency being a big thing; as far as I know it's almost entirely due to driver implementation and unless someone uses a particularly lovely driver it's not really going to matter who built the board.
|
|
#
?
May 16, 2014 09:19
|
|
|
Malcolm XML posted:Whats the deal with DPC latency being a big thing; as far as I know it's almost entirely due to driver implementation and unless someone uses a particularly lovely driver it's not really going to matter who built the board. It's both
|
|
#
?
May 16, 2014 15:05
|
|
|
I'm finally trying to be not-lazy and figure out memory overclocking. I manually set it to 1866 mhz but it bumped my timings from 9-9-9-24 to 11-11-11-28 automatically. Should I be more focused on keeping the timings constant with the highest possible clock or is a little higher okay for timing if it means a higher clock? It allows me to manually set timings... although the wording and menus are somewhat intimidating There is an XMP drop down menu that gives me two options: Disabled and Profile 1. I can't figure out what this does. Because this is a Haswell chip, should I absolutely not mess with the ram voltage at all?
|
|
#
?
May 16, 2014 17:27
|
|
|
There's a good overview of the effects of RAM speeds vs latencies here. In short: MHz are far more important than timing latencies:  XMP is Intel's memory overclocking profile(s). You should enable it. I do not recommend raising DDR3 voltages on Haswell by much, if at all, though there are sticks that use 1.65v. They are not recommended in the OP of the build-your-PC thread. Welmu fucked around with this message at 12:23 on May 17, 2014 |
|
#
?
May 17, 2014 06:12
|
|
|
Welmu posted:XMP is Intel's memory overclocking profile(s). You should enable it. I do not recommend raising DDR3 voltages on Haswell by much, if at all, though there are sticks that use 1.65v. They are not recommended in the OP of the build-your-PC thread. future ghost fucked around with this message at 17:53 on May 17, 2014 |
|
#
?
May 17, 2014 17:50
|
|
|
Conversely, would running low-voltage (1.35v) DDR3 put less heat strain on a Devil's Canyon CPU?
|
|
#
?
May 17, 2014 17:59
|
|
|
Welmu posted:Conversely, would running low-voltage (1.35v) DDR3 put less heat strain on a Devil's Canyon CPU? future ghost fucked around with this message at 18:06 on May 17, 2014 |
|
#
?
May 17, 2014 18:04
|
|
|
Holy crap, nice, thanks for the chart. I'm glad this old rear end ram went up to 1866 so effortlessly I guess I'll work on pushing it a little higher
|
|
#
?
May 17, 2014 19:17
|
|
|
Did you know that Intel NUCs actually overclock pretty well? I built a system in a D54250WYKH with a Core i5 4250U, nominally a 15W TDP with boosting to 25W for 28 seconds. You can't overclock the CPU, but you can adjust both the base and boost TDPs, as well as the boost duration, and can overclock the GPU. First I confirmed that my fears about CPU performance were largely unfounded, at stock settings it could happily run Prime95 at 2.1-2.2Ghz after the boost period ran out. By bumping the TDP up to 18W I was able to prevent it from throttling under Prime95 at all, it sat at max Turbo speed of 2.3Ghz. By playing with Unigine Valley I found it required about 32W to run the GPU at stock 1000Mhz without any throttling, with 1-2W on the CPU to run the drivers. I didn't actually try overclocking the GPU but 1250MHz should be easy. The real question is how much noise you can tolerate. The fan runs from 3000-6000rpm, and it seems to have about four bands it switches between depending on temperature, the bottom two are quiet, the third is very noticeable but not objectionable, and the fourth (>5000rpm) is quite loud. I found that the system could sustain a TDP of up to 18W indefinitely without speeding up the fan into the highest band. I was able to do burst TDP of 32W for 32 seconds with the sustained TDP of 18W without entering the highest fan band either. Eventually I settled on 18W/28W/32 seconds (from 15W/25W/28 seconds) as a slight bump for improved performance under sustained load. Overall I was pretty drat impressed with the NUC, though I feel like it needs a bigger TDP CPU, like a regular 37W laptop i5, and an appropriate cooler. Two 2.3Ghz Haswell cores aren't bad, but there were definitely times such as installing updates when it was sitting there with all four threads pegged and was obviously slower than a cheap laptop with a Core i3. The fact that the CPU cores only used 18W max boost also made me think they could unlock some more Turbo bins since you're allowed up to 25W. I'm really glad I didn't go with the Core i3 4010U, going down to 1.7Ghz would really suck.
|
|
#
?
