|
Arivia posted:I’m really looking forward to seeing how that’s not going to work. Sticking X299 and a Vega desktop CPU in an AIO form factor designed for mobile GPUs was crazy when they announced it but could MAYBE work. Now that we know one is a fusion reactor and one is a core sample of the sun, it’s going to be absurd. Oh, and Apple already staked their ground on it being the exact same size as the regular iMac. Apples new Planned Obsolescence Overheating Phenomena.
|
#
?
Sep 28, 2017 12:34
|
|
|
|
| # ? Jun 4, 2024 01:56 |
|
|
Yeah the only way that's going to work is to use binned products, then reduce voltages frequencies and throttle those massively under sustained load. Apple is no stranger to either of those strategies though, iirc they often used cut down custom NVIDIA SKUs for MBPs. Even then the air cooled combination of X299 and Vega in a Laptop form factor is going to be hilarious. 
|
|
#
?
Sep 28, 2017 12:45
|
|
|
Arivia posted:I’m really looking forward to seeing how that’s not going to work. Sticking X299 and a Vega desktop CPU in an AIO form factor designed for mobile GPUs was crazy when they announced it but could MAYBE work. Now that we know one is a fusion reactor and one is a core sample of the sun, it’s going to be absurd. Oh, and Apple already staked their ground on it being the exact same size as the regular iMac. Even the current 7600k+RX580 iMacs barely keep themselves from melting: AppleInsider posted:Previous iMacs would thermal throttle under extended 100 percent CPU load, and this year's machine is no exception. After about a 15 seconds of full load, the Kaby Lake iMac reached 99 celsius and slowed down from 4.4GHz to 4.3 and after a few more seconds was down to 4.2GHz. For the next 10 minutes of our test, it stayed at about 97 celsius and 4.2GHz, which is the rated clock speed. And that's just CPU load, running the GPU simultaneously will make it even worse. Good luck with X299+Vega Apple
|
|
#
?
Sep 28, 2017 14:26
|
|
|
Lol 100 degrees on a 7600k in 15 seconds. Im literally impressed thats even possible and very amused you get to pay Apple for the privelege to witness such a thing
|
|
#
?
Sep 28, 2017 16:10
|
|
|
Yeah I feel retroactively sorry for putting my daily-driver macbook pro through that for 5 years. I'm sorry.  My current CPU has an actual cooler and air and it's much happier now! 
|
|
#
?
Sep 28, 2017 16:13
|
|
|
1gnoirents posted:Lol 100 degrees on a 7600k in 15 seconds. Im literally impressed thats even possible and very amused you get to pay Apple for the privelege to witness such a thing If those are the internal core sensors, that's not even surprising tbh, they show a very different picture than "cpu temp".
|
|
#
?
Sep 28, 2017 16:18
|
|
|
I mean it's not like apple couldn't do some kind of low profile liquid loop for this poo poo connected to a 280mm rad (or two 140mm ones) but lmao no they'll just let the plastic slough off after an hours use and ask you for another 5k for the privilege of using a similar PoS.
|
|
#
?
Sep 28, 2017 16:19
|
|
|
Could an old PSU cause GPU usage drops that seem to cause stuttering? I upgraded my wife's 2500k to a 2600k and her nvidia 460 to an AMD 480 so she can play PUBG. I've tried DDU with different drivers, playing with voltages, power limits, clocks, etc. The PSU is an antec earthwatts 500W but is over six years old at this point.
|
|
#
?
Sep 28, 2017 16:20
|
|
|
It'd probably just result in your drivers crashing instead, but it could maybe be possible? You should probably change the PSU anyway though.
|
|
#
?
Sep 28, 2017 16:22
|
|
|
Truga posted:If those are the internal core sensors, that's not even surprising tbh, they show a very different picture than "cpu temp". From my past experience, no they aren't.
|
|
#
?
