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
BurritoJustice
Oct 9, 2012

Paul MaudDib posted:

That's an NF-F12, which is an airflow-optimized fan (when you are trying to do traditional "cycle the air" designs). Those suck for radiators, so yeah, not surprising. You need the NF-P12, static pressure optimized model.

Noctuas are pretty fantastic given the noise. The PWM model is louder but moves even more air, but that's kinda not Noctua's niche.

That naming convention would make sense, but the F12 is noctua's static pressure optimised fan.

Outside of niches like half thickness fans, there are fans in every category that beat Noctua on noise/performance that are also usually cheaper. The F12 is probably the worst compared to their competition, they're poo poo radiator fans.

You're basically always better going with Phanteks MPs, eLoops, Vardars/GTs, Corsair MLs etc. The ML's are probably the best, the Maglev bearings actually work and work well.

Adbot
ADBOT LOVES YOU

Craptacular!
Jul 9, 2001

Fuck the DH

Yeah, I don't research my PC purchases too well except to make sure that they aren't disasters. I built this off a combination of MMO Champion's listings (gawd) and trying to ape a streamer. Now I'm just going fanboy for specific brands based on warranty or personal experiences. In this case, I liked the look of my new GPU, and since EVGA is just slapping their name on an Asetek cooler anyhow I might as well buy that for a matching set. I genuinely liked some of the concepts for their next case. They're one of those brands people seem to love on the customer service front, which is similar to how I feel about Corsair (this year I RMA'd my Corsair PSU and they gave me a free upgrade from a semi-modular gold to a new fully-modular platinum) except I think Corsair's cases are either ugly or overpriced.

I don't have any pets, but I live in a part of the desert where a fine silky dust, the kind that doesn't pick up very well until you practically scrape it off, seems to eventually settle over everything that doesn't exist anywhere with more humidity. In spite of that, I check my PC regularly but everything is pretty clean in there. I could go for air cooling, but am not inclined for it. Nonetheless might get a Core Frozr L. I think the colored plastic of Noctua's fans look ugly, I'd compare them to Be Quiet or something if I were in the market for any.

Again, I'm simple and buy stuff from brands that either have earned my attachment or just what I think looks cool that can achieve a 3.5 star average.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

BurritoJustice posted:

That naming convention would make sense, but the F12 is noctua's static pressure optimised fan.

Oh whoops.

Yeah looks like the F12 is the PWM optimized pressure fan. I guess the airflow fan is the S12. But Noctua still sells the P12 for radiators and other high-pressure applications:

quote:

The NF-P12 has been specifically developed for applications such as CPU coolers that demand superior pressure and airflow performance.

Fair enough then, what's the best nowadays for static pressure at reasonable noise? That's actually what I'm going for here... but you do need to figure on buying some decent fans. The stock ones suck.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Craptacular! posted:

I don't have any pets, but I live in a part of the desert where a fine silky dust, the kind that doesn't pick up very well until you practically scrape it off, seems to eventually settle over everything that doesn't exist anywhere with more humidity. In spite of that, I check my PC regularly but everything is pretty clean in there.

RIP your wallet. You totally have an excuse to do a bad-rear end full-passive build. Imagine this as the size of a decent stereo receiver.



You can put 90+90W into that (CPU+GPU). My Haswell 6-core is about 90W under load, and clocked down to 90W total card power (50% TDP ) my 1080 is around 1070 performance, ish. You might be able to do even better on a 1080 Ti clocked way down - although I dunno if you can actually set <50% power limit even on command line, and you might throttle given a >90W TDP.

That's actually one of the major niches of full-passive builds. Sand/marine environments, where there's no moving parts and the chassis is super easy to wipe down and clean.

That with Coffee Lake, or Ryzen2? Yes please. It also works with X99 Narrow ILM for that matter - the long-sought excuse to purchase Asrock's badass X99 board. :kheldragar:



(have not done this build myself but it seems possible on paper)

Paul MaudDib fucked around with this message at 04:15 on Sep 27, 2017

Craptacular!
Jul 9, 2001

Fuck the DH
I'll have to give the same response I give the Mormons: Interesting, but, uhh, I'll pass.

