|
So I don't want to complain but my CPU is too cool running. Im only hitting like 62C running OCCT burn test. Ryzen Master says 73W. That isn't even enough for this thing? Surely I can get more MHZ somehow? [edit] R9 3900x redeyes fucked around with this message at 18:43 on Apr 1, 2020 |
# ? Mar 31, 2020 23:18 |
|
|
# ? May 15, 2024 01:10 |
|
62F as in fahrenheit? Are you sure it’s switched on? Are you using it outdoors?
|
# ? Apr 1, 2020 00:00 |
|
Soricidus posted:62F as in fahrenheit? Are you sure it’s switched on? Are you using it outdoors? Even 62c is a touch on the low end if that is running full tilt if it's on the stock cooler. Not that that's a bad thing but points to auto boost might not be working. Now a custom cooler, yeah you might be able to do that while boosting.
|
# ? Apr 1, 2020 00:28 |
|
Beautiful Ninja posted:Techspot has iGPU benchmarks - https://www.techspot.com/review/2003-amd-ryzen-4000/ FINALLY!
|
# ? Apr 1, 2020 03:05 |
|
Seeing how close the 4900HS performs in video editing and how much of a jump over the previous gen is intriguing, should resonate well with tech youtubers. Battery life looks very good on a 1080p screen reaching 11ish hours watching video and doing office work.
|
# ? Apr 1, 2020 03:59 |
|
Beautiful Ninja posted:Techspot has iGPU benchmarks - https://www.techspot.com/review/2003-amd-ryzen-4000/ Not bad at all. A APU that offers "good enough" 1080p gaming crosses an important threshold for a lot of people. Vega 11 wasn't quite there.
|
# ? Apr 1, 2020 04:01 |
|
Are 4xxxx Ryzens the last generation on AM4?
|
# ? Apr 1, 2020 04:34 |
|
teagone posted:Are 4xxxx Ryzens the last generation on AM4? I don't think it's ever been officially confirmed, but everyone is expecting desktop Intel and AMD to shift to DDR5 in 2021. That'd be a good opportunity to switch to a new socket and avoid monumental compatibility headaches.
|
# ? Apr 1, 2020 04:40 |
|
2021 would also put it on track for USB4. Finally, industry-standard external PCIe lanes that resemble but are legally distinct from Thunderbolt!
|
# ? Apr 1, 2020 08:41 |
|
SwissArmyDruid posted:2021 would also put it on track for USB4. Finally, industry-standard external PCIe lanes that resemble but are legally distinct from Thunderbolt! I keep trying to think of ways the USB-IF can make USB4 as weird and obtuse as possible but I always give up and realize that my imagination is nowhere near as good as theirs is.
|
# ? Apr 1, 2020 08:54 |
|
USB 2.0 Series U
|
# ? Apr 1, 2020 08:55 |
|
X USB 3.60 S X USB 3.60 X X USB Series X
|
# ? Apr 1, 2020 09:05 |
|
Joking aside, I am more concerned over how, for the next two years or so, until USB4 products start showing up, the top-of-the-line laptops may very well be AMD-based, but not have Thunderbolt unless the OEM sticks a Titan Ridge chip in there. And considering how hard it is to get Thunderbolt on desktop AMD platfroms as it stands... does not lead me to be optimistic.
|
# ? Apr 1, 2020 09:27 |
|
FuturePastNow posted:Not bad at all. Fiddling with numbers a bit from the review, Vega 8 Plus has a 40% lead in games over Vega 10. If we apply that to known Firestrike scores, for Vega 10 (2745), we get 3740 for Vega 8 Plus. That puts it only slightly behind desktop Vega 11 in mobile form factor (4037). Assuming it can clock like a Vega VII, you might be looking at 15-20% more performance on top of mobile numbers, I think. I doubt there is going to be as much daylight between Vega 8 Plus Mobile and Vega 8 Plus Desktop, but if there is you're looking @ a ~45% gap, not a 15-20% gap. In scenario #1, that's 4488 at best on desktop; In scenario #2 we'd be looking at 5423. The average between is 4956. Desktop GTX 1050 is ~7645, or about 70% faster in scenario #1, 40% faster in scenario #2, and 55% faster in scenario #3. However the GTX 1650 Max-Q is only 55% faster in those reviews on average, and also has similar Firestrike scores to a GTX 1050, so my math might already be really off for Vega 8 Plus Desktop and it could legit be a hella lot closer once OCed, which is nuts. That'd put it between an RX 560 and GTX 1050, which is 1080p medium-high for a lot of esportsish or low demand titles, and it'd absolutely munch on older poo poo. Klyith posted:Eh, the 3600 will stop being produced once the 4000 series comes out. AMD keeps making 12nm chips because they have cheap fab capacity at GloFlo. Er, probably not TBH. AMD will continue to produce Zen2 for quite some time yet as it fullfills contractual obligations for Rome orders. We're going to keep getting Ryzen 3000 parts for the next 2 years, and on the cheap, because not all dies will meet the binning requirements. In contrast, there's little reason to keep 12nm parts chugging along if you can push your old customers to move to Rome and Milan (the performance gains are too real to not TBH), and GloFo will be going full tilt on 12nm to supply Rome and Milan parts.