Sep 28, 2017 16:36
|
|
|
FaustianQ posted:Apples new Planned Obsolescence Overheating Phenomena. O/T and You're probably joking but I bet there are millions of fairly recent rMBPs out there that throttle to 1.2-1.3 Ghz under load because Apple's TIM turned into a dry and brittle mess and the heatsinks/fans are filled with dust and lint. CPU won't perform well anymore, device gets hot, iPhoto (or whatever) becomes laggy, user goes out to buy a new MBP because the old one is slow. Mine was one of them but repasting fixed it. Their solution for the new MBPs? They mounted CPU and GPU on the other side of the PCB so you have to remove the entire logic board to access the CPU/HSF, making repasting a lot harder and more time consuming for users and repair shops.  e: if any mac user is curious about this, use this old Intel tool and Handbrake to see if the CPU even holds nominal clockspeed. My old 15" 2.3 Ghz rMBP really throttled to ~15W/1.2 Ghz in Unigine Heaven and boosts to ~3.3 Ghz after cleaning and repasting. eames fucked around with this message at 16:44 on Sep 28, 2017 |
|
#
?
Sep 28, 2017 16:41
|
|
|
B-Mac posted:Could an old PSU cause GPU usage drops that seem to cause stuttering? OTOH, AMD drivers are strange and if you want to know if it's power for sure run Furmark and Prime and see how it responds to that. Furmark isn't a good GPU stability test because it touches a very tiny section of the GPU, but it does draw as much power as you'll ever see to test cooling and power delivery. Similar deal with Prime, it should use more power than you ever see in real world use.
|
|
#
?
Sep 28, 2017 16:48
|
|
|
Arivia posted:From my past experience, no they aren't. Yikes. The cores must be running at like 120 degrees at that point.
|
|
#
?
Sep 28, 2017 16:52
|
|
|
eames posted:O/T and You're probably joking but I bet there are millions of fairly recent rMBPs out there that throttle to 1.2-1.3 Ghz under load because Apple's TIM turned into a dry and brittle mess and the heatsinks/fans are filled with dust and lint. CPU won't perform well anymore, device gets hot, iPhoto (or whatever) becomes laggy, user goes out to buy a new MBP because the old one is slow. Mine was one of them but repasting fixed it. No matter one's stance on the concept of planned obsolescence, I believe we call all agree that if it were happening Apple is at the top of that list.
|
|
#
?
Sep 28, 2017 17:38
|
|
|
1gnoirents posted:No matter one's stance on the concept of planned obsolescence, I believe we call all agree that if it were happening Apple is at the top of that list. It's surprising how much that is not the case with iOS devices. It took 5 years until my 4th gen iPad stopped receiving the latest version of iOS 11, but it still gets updates for the iOS 10. I wish I could get an Android with that level of support.
|
|
#
?
Sep 28, 2017 18:41
|
|
|
It's not really "supported" but you can load the latest Android on 5 year old phones too, you just need to turn to something like LineageOS. 7.1 works great on my Nexus 4.
|
|
#
?
Sep 28, 2017 18:50
|
|
|
Looks like Gigabyte may release custom Vega cards after all? http://www.guru3d.com/news-story/gigabyte-will-release-custom-radeon-rx-vega-64-cards.html
|
|
#
?
Sep 28, 2017 18:53
|
|
|
Saukkis posted:It's surprising how much that is not the case with iOS devices. It took 5 years until my 4th gen iPad stopped receiving the latest version of iOS 11, but it still gets updates for the iOS 10. I wish I could get an Android with that level of support. I meant more along the lines of the degree Apple designs everything to be non upgrade able and repairable more than designing poo poo to melt down like above. Phones are tablets largely excluded I guess since "everybody does it", but I can tell you right now you are in a minority there. I'm impressed your battery is still ok after all that time. My work phone is an iphone and I really hate the drat thing more all the time.
|
|
#
?
Sep 28, 2017 18:53
|
|
|
1gnoirents posted:No matter one's stance on the concept of planned obsolescence, I believe we call all agree that if it were happening Apple is at the top of that list. I personally think its less a thing companies actively plan for, and more companies not giving a flying gently caress about serviceability/repairability due to consumers generally not caring about those things when making purchasing decisions. It's true in other markets too, like the packaging of engines in cars (yes I would really like to take the whole loving engine out in order to replace the alternator, thanks guys). I'm guilty of it too. I absolutely put things like thickness and weight above easy serviceability for laptop/tablet/phone purchases.
|
|
#
?
Sep 28, 2017 18:54
|
|
|
Cygni posted:I absolutely put things like thickness and weight above easy serviceability for laptop/tablet/phone purchases. It's you, you're the reason everything is terrible. I on the other hand pretty much treat iFixit's repairability score as the de facto review score for any product.
|
|
#
?