I don't have the money for exotic computing. The reason I'm not getting an 8770 non-K is because I'm hoping my finances will be better in a year. In the past I've had Lian-Li cases and passively cooled GPUs and I'm just not into that anymore.

Cygni
Nov 12, 2005

raring to post

JawnV6 posted:

Did you go back any further or is this just based on these two shots?

The shots mainly. Based on what I've read, both Intel and AMD have used synthesized designs heavily in the lower power markets, including the Atom and Cat cores respectively. They've also used it in other areas, like I/O controllers, graphics, and memory bus for a while. But until recently, both had been reluctant to roll it out in the area of performance CPU cores for various reasons.

AMD went first, using synthetic design in some of the later Bulldozer designs, and then Zen has the tell tales of synthetic work as well. This isnt to say there isnt hand work. In fact, the synthetic stuff being used by AMD and Intel seems more focused on optimizing hand laid designs rather than a from scratch synthetic design.

I'm sure we have some folks here who know more than me about this, but the smushy blobby scribbly marks inside Coffee Lake seem like a dead give away that synthesis was done on the design in some way.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Craptacular! posted:

I'll have to give the same response I give the Mormons: Interesting, but, uhh, I'll pass.

I don't have the money for exotic computing. The reason I'm not getting an 8770 non-K is because I'm hoping my finances will be better in a year. In the past I've had Lian-Li cases and passively cooled GPUs and I'm just not into that anymore.

I absolutely wouldn't overclock a rig like that any farther than it could go at stock voltages - so yeah you'd have to ask real hard whether the 8700K is worth it over the 8700 non-K (although even at stock voltage you will likely get all-core turbo that's several hundred MHz higher than officially specced). But the point of this is that it's emitting minimal heat for a given level of performance, and that it's conducting everything directly to the outside of the case (no internal heatsinks/fans to collect dust = easy to clean) . By all means, in those kind of environments passive cooling may not be the best idea at all, but regardless of how you dissipate it, you're going to have to think about minimizing how much heat you're emitting (and underclocking a fast GPU is the best way to do that). An ounce of prevention is worth a pound of cure. :frogbon:

Paul MaudDib fucked around with this message at 04:41 on Sep 27, 2017

JawnV6
Jul 4, 2004

So hot ...

Cygni posted:

The shots mainly.
Can you classify these two based on "blobby" or "hand drawn?"

BurritoJustice
Oct 9, 2012

Paul MaudDib posted:

Oh whoops.

Yeah looks like the F12 is the PWM optimized pressure fan. I guess the airflow fan is the S12. But Noctua still sells the P12 for radiators and other high-pressure applications:


Fair enough then, what's the best nowadays for static pressure at reasonable noise? That's actually what I'm going for here... but you do need to figure on buying some decent fans. The stock ones suck.

EK Vardars or Darkside Gentle Typhoons are best for radiator performance at a certain RPM (which is reasonably arbitrary). Corsair ML series are best for performance to noise, which means they're probably the best overrall as they're REALLY quiet and you can run them faster without hearing them. Phanteks makes the best heatsink (think tower air) fans, and their case fans are fantastic all-rounders that make their cases great value. The stock fans on most Phanteks cases beat out most of the market for general use fans and it's not really worth replacing them for that use. Depending on availability and price their MP series radiator fans are a great buy, they're right up there with GTs et al. and usually cheaper and more available.


Personally I'd just buy Corsair ML fans for every application but air towers. They're great as both case and radiator fans and they are absurdly loving quiet. They are subjectively pleasant sounding as well, very smooth with no ticking. They do have quality control issues and are a bit fragile however, it's not uncommon to get one that has a whining bearing. I bought them from a local shop so it was an easy swap for me, and they're easily my favourite fans out. I also own Vardars and almost every Phanteks fan produced, for reference.

