|
SwissArmyDruid posted:I would not call him a fanboy. Seems like a bit of an insult; I find his analysis to be realistic and even-keeled.
|
#
?
Dec 15, 2016 22:38
|
|
|
|
| # ? May 15, 2024 17:39 |
|
|
The Only thing that slows down the web for any modern system, is the onslaught of pop in, auto play, interactive, infected ads thrown every friggin where on sites. Most sites work just fine in themselves, (as shown when you visit the same sight with a good ad blocker) but the evolution from popup ads to the new age interactive media heavy bs ones they are trying to shove down our throats these days is the only reason any system feels slow on the web today. Anime Schoolgirl posted:apparently they set the render sampling to 128 samples (someone with a 5960x have gotten the 36 second time that way), which might have been a default years ago but the default for blender is now 200 which is making people go "huh?" Could they have lowered it just to keep the "demo" to a reasonable period of time? (60 sec target) Though that all went out the window with the "Live Vega/Ryzen demo"
|
|
#
?
Dec 15, 2016 22:39
|
|
|
I don't know where the notion that non-nerds don't give a poo poo about how fast their computer is comes from, no one likes waiting for their computer to do poo poo. I hear people at work complaining about computer slowness all the time although much less so now that they mostly have SSDs.
|
|
#
?
Dec 15, 2016 22:39
|
|
|
Anime Schoolgirl posted:apparently they set the render sampling to 128 samples (someone with a 5960x have gotten the 36 second time that way), which might have been a default years ago but the default for blender is now 200 which is making people go "huh?" quote:https://www.reddit.com/user/AMD_james Seems like they say its 150
|
|
#
?
Dec 15, 2016 22:47
|
|
|
Risky Bisquick posted:Seems like they say its 150 
|
|
#
?
Dec 15, 2016 22:50
|
|
|
Boiled Water posted:It'll be good enough for most people who buy facebooking laptops that it won't matter.
|
|
#
?
Dec 15, 2016 23:09
|
|
|
SwissArmyDruid posted:Plz no, I'm probably already going to nab a EK Predator, I need AIOs floating around in my IT closet like I need another hole in the head.
|
|
#
?
Dec 15, 2016 23:23
|
|
|
Josh Lyman posted:What's wrong with AIOs? Note I've only ever used air cooling. I have a Cooler Master Nepton 280L that I am about to replace with a Noctua air cooler and I can tell you that it is loud (probably could fix this with new fans), the pump makes a whining noise pretty much all the time, and it really doesn't perform that much better than any air cooler I have used before.
|
|
#
?
Dec 15, 2016 23:25
|
|
|
Lowen SoDium posted:I have a Cooler Master Nepton 280L that I am about to replace with a Noctua air cooler and I can tell you that it is loud (probably could fix this with new fans), the pump makes a whining noise pretty much all the time, and it really doesn't perform that much better than any air cooler I have used before.
|
|
#
?
Dec 15, 2016 23:25
|
|
|
Josh Lyman posted:What's wrong with AIOs? Note I've only ever used air cooling. I don't need an extra one that I will likely never use taking up space in my IT closet, because I will already have a semi-custom watercooling loop in the form of the EK Predator, just as soon as EK comes up with a version with an AMD compatible waterblock and/or mounting hardware. Sorry if I wan't clear on that.
|
|
#
?
Dec 15, 2016 23:37
|
|
|
fishmech posted:No it absolutely will not. I don't know where this idea keeps coming from but it's not 2002 anymore, average users aren't using sites that are mostly static text with a few images, they're using websites with javascript poo poo and HTML5 video splattered all over the place and it's quite performance sensitive. Especially when they're on Facebook and playing Facebook games. It's "just a browser" but it's a browser running so much stuff. Maybe someone who hadn't gotten their hands on a new computer in 10 years wouldn't notice the difference, but that's not going to be a lot of people. Even -U series mobile chips can and do bog down all the loving time on random websites everywhere, to say nothing of lol ARM chips. (I had an XPS13 and ended up going back to a laptop with a full 45w mobile quad because web performance with more than like 5 tabs in either browser just jumped right in the toilet.)
|
|
#
?