|
# ? Apr 1, 2020 14:47 |
|
The talk of USB4 remained me that all the front ports on my case are USB 2.
|
# ? Apr 1, 2020 15:19 |
|
OhFunny posted:The talk of USB4 remained me that all the front ports on my case are USB 2. That's the only reason I get a new case last rebuid. Then I got a monitor with 3.0 hub in it so I used the case ports ~twice for actual 3.0 things
|
# ? Apr 1, 2020 16:00 |
|
OhFunny posted:The talk of USB4 remained me that all the front ports on my case are USB 2. Munkeymon posted:That's the only reason I get a new case last rebuid. If you have a case with USB 2 that you like and is still a decent case, it often isn't the most difficult mod to take apart the front panel and upgrade it to USB 3. All you need is a cheap 3.5" bay panel that looks like this, with the 2 individual connectors. Take those out of the panel and rig up whatever works to fix them where the USB 2 jacks are. The only trouble is that in some cases they use a PCB that hooks up both usb and audio jacks, at which point you have to either desolder the USB 2 ports to fit the new ones, or just throw away the PCB if you don't care about front panel audio. EmpyreanFlux posted:In contrast, there's little reason to keep 12nm parts chugging along if you can push your old customers to move to Rome and Milan (the performance gains are too real to not TBH), and GloFo will be going full tilt on 12nm to supply Rome and Milan parts. From general news articles I had the impression that 7nm is very tight on capacity versus 14/12, so I have a hard time imagining that GF is entirely occupied just producing IO dies? I figured that was why AMD was still producing 12nm zen in pretty decent volumes, for example the 1600AF as a 'new' product well after the 3000s came out. I forgot about epyc though, that does mean the changeover wouldn't be nearly as fast.
|
# ? Apr 1, 2020 16:55 |
|
Like, I guess, maybe? I think the I/O dies meet all contractual obligations with GloFo, I could be wrong. I'm not even sure why AMD would continue making Gen1 parts if they didn't have to, the demand for Gen1 is low and there is no way the DIY market would compel AMD to keep making them. Like, what would AMD even continue to make on 12nm that's worth anyone's time? No one is going to be interested in 12nm RDNA2 or Zen2 I think, and that's also wasted time and money. Polaris is looking long in the tooth as well, you'd only need it to meet contractual obligations at this point. Uhh...chipsets with ASMedia?
|
# ? Apr 1, 2020 17:32 |
|
SwissArmyDruid posted:Joking aside, I am more concerned over how, for the next two years or so, until USB4 products start showing up, the top-of-the-line laptops may very well be AMD-based, but not have Thunderbolt unless the OEM sticks a Titan Ridge chip in there. It will be interesting to see what Apple does. Macbook would be a perfect use-case for a 4900H
|
# ? Apr 1, 2020 18:00 |
|
OhFunny posted:The talk of USB4 remained me that all the front ports on my case are USB 2. Mine has a USB-C but its dead because the only way to plug it into a motherboard is with a $25 adapter. It's like a weird SATA cable design with an extra plastic corner
|
# ? Apr 1, 2020 18:03 |
|
Yeah sorry, about 64C maxed out with OCCT burn test. I am using a Noctua 140mm cooler, so i guess that helps. I ran HWinfo for 2ish days and this is what it shows: It's not bad, but drat, I have a lot more thermal headroom.
|
# ? Apr 1, 2020 18:46 |
|
redeyes posted:Yeah sorry, about 64C maxed out with OCCT burn test. I am using a Noctua 140mm cooler, so i guess that helps. I ran HWinfo for 2ish days and this is what it shows: I think one of the long-term consequences of the ongoing Enermax fuckups is that a lot of people are realizing they don't really need an AIO, a fuckoff huge Noctua does really well too. And the enthusiasts who are buying those $4000 processors are "thought leaders" who chatter in tech forums and that kind of thing. I mean, if you can cool a 64C processor with a 140mm air cooler, why exactly do you need liquid for your baby 8C processor build? Enermax isn't just killing their brand, they're basically killing the AIO market for all the other companies too. It's been a long-term trend back towards air cooling over the last couple years but Enermax really isn't helping.