Craptacular!
Jul 9, 2001

Fuck the DH

Paul MaudDib posted:

I absolutely wouldn't overclock a rig like that any farther than it could go at stock voltages - so yeah you'd have to ask real hard whether the 8700K is worth it over the 8700 non-K (although you will likely get all-core turbo that's several hundred MHz higher than officially specced). But the point of this is that it's emitting minimal heat for a given level of performance, and that it's conducting everything directly to the outside of the case (no internal heatsinks/fans to collect dust = easy to clean) . By all means, in those kind of environments passive cooling may not be the best idea at all, but regardless of how you dissipate it, you're going to have to think about minimizing how much heat you're emitting (and underclocking a fast GPU is the best way to do that). An ounce of prevention is worth a pound of cure. :frogbon:

For now I'm just going to keep my Ivy and see what I can reach at 1.2v or something. I saw some guy achieved 4.5ghz at 1.4 and everyone was telling him that that was too much voltage and he was killing his chip, so I figure at either stock voltage or just a little over I'll see if I can get what I want and that's that. Delidding is absolutely not an option as I can't afford to replace a CPU right now, and actually to be honest I could totally just get by without OC since it's not like I ever complain about this PC. I just decided to do it now that the generation is five years old and no more documentation about how to OC it is going to appear than what's already out there.

This would, literally, be the first time I even try to OC anything. I go back a long ways (386 -> K6-2 -> Coppermine P3 -> Prescott P4 -> Conroe C2D -> Ivy i7) and never bothered to until now. But I'd feel more comfortable replacing this nearly 5 year old Corsair pump with something new, and I also thought my case could be prettier even when it was new, so why not.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Craptacular! posted:

For now I'm just going to keep my Ivy and see what I can reach at 1.2v or something. I saw some guy achieved 4.5ghz at 1.4 and everyone was telling him that that was too much voltage and he was killing his chip, so I figure at either stock voltage or just a little over I'll see if I can get what I want and that's that. Delidding is absolutely not an option as I can't afford to replace a CPU right now, and actually to be honest I could totally just get by without OC since it's not like I ever complain about this PC. I just decided to do it now that the generation is five years old and no more documentation about how to OC it is going to appear than what's already out there.

This would, literally, be the first time I even try to OC anything. I go back a long ways (386 -> K6-2 -> Coppermine P3 -> Prescott P4 -> Conroe C2D -> Ivy i7) and never bothered to until now. But I'd feel more comfortable replacing this nearly 5 year old Corsair pump with something new, and I also thought my case could be prettier even when it was new, so why not.

My :effort: philosophy here, again, has been to just overclock however much I can get at stock voltage. Increase multiplier till it's unstable, back off, done. And I actually think the AVX offset might be the only thing holding me back from another step here, just haven't been assed enough to gently caress with it in a year.

Any core is rated to run at least its turbo frequency, at regular voltage. I'm doing all-core at the same voltage. Literally can't hurt the processor, and package power shows as 90W under a Handbrake encoding load on a "135W TDP" processor. Coffee Lake and Ryzen2 are both going to step that performance up a bit.

Just get the most performance you can get at whatever TDPs/cleaning intervals you find reasonable and be happy, is my advice.

Paul MaudDib fucked around with this message at 04:57 on Sep 27, 2017

Cygni
Nov 12, 2005

raring to post

JawnV6 posted:

Can you classify these two based on "blobby" or "hand drawn?"



Not really? I really only noticed because I had the die shots open side by side and despite Intel saying they have the same architecture, they don't look alike at all. Like I said, I'm sure we have some folks here who know more than me about this. Just making an observation.

Paul MaudDib
May 3, 2006

TEAM NVIDIA:
FORUM POLICE

Cygni posted:

Not really? I really only noticed because I had the die shots open side by side and despite Intel saying they have the same architecture, they don't look alike at all. Like I said, I'm sure we have some folks here who know more than me about this. Just making an observation.

The interesting bullet point from what I read was that the CPU area is now bigger than the iGPU area. Yeah, that's definitely panic right there. Even if the GPU partition ends up being more efficient per mm^2.