Dec 15, 2016 23:38
|
|
|
Josh Lyman posted:Is it louder than an air cooler though? Considerably.
|
|
#
?
Dec 15, 2016 23:56
|
|
|
Re: adoredtv, I didn't mean it like that and should have probably used "fan" I like his videos too, though sometimes they're quite optimistic
|
|
#
?
Dec 16, 2016 01:00
|
|
|
I know next to nothing about AMD since the days I bought a few Durons. Why are people thinking/hoping that the motherboards will be significantly cheaper than Intel's? Do AMD boards typically cost less in general?
|
|
#
?
Dec 16, 2016 01:12
|
|
|
WAR DOGS OF SOCHI posted:I know next to nothing about AMD since the days I bought a few Durons. Why are people thinking/hoping that the motherboards will be significantly cheaper than Intel's? Do AMD boards typically cost less in general?
|
|
#
?
Dec 16, 2016 01:21
|
|
|
And the reason they do, is that they suck. No one in their right mind would pay any more. I still can't believe how awful all the amd/via/nvidia chipsets I had were.
|
|
#
?
Dec 16, 2016 01:42
|
|
|
Ihmemies posted:And the reason they do, is that they suck. No one in their right mind would pay any more. I still can't believe how awful all the amd/via/nvidia chipsets I had were. All motherboards from that era suck mostly due to that whole capacitor scandal. My nforce board worked great until it... didn't.
|
|
#
?
Dec 16, 2016 01:44
|
|
|
I watched the Zen conference and it wasn't horrible actually. I'm used to things like E3 cringeshows and Apple's reality-disconnected conferences. But this was surprisingly tolerable. I just topped up my computer with an intel processor for gaming for the next couple of years. Once it starts getting long in the tooth, I hope that I could try out AMD for a generation, provided its economically smart to do so.
|
|
#
?
Dec 16, 2016 03:02
|
|
|
Ihmemies posted:And the reason they do, is that they suck. No one in their right mind would pay any more. I still can't believe how awful all the amd/via/nvidia chipsets I had were. shame about the CPUs though
|
|
#
?
Dec 16, 2016 03:04
|
|
|
defaultluser posted:Can anybody here help out with trying to recreate the 36 seconds AMD got for an live demo of Blender test? If you have an 8-core, it would be incredibly helpful defaultluser posted:Just did a quick sanity check, and my Ivy Bridge Core i3 3.2 GHz took 4:14 to run the same test. I managed 61 (70 @ 200 samples) seconds with 2600k@4.8Ghz. Fury did it in 25.5 once the render kernels were loaded (1:15 without) 36 seconds is the kind of performance I expected to see by 2013 when I built this, in 2011. Now that it's 2016 I can stand to wait a while longer
|
|
#
?
Dec 16, 2016 05:42
|
|
|
defaultluser posted:Just did a quick sanity check, and my Ivy Bridge Core i3 3.2 GHz took 4:14 to run the same test. Tested on my Core i3 6100 @ 3.7GHz and it took 1:43.
|
|
#
?
Dec 16, 2016 05:54
|
|
|
Core i5 4590, 1:53, so I guess it may be worth it to get Zen? I'd be moving from an H87/i5 to an enthusiast build.
|
|
#
?
Dec 16, 2016 06:50
|
|
|
What are you guys using to trim the video file?
|
|
#
?
Dec 16, 2016 08:51
|
|
|
I currently have a Core i5 4570. I would love to give AMD money since I opted to get a GTX 1060 over an RX 480. Please be awesome and good Ryzen.
|
|
#
?
Dec 16, 2016 10:10
|
|
|
Ryzen looks pretty good considering that people with 8-10 core Intel chips are close or falling short of the benchmark. Nearly double the speed of a 2600k seems like a good upgrade but I don't have a 4 year old computer so maybe my perspective is different.
|
|
#
?
Dec 16, 2016 15:18
|
|
|
if anyone's hoping for a single threaded pipe dream it ain't here, but bringing up average # of haswell-grade cores per user would really help overall
|
|
#
?
Dec 16, 2016 15:41
|
|
|
Anime Schoolgirl posted:if anyone's hoping for a single threaded pipe dream it ain't here, but bringing up average # of haswell-grade cores per user would really help overall Yet to be seen how high it can clock with its 25mhz fractional/crazyhigh multiplier boost. 4ghz+ seems plausible which will make the platform very attractive.
|
|
#
?