May 18, 2014 05:05
|
|
|
I finally got my new i5 rig bought, pieced together, and running!  It feels like a HUGE step up from my older Phenom II x4 system, the boot time alone was cut down by at least 10-15 seconds. I reused an Intel 160GB SSD I had Windows loaded on, and on the old AMD system, it was around 20-25 seconds to boot (this was after tweaking the startup items in the BIOS) and now it's literally sub-15 seconds from hitting the power button to booting up. I thought something was wrong when the Win7 logo didn't fully load at first, until it went to the login and I was all  Anyhow, current system is chugging along, here's what I ended up getting to try to keep a nice price/performance ratio: Gigabyte GA-Z87-HD3 mainboard i5-4670K CPU 8GB G-Skill ARES DDR3-1866 Cooler Master HAF 922 case Corsair H100 CPU cooler with 2 Scythe slipstream fans (68CFM each) Stuff reused from my other AMD rig: GeForce GTX 670 FTW 2GB 2 x 160GB Intel SSD 1 x 1TB Seagate Barracuda 7200RPM Asus Xonar DX sound card Antec TPQ 850W modular PSU So far the beast is chugging along happily at 4.2Ghz, running a Prime stress test right now to test stability and cores haven't gone above 60*C at all. Surprisingly, the de-lidding of the IHS wasn't as bad as I thought it'd be...I don't have a vice or other means so I went with the tougher razor blade method and being really, really, REALLY careful not to damage the core or cut myself. Cleaned off the old glue on the IHS and CPU wafer, put a little line of AS5 on the die itself and re-seated the IHS, and left the stock thermal pad on the H100 block since I didn't feel like cleaning and reapplying AS5 to that. Got some nice deals off Newegg, the price of the CPU had dropped $20 right before I bought it, and had some promo codes for the RAM and mainboard, along with the H100 being a refurb unit that cost $55 and thankfully had no issues at all. Very pleased thus far, glad I made the move and it didn't completely break the bank 
BOOTY-ADE fucked around with this message at 23:48 on May 18, 2014 |
|
#
?
May 18, 2014 23:46
|
|
|
I have an older i5 on an asus thats been keeping me going fine for the past couple of years. however, I happened to get my hands on an lga 2011 xeon cpu, chucked it onto a gigabyte mobo and its pretty nice, despite making a shoddy flat plastic coupling between the chip and heatsink. However again, I am an asus fanboy, after getting a "used" rog lga 2011 board, and finding dubious pins on the cpu socket, I decided to double down with another Rog x79 board and an x79 pro, both as shady looking as the previous one, however, I have a good feeling they will turn out to be ok. To my question; can any UK goons recommend some place that can replace a cpu socket? None of the boards have cost me more than £60 so far, around the same price as the cpu, plus, I will probably be getting a similar xenon at the same price.
|
|
#
?
May 19, 2014 03:11
|
|
|
Are the pins just bent, or way more messed up than that? I can't imagine it would be very easy (nor inexpensive and quick) to replace a CPU socket itself...
|
|
#
?
May 19, 2014 05:00
|
|
|
The PC Part thread says Liquid cooling goes here so: I have a new system with an H55 from Corsair. Every now and then I hear what sounds like running water coming from the case. The first time I panic'ed and pulled the plug, but everything was still bone dry. Is that just natural noise for a closed loop system?
|
|
#
?
May 19, 2014 05:18
|
|
|
Xae posted:The PC Part thread says Liquid cooling goes here so: I imagine you're hearing the water return splashing in the radiator but I thought they'd be full or nearly full where that wouldn't happen .
|
|
#
?
May 19, 2014 06:11
|
|
|
Xae posted:The PC Part thread says Liquid cooling goes here so: How long does the running water sound last? does it effect the temperature of your cpu when it starts making a noise? Might be an air bubble trapped inside as they arent completely full, have you moved or shook the PC recently that could of allowed a bubble to get in the loop maybe thats traapped somewhere.
|
|
#
?
May 19, 2014 12:45
|
|
|
Ozz81 posted:Are the pins just bent, or way more messed up than that? I can't imagine it would be very easy (nor inexpensive and quick) to replace a CPU socket itself... Modern sockets basically have crazy miniscule pins at an angle, while I am happy to bodge a bunch of stuff, the connection between the cpu and motherboard is beyond my scope, I'd rather pay asus or some company to flip a new socket in than chance it. I do admire people who are quite happy to poke away at the most delicate parts of a £400 motherboard in the hope that it will not fry their £300-£1000 cpu. Xae posted:The PC Part thread says Liquid cooling goes here so: As its a closed loop system it has to be full at every point, if you hear anything other than the fan or the pump, you have air in there, The much more expensive modular kits with large reservoirs can take some air at the top of the res, yours is not designed like that, if you bought it new, I suggest you return it. not from danger to it vomiting coolant all over your motherboard, more from it not actually cooling your cpu. staberind fucked around with this message at 14:40 on May 19, 2014 |
|
#
?
May 19, 2014 14:19
|
|
|
staberind posted:Modern sockets basically have crazy miniscule pins at an angle, while I am happy to bodge a bunch of stuff, the connection between the cpu and motherboard is beyond my scope, I'd rather pay asus or some company to flip a new socket in than chance it. We are the same people who happily put a hammer to our cpus too but even so, if there is serious damage to the pins I wouldn't think it'd be feasible. Although I'd try. You can see the how "large" the contact points are so they just have to be in those squares. But I can see them just breaking off instead or bending too low permanently
|
|
#
?
May 19, 2014 14:26
|
|
|
Ignoarints posted:We are the same people who happily put a hammer to our cpus too The course of events was this: got a mobo from ebay for about 70 incl postage, received it, immediately took off CPU cover to asses, noticed some paler discoloration, like impacted plastic, saw that some pins were dodge. So I start a claim on ebay, stating that its not used but actually damaged, and go hunting for companies that actually repair computers, as opposed to bodging like I do. Seem to have hit a brick wall, as few companies do.
|
|
#
?
May 19, 2014 14:39
|
|
|
|
| # ? Jun 2, 2024 22:05 |
|
|
Yeah I agree if there are too many that can get unreasonable really fast. There was someone here recently who put that cpu cover back on the wrong way and hosed up a lot of pins as well. An honest mistake, I can't believe they are (or were) designed in a way where that's possible really.
|
|
#
?
May 19, 2014 14:43
|
|