(note that I am not saying it's true but that would be a falsifiable test for the above)

There are also rumors that it clocks sliiiiightly higher on lower voltage - which might be possible with some minor revisions on certain logic cells. Again, not saying that's necessarily true, wait for benchmarks. Wouldn't be inconsistent with an improvement in Intel's layout technique either tho.

Paul MaudDib fucked around with this message at 05:07 on Sep 27, 2017

BIG HEADLINE
Jun 13, 2006

"Stand back, Ottawan ruffian, or face my lumens!"

BurritoJustice posted:

Personally I'd just buy Corsair ML fans for every application but air towers. They're great as both case and radiator fans and they are absurdly loving quiet. They are subjectively pleasant sounding as well, very smooth with no ticking. They do have quality control issues and are a bit fragile however, it's not uncommon to get one that has a whining bearing. I bought them from a local shop so it was an easy swap for me, and they're easily my favourite fans out. I also own Vardars and almost every Phanteks fan produced, for reference.

Corsair MLs also carry a five year warranty, which is kind of unheard of for a fan they have to assume might be getting run 24/7.

BurritoJustice
Oct 9, 2012

BIG HEADLINE posted:

Corsair MLs also carry a five year warranty, which is kind of unheard of for a fan they have to assume might be getting run 24/7.

Yeah I could waffle on about fans forever but I think the PC Building thread should just have a line in the OP saying "if you're adding fans just loving buy Corsair ML fans always". Atleast until someone else markets Maglev for consumers.

Combat Pretzel
Jun 23, 2004

No, seriously... what kurds?!

Cygni posted:

Not really? I really only noticed because I had the die shots open side by side and despite Intel saying they have the same architecture, they don't look alike at all. Like I said, I'm sure we have some folks here who know more than me about this. Just making an observation.
JawnV6 works or has worked at Intel, is trying to prove you wrong and is just being an rear end about it. Like always.

SwissArmyDruid
Feb 14, 2014

by sebmojo
http://wccftech.com/intel-arbitrarily-breaks-coffee-lake-compatibility-z270-force-users-buy-new-z370-motherboards/

WCCFT article, salt now so you're not salty later.

coke
Jul 12, 2009

Combat Pretzel posted:

JawnV6 works or has worked at Intel, is trying to prove you wrong and is just being an rear end about it. Like always.

Eh it's always good to see a good smackdown of baseless assumptions. As it's hard to say one is better than the other since manually laid out core blocks can be faster than the automatic layout sometimes.

The Lord Bude
May 23, 2007

ASK ME ABOUT MY SHITTY, BOUGIE INTERIOR DECORATING ADVICE

BurritoJustice posted:

Yeah I could waffle on about fans forever but I think the PC Building thread should just have a line in the OP saying "if you're adding fans just loving buy Corsair ML fans always". Atleast until someone else markets Maglev for consumers.

Now to convince them to make a 230mm one for my prodigy. I've always been a little disappointed at the overall noise level of my current PC.

craig588
Nov 19, 2005

by Nyc_Tattoo
I mostly only have 140mm fans in my case, but are the Corsair ML 120s better than the old Panaflo 120x38 fans? I have 1 120mm only slot. I think actually even if they are better for most people my specific use case would be better for a x38 fan, the space is right behind my heatsink where by being longer not only does it move more air, but it's closer to the heatsink to kind of act like a 3rd fan. It's even part of Noctuas recommendation to not put a 3rd puller fan on the D15 and let the case fan back there act as one.

GRINDCORE MEGGIDO
Feb 28, 1985


Jeez, I haven't used panaflo for 20 years.

craig588
Nov 19, 2005

by Nyc_Tattoo
Yeah, you can't get them anymore, but I have 6 of them left over from when they were still for sale for normal prices and not the 100 dollars people charge for them now. People are crazy about panaflos, if I didn't have a stock of them no way would I be getting them now.

Cygni
Nov 12, 2005

raring to post

Combat Pretzel posted:

JawnV6 works or has worked at Intel, is trying to prove you wrong and is just being an rear end about it. Like always.

I figured. Still doesn't explain why the die shots look so different in a way that looks telling. If it's not synth, would be cool to find out more.

e: maybe shots from different parts of the process?