Dec 16, 2016 16:10
|
|
|
Risky Bisquick posted:Yet to be seen how high it can clock with its 25mhz fractional/crazyhigh multiplier boost. 4ghz+ seems plausible which will make the platform very attractive. We know this how? (Why is 4+ plausible? I gotta see some facts before  )
|
|
#
?
Dec 16, 2016 22:23
|
|
|
Potato Salad posted:We know this how? (Why is 4+ plausible? I gotta see some facts before Broadwell-E overclocks to low-mid 4ghz and it's not an unreasonable upper range given the gains seen in the last several generations of chips from both vendors. Or maybe the automatic overclocking thingy is already pushing them to the upper limits and they won't go much beyond what it says on the spec sheet.
|
|
#
?
Dec 17, 2016 00:05
|
|
|
Potato Salad posted:We know this how? (Why is 4+ plausible? I gotta see some facts before Ack, I just saw a thing which reminded me that I was going to answer this question once I got back to a keyboard. Stock Broadwell clocks about 3.2, up to about 3.7 base, and then overclock to low 4.0. If Ryzen is comparable, they're starting up 200 MHz base, which would make 4.0 on a quadcore part trivial. Hypotheticals are not reality, and you should still wait to see what it actually does. (Take some salt now, so you're not salty later.)
|
|
#
?
Dec 17, 2016 01:10
|
|
|
Potato Salad posted:We know this how? (Why is 4+ plausible? I gotta see some facts before The Intel 6800 and higher chips are similar in terms of base clock and OC to comparable levels. There's nothing to suggest that Ryzen couldn't hit comparable levels.
|
|
#
?
Dec 17, 2016 02:47
|
|
|
If GlobalFoundries is doing Zen, then I would expect moderate overclocks at best. I'd like to be wrong, of course, but I'm trying to stay at least somewhat grounded here.
|
|
#
?
Dec 17, 2016 03:00
|
|
|
Didn't AMD confirm that Zen over shot their goal of 40% increase in IPC? If that's the case, I don't see how single threaded performance could be anything but pretty good.
|
|
#
?
Dec 17, 2016 03:59
|
|
|
Overshooting their goal only brings them up to parity with Broadwell/Broadwell-E, when Intel are moving off of Skylake and onto Kaby Lake. In short, this is not X64 2: Electric Boogaloo, and 40% increase in IPC doesn't tell us how much optimization they have left in the tank.
|
|
#
?
Dec 17, 2016 04:18
|
|
|
Kaby Lake has basically no IPC optimizations and minimal clock improvements though, right? Also, interesting poll: http://www.strawpoll.me/11885898/r
|
|
#
?
Dec 17, 2016 04:38
|
|
|
Kaby Lake is terrible unless you delid it, and even then it's just Skylake minus a few watts.
|
|
#
?
Dec 17, 2016 04:40
|
|
|
PerrineClostermann posted:Kaby Lake has basically no IPC optimizations and minimal clock improvements though, right? My point is that even as Intel has left an opening for AMD to catch up, they have not done so, and even then the window will close by the time the back half of 2017 comes around with Cannonlake. Also, those poll numbers don't look good for AMD. I think they would still continue to be a good value up to the $700 mark relative to a top=flight Skylake's $1100, but this leads me to believe that people are too skittish and still think construction cores are synonymous with AMD.
|
|
#
?
Dec 17, 2016 04:46
|
|
|
I think the results are more indicative of a desire for higher clock speed/single core performance. If they can manage over 4GHz in general, I imagine people would be willing to pay more.
|
|
#
?
Dec 17, 2016 04:48
|
|
|
oh so Broadwell/Broadwell-E have poor signal threaded performance? Is that what you're saying?
|
|
#
?
Dec 17, 2016 05:03
|
|
|
|
| # ? May 15, 2024 17:39 |
|
|
Bloody Antlers posted:oh so Broadwell/Broadwell-E have poor signal threaded performance? Is that what you're saying? Just worse than Skylake, intel's current offering.
|
|
#
?
Dec 17, 2016 05:15
|
|