Sep 28, 2017 18:59
|
|
|
Cygni posted:I personally think its less a thing companies actively plan for, and more companies not giving a flying gently caress about serviceability/repairability due to consumers generally not caring about those things when making purchasing decisions. It's true in other markets too, like the packaging of engines in cars (yes I would really like to take the whole loving engine out in order to replace the alternator, thanks guys). I believe its not so cut and dry. But as far as Apple is concerned, there is a lot of direct evidence of them actually spending more money on a product to make it difficult to service or upgrade. Its probably obvious I'm a pretty big Apple hater (for their computers) but I really do hate that practice. Its one thing to include the fact its just going to be replaced in 2 years when you design something, as that can and should save you money developing it. Basically everything is guilty of this outside of a regular desktop computer, but the way Apple deals with it puts them in a special place in my heart. My job is ultimately providing computers to corporate clients and lightly speaking supporting the equipment long term which means I have to dick around with the occasional Apple thing and I'm rarely not disappointed with them (and we only get the retarded $4000 machines that are supposed to be good too). The best I can say is "wow it looks nice"
|
|
#
?
Sep 28, 2017 19:01
|
|
|
Speaking of antialiasing in places you wouldn't expect, the Nintendo DS actually natively antialiased everything, and with a slightly novel way of accomplishing it, which you can read here as a dec tried to implement it in a an emulator the way the original hardware accomplishes it, rather than just applying typical gpu solutions towards it Of course, the DS renders at such a low resolution it barely helps anything really
|
|
#
?
Sep 28, 2017 19:08
|
|
|
https://www.newegg.com/Product/Product.aspx?item=N82E16814125908 A gigabyte gtx1080 for 499$ is a pretty good price right? Even the 1070s are going for $500+ due to bitcoins. But i do know that a new card (the 1070ti?) is coming soon so any comments on if this is a good buy?
|
|
#
?
Sep 28, 2017 19:29
|
|
|
https://www.reddit.com/r/buildapcsales/comments/71scqq/gpu_gigabyte_geforce_gtx_1080_windforce_oc_44099/ See ^^this^^ Just check out buildapcsales or slickdeals Risky Bisquick fucked around with this message at 19:44 on Sep 28, 2017 |
|
#
?
Sep 28, 2017 19:36
|
|
|
ughhhh posted:https://www.newegg.com/Product/Product.aspx?item=N82E16814125908 That's a blower style card (and not even the half-decent Nvidia ones with vapour chambers), I would avoid it unless you don't care about noise. Or you have a really cramped case/insufficient case fans and need help exhausting hot air out the back.
|
|
#
?
Sep 28, 2017 19:42
|
|
|
sauer kraut posted:That's a blower style card (and not even the half-decent Nvidia ones with vapour chambers), I would avoid it unless you don't care about noise. Or you have a really cramped case/insufficient case fans and need help exhausting hot air out the back. I was actually looking specifically for a blower style card because I have a mitx case (evga hadron) with a gtx970 founders edition (which has been running really well in this case). Sad to hear that the 3rd party blower style cards went as good as the founders edition ughhhh fucked around with this message at 19:56 on Sep 28, 2017 |
|
#
?
Sep 28, 2017 19:51
|
|
|
I kinda feel like upgrading from my 1070 to a reference 1080...wouldn't even need a new block...
|
|
#
?
Sep 28, 2017 19:54
|
|
|
ughhhh posted:I was actually looking specifically for a blower style card because I have a mitx case (evga hadron) with a gtx970 founders edition (which has been running really well in this case). Sad to hear that the 3rd party blower style cards went as good as the founders edition Honestly it's hard to say, those black box cards are mostly used by integrators/prebuilt machines or people who don't know the difference, and almost never get properly reviewed. Unless they're white and have Asus printed on top..
|
|
#
?