Cygni fucked around with this message at 16:19 on Sep 27, 2017

JawnV6
Jul 4, 2004

So hot ...

Cygni posted:

I figured. Still doesn't explain why the die shots look so different in a way that looks telling. If it's not synth, would be cool to find out more.

e: maybe shots from different parts of the process?

Yeah, sorry, I normally swoop in to rip apart weird theories about CPU design process, but I was kinda curious if you'd picked up on a genuine methodology difference between the two big Intel CPU teams and grabbed one from each.

Die shots are mercilessly doctored. They're summations of different angles, lighting, and other post processing. That's not to say you couldn't get real design information out of them. One of the apocryphal stories is that AMD got their start by taking a high--quality die picture of a 386(maybe earlier?), blowing it up to be the size of their cafeteria, and having their architects walk around it to reverse engineer the functionality. There was a leaked picture of the Sandybridge die some years before it hit the market, it was damaging not just because it got out, but because it was pretty easy to suss out it was a 4 core part with integrated graphics, at the time Intel didn't know AMD's roadmap that far out. The ones you have here are marketing shots though, more aesthetic than a real map of the territory.

Regarding synthesis, it's always been in use at some level. We should separate out two concerns, one is the 'circuit' level of 'what wires go to what transistors' and the other is 'floorplan' or 'how many square millimeters does each chunk of the chip take up.' In die photos, there is little to no remaining circuit information and you're largely seeing the floorplan. I sorta remember the noise a few years back when AMD was claiming full-auto on synthesis, and it's just one of those areas that always sounds like tooling could save you ~30 bodies but hardly works out in practice. Like, everyone's been automating data path stuff, is the 34'th bit of data really all that different from data[12]? Eh, probably not. But the "invalidate line" bit that stretches across the entire cache line should probably have an expert's touch.

MaxxBot
Oct 6, 2003

you could have clapped

you should have clapped!!
I like this shot of the 8008 die because everything is so big you can see the circuit level stuff, only 3500 transistors.

MaxxBot fucked around with this message at 18:35 on Sep 27, 2017

Cygni
Nov 12, 2005

raring to post

JawnV6 posted:

Yeah, sorry, I normally swoop in to rip apart weird theories about CPU design process, but I was kinda curious if you'd picked up on a genuine methodology difference between the two big Intel CPU teams and grabbed one from each.

Die shots are mercilessly doctored. They're summations of different angles, lighting, and other post processing. That's not to say you couldn't get real design information out of them. One of the apocryphal stories is that AMD got their start by taking a high--quality die picture of a 386(maybe earlier?), blowing it up to be the size of their cafeteria, and having their architects walk around it to reverse engineer the functionality. There was a leaked picture of the Sandybridge die some years before it hit the market, it was damaging not just because it got out, but because it was pretty easy to suss out it was a 4 core part with integrated graphics, at the time Intel didn't know AMD's roadmap that far out. The ones you have here are marketing shots though, more aesthetic than a real map of the territory.

Regarding synthesis, it's always been in use at some level. We should separate out two concerns, one is the 'circuit' level of 'what wires go to what transistors' and the other is 'floorplan' or 'how many square millimeters does each chunk of the chip take up.' In die photos, there is little to no remaining circuit information and you're largely seeing the floorplan. I sorta remember the noise a few years back when AMD was claiming full-auto on synthesis, and it's just one of those areas that always sounds like tooling could save you ~30 bodies but hardly works out in practice. Like, everyone's been automating data path stuff, is the 34'th bit of data really all that different from data[12]? Eh, probably not. But the "invalidate line" bit that stretches across the entire cache line should probably have an expert's touch.

This is awesome info! Thanks dude!

Saukkis
May 16, 2003

Unless I'm on the inside curve pointing straight at oncoming traffic the high beams stay on and I laugh at your puny protest flashes.
I am Most Important Man. Most Important Man in the World.

Paul MaudDib posted:

I'm trying real hard to cut down on computer junk, so I'm looking at smaller rather than larger. I'd really like to maybe have a RVZ01 or that new water-cooled DAN case that I can put on my desk as a primary system, and then either have a short+shallow half-height rack with a couple 2U/3U chassis for processing servers and a small portable Synology-sized NAS (U-NAS NSC-810a case).

Have you considered attaching your PC under the desk? When it's out of the way it doesn't really matter how big the case is. I don't think there is small enough case that I would want to put it on my desk.

Rastor
Jun 2, 2001

MaxxBot posted:

I like this shot of the 8008 die because everything is so big you can see the circuit level stuff, only 3500 transistors.



Have you read Ken Shirriff's writeup on it?

http://www.righto.com/2016/12/die-photos-and-analysis-of_24.html

It's interesting stuff. Notice the triangular shape of the blocks on the left side?

quote:

Below the registers is the carry look ahead circuitry. For addition and subtraction, this circuit computes all eight carry values in parallel to improve performance. Since the low-order carry depends on just the low-order bits, while the higher-order carries depend on multiple bits, the circuit block has a triangular shape.

The triangular layout of the ALU is unusual. Most processors stack the circuitry for each bit into a regular rectangle (a bit-slice layout). The 8008, however, has eight blocks (one for each bit) arranged haphazardly to fit around the space left by the triangular carry generator.

Khorne
May 1, 2002

Craptacular! posted:

For now I'm just going to keep my Ivy and see what I can reach at 1.2v or something. I saw some guy achieved 4.5ghz at 1.4 and everyone was telling him that that was too much voltage and he was killing his chip, so I figure at either stock voltage or just a little over I'll see if I can get what I want and that's that. Delidding is absolutely not an option as I can't afford to replace a CPU right now, and actually to be honest I could totally just get by without OC since it's not like I ever complain about this PC. I just decided to do it now that the generation is five years old and no more documentation about how to OC it is going to appear than what's already out there.

This would, literally, be the first time I even try to OC anything. I go back a long ways (386 -> K6-2 -> Coppermine P3 -> Prescott P4 -> Conroe C2D -> Ivy i7) and never bothered to until now. But I'd feel more comfortable replacing this nearly 5 year old Corsair pump with something new, and I also thought my case could be prettier even when it was new, so why not.
I run an i7 3770k at 4.5ghz with 1.35 (1.31-1.32) draw. It might actually be 1.325 or 1.335 or something, I'm not sure and not at my desktop. I have since release, so over 5 years at this point. On air (noctua nh-d14) and it's fine. For most of its life it kept itself fully clocked and I'd run 100% load stuff overnight (8-12 hours per day).

My advice would be to go higher. Just pick a safe enough voltage and see where you get. My chip doesn't run stable at 4.6 no matter how much voltage I throw at it. It's just as unstable as it is at around 1.35 or so.

Has anyone actually had a CPU die in the past 5 years? I can only think of one anecdotally, and it was an amd chip crushed by a too heavy copper heatsink that required multiple high risk modifications to even get into that state.

I probably shouldn't post this here, but I have a P3 that ran 24/7 for over 12 years with a nickel jammed between the heatsink and cpu to force contact on the other half of the heatsink. I retired it because I had no use for it anymore and a better replacement. Why there were coins in between my heatsink and cpu is another story entirely.

Khorne fucked around with this message at 00:18 on Sep 28, 2017

Stanley Tucheetos
May 15, 2012

Khorne posted:

I run an i7 3770k at 4.5ghz with 1.35 (1.31-1.32) draw. It might actually be 1.325 or 1.335 or something, I'm not sure and not at my desktop. I have since release, so over 5 years at this point. On air (noctua nh-d14) and it's fine. For a few years I even had the underclocking turned off so it just ran full blast 24/7 for 3+ years. When I slept I ran physics simulations that maxed cpu for 4 of the years.

My advice would be to just go for 4.4 @ 1.325. That's what I ran at for a long time. I forget my exact oc settings now, but I can hit 4.6 "somewhat stable" at 1.35 or so. My chip in particular seems to not benefit much from more voltage, like 1.45v doesn't make 4.6ghz any more stable. 4.6ghz is "stable enough" for cpu demanding games, but it's not stable enough for computational physics or prime95 all night. 4.5 is just 100% stable.