Sep 28, 2017 20:00
|
|
|
ughhhh posted:I was actually looking specifically for a blower style card because I have a mitx case (evga hadron) with a gtx970 founders edition (which has been running really well in this case). Sad to hear that the 3rd party blower style cards went as good as the founders edition I'm not familiar with the thermals of that case so it might be an exception, but I've yet to see any evidence that blowers work better than open air coolers in a single GPU system in modern times. The issue with blower coolers today is that they are guaranteed to throttle quite a bit and even at their best the noise they make under full load is significant. A solid open air cooler does have to work harder, but it doesnt approach throttling or maxing fan speed. There is probably testing out there to verify it for that case but its actually very hard to come by in general. Reviewers of cases tend to put blower cards into MITX cases right off the bat. I have a RVZ02 which is much smaller than that case and it only increased my temperatures 5* and fan speed a bit depending on the card. There is another caveat here where I can attest that particular cooler on that 1080 isn't the greatest, but it sure as hell will walk all over any blower version. I've had that exact card in the RVZ02 and I remember my fanspeed hit 60% max and it boosted itself to ~1950. Under ideal conditions not even the FE cooler could do that on an open air test bench for more than 15 minutes. edit: I just looked at some pictures and that case is smaller than I thought and does it have the GPU literally facing the PSU? lol. Like I said, that case might be an exception, however I'd look into it if I were in the same boat 1gnoirents fucked around with this message at 20:12 on Sep 28, 2017 |
|
#
?
Sep 28, 2017 20:07
|
|
|
Linus video on the Marseille voodoo cable. https://www.youtube.com/watch?v=MjJzibFTaqA TL;DR: WTF is this sorcery?
|
|
#
?
Sep 28, 2017 20:34
|
|
|
it's me, i'm the computer guy who's amazed that it's possible to do anti-aliasing with a filter when sweetfx/reshade have been doing exactly that for years
|
|
#
?
Sep 28, 2017 21:11
|
|
|
But *in a cable*.
|
|
#
?
Sep 28, 2017 21:12
|
|
|
B-Mac posted:Could an old PSU cause GPU usage drops that seem to cause stuttering? It could also be an issue with Windows. When I switched from a R9 280 to a 1070 I was having serious performance issues in some games and none of the fixes I tried worked, including trying DDU several times. I eventually gave up and reinstalled Win 10 and that solved the issue.
|
|
#
?
Sep 28, 2017 21:13
|
|
|
repiv posted:it's me, i'm the computer guy who's amazed that it's possible to do anti-aliasing with a filter when sweetfx/reshade have been doing exactly that for years I'm the guy looking at the guy going im the guy at something who isnt amazed by the HDMI cable that does anti aliasing 
|
|
#
?
Sep 28, 2017 21:52
|
|
|
repiv posted:it's me, i'm the computer guy who's amazed that it's possible to do anti-aliasing with a filter when sweetfx/reshade have been doing exactly that for years Installs sweetfx on my sega saturn
|
|
#
?
Sep 28, 2017 21:59
|
|
|
I'm not at all impressed that they could put antialiasing on a chip in a cable however I am a little impressed that no reviewer has been able to detect any added latency
|
|
#
?
Sep 28, 2017 22:01
|
|
|
SwissArmyDruid posted:But *in a cable*.
|
|
#
?
Sep 28, 2017 22:07
|
|
|
I dont know why I didnt mention this but I did ask that company when the PCP article was posted if they were going to do displayport and this was their responsequote:Thank you for your inquiry. Display Port has been on our list and frankly speaking, I personally consider it a more advanced protocol than HDMI. But currently we do not have an immediate plan to release it, at least not for next year. If we start working on it next year, we are looking at a release in 2019. This is mainly because we are still not a huge company so we have to concentrate our engineering resources. And currently the main protocol for home entertainment devices are still HDMI. They might change their tune after all this attention though. edit: I just realized she said keep in touch is this how girlfriend happen
|
|
#
?
Sep 28, 2017 22:07
|
|
|
How much difference are you really gonna see on one iteration of a particular card to another? Especially within the same manufacturer?
|
|
#
?
Sep 28, 2017 22:46
|
|
|
|
| # ? Jun 4, 2024 01:56 |
|
|
codo27 posted:How much difference are you really gonna see on one iteration of a particular card to another? Especially within the same manufacturer? I feel like I just posted this in a different thread, but basically there's no difference in binning anymore, especially NVidia, for instance all the EVGA reference/superclocked stuff gets the same max stable clocks in practice. You're paying extra for lower noise at the end of the day and not really anything else.
|
|
#
?
Sep 28, 2017 22:50
|
|