My ram sucks no idea if that influences anything. I haven't even tried to mess with it. Maybe I'm running it at worse settings than I could be, too.

I must of hit the jackpot with my 3570k then. For the past 5 years I've been running it at 4.7ghz at 1.35 on air. It will do 4.8 at 1.4 but I can't make anything stable past that. I'm satisfied for the most part with it to this day but lacking some of the newer features is tempting me to go in on coffee lake.

hitze
Aug 28, 2007
Give me a dollar. No, the twenty. This is gonna blow your mind...

Time to replace my 4590

GRINDCORE MEGGIDO
Feb 28, 1985


Stanley Tucheetos posted:

I must of hit the jackpot with my 3570k then. For the past 5 years I've been running it at 4.7ghz at 1.35 on air. It will do 4.8 at 1.4 but I can't make anything stable past that. I'm satisfied for the most part with it to this day but lacking some of the newer features is tempting me to go in on coffee lake.

That's a good chip. My friends 3570k was stock until the last 6 months of it's life, when he ran it at 4.8 - 1.35(I think). Then the board died.

spasticColon
Sep 22, 2004

In loving memory of Donald Pleasance

GRINDCORE MEGGIDO posted:

That's a good chip. My friends 3570k was stock until the last 6 months of it's life, when he ran it at 4.8 - 1.35(I think). Then the board died.

This is what I'm worried about with 2500K rig. The motherboard will croak at some point and that will force me to upgrade even if the CPU is still good.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

spasticColon posted:

This is what I'm worried about with 2500K rig. The motherboard will croak at some point and that will force me to upgrade even if the CPU is still good.

This is what happened to me. The cost of a new motherboard wasn't much less than the cost of selling the 2500k / RAM and buying a 5820k rig.

GRINDCORE MEGGIDO
Feb 28, 1985


Same :(

I was totally happy with my 2500K and was looking forward to putting a 1080 in it, the board died last year.

I wanted to keep that chip going for a perversely long time too.

BurritoJustice
Oct 9, 2012

When my Z77 board croaked I went on Australia's equivalent of Craigslist and found a listing for the exact same model, perfect condition, for $25 and a ten minute drive. He'd listed it six minutes before I saw it, it was a surreal experience. The same board sells on eBay for $200-300

PerrineClostermann
Dec 15, 2012

by FactsAreUseless
Still running a 2600k @ 4.2GHz. I used to run it higher, but it got unstable. With this I could lower the voltage a little, too.

Craptacular!
Jul 9, 2001

Fuck the DH

BurritoJustice posted:

The same board sells on eBay for $200-300

Christ. You've caused me to realize I could probably upgrade from what I have to an 8700 non-K for $100, maybe less if chip value doesn't tank too bad before Intel releases H boards. Maybe I should just keep this stuff stock for the next few months until the demand for more cores sinks it's resale.

craig588
Nov 19, 2005

by Nyc_Tattoo
I got a 5820K when my 680I 775 motherboard died. Caps started bursting on it but it kept going for another year with broken caps. It's cool to keep stuff until it's physically failing and not have to replace it just for lack of performance. Don't get it wrong, there still was a large performance difference, but imagine keeping a Pentium 1 for 9 years, it'd be useless instead of just noticeably slow.

Adbot
ADBOT LOVES YOU

Ak Gara
Jul 29, 2005

That's just the way he rolls.
I hope the 8700k is as good as it looks. Might finally upgrade!

PerrineClostermann posted:

Still running a 2600k @ 4.2GHz. I used to run it higher, but it got unstable. With this I could lower the voltage a little, too.

I had to replace the stock thermal paste on my 2500k @ 5.0 + Corsair H100 after a few years as my temps kept going up every few months. That poo poo had baked on, I had to use a razor to remove it! Temps dropped 20c afterwards. Although I don't know how much that was new paste vs old paste or corsairs pre applied paste vs some good poo poo. Vacuuming the radiator probably helped, too.

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