|
# ? Apr 1, 2020 19:14 |
|
yeah but AIOs look cool so!!! i mean that is literally why i bought one for my personal gaming computer. i went from a Dark Rock Pro 3 to a 280 AIO and i'm pretty sure the cooling was actually better on the air cooler, ha
|
# ? Apr 1, 2020 19:27 |
|
Cygni posted:yeah but AIOs look cool so!!! that or people who want ease of access or move their case around a lot and don't want a big heatsink torquing on their board, but yeah, it's reaching gimmick territory at this point, you don't do it for the cooling or the noise I came into my office yesterday and pushed the button to turn my PC on. Whoops, I had left it turned on, the D15S is just that quiet.
|
# ? Apr 1, 2020 19:42 |
|
Because air coolers did this weird brinksmanship thing around the beginning of the millenium where they just kept getting bigger and taller and heavier to beat out everyone else in the market until the point where delamination of the motherboard at or around the socket or other localized damage was a real concern. And then Socket 370 and Socket A went away, and through-board mounting points became a thing, and processors did this thing where they started drawing less and less power, and now we don't have 300W CPUs unless someone has hosed up badly.
|
# ? Apr 1, 2020 19:46 |
|
USB4 with a new triangular connector that only works in one orientation
|
# ? Apr 1, 2020 19:47 |
|
IMO the main reason why AIO watercoolers gained so much mindshare is because they make prebuilt high-end PCs easier/safer to ship and the "large botique" (i know) manufacturers tend to invest a lot of money in marketing. Almost nobody sells prebuilts with NH-D15s at scale because packaging and shipping would be a nightmare. It doesn't have to be that way but good luck changing the ATX spec. eames fucked around with this message at 19:52 on Apr 1, 2020 |
# ? Apr 1, 2020 19:50 |
|
SwissArmyDruid posted:Because air coolers did this weird brinksmanship thing around the beginning of the millenium where they just kept getting bigger and taller and heavier to beat out everyone else in the market until the point where delamination of the motherboard at or around the socket or other localized damage was a real concern. Or unless someone really REALLY wants to move 200 Mhz farther right on that volts/hertz curve. eames posted:IMO the main reason why AIO watercoolers gained so much mindshare is because they make prebuilt high-end PCs easier/safer to ship and the "large botique" (i know) manufacturers tend to invest a lot of money in marketing. I always thought it was because of case windows. An AIO looks a shitload sleeker than a gigantic aluminum cube with a fan on one side of it, especially given all the other poo poo motherboards do these days to look 2fast2furious. Being able to ship the unit fully assembled without worrying about the package monkeys in the warehouse flinging it down three flights of stairs and cracking the mobo in half also helps. Methylethylaldehyde fucked around with this message at 19:53 on Apr 1, 2020 |
# ? Apr 1, 2020 19:50 |
|
Y'all act like Instapak wasn't a thing, even back then. It was a thing, even back then. $3.72 per, in bulk. https://www.amazon.com/EcoBox-Inches-Instapak-Temperature-2030-1/dp/B00KAFU2WG/ SwissArmyDruid fucked around with this message at 19:57 on Apr 1, 2020 |
# ? Apr 1, 2020 19:54 |
|
I'm convinced AOIs became popular because they're another way for idiots to add RGB.
|
# ? Apr 1, 2020 19:55 |
|
SwissArmyDruid posted:Y'all act like Instapak wasn't a thing, even back then. I'm well aware these exist but they're not going to save a motherboard with a mounted NH-D15S when the whole unit is dropped from 3 feet height.
|
# ? Apr 1, 2020 20:02 |
|
eames posted:I'm well aware these exist but they're not going to save a motherboard with a mounted NH-D15S when the whole unit is dropped from 3 feet height. Also you ever try to get one of those things out of the fins in a radiator without bending the poo poo out of it?
|
# ? Apr 1, 2020 20:13 |
|
SwissArmyDruid posted:Because air coolers did this weird brinksmanship thing around the beginning of the millenium where they just kept getting bigger and taller and heavier to beat out everyone else in the market until the point where delamination of the motherboard at or around the socket or other localized damage was a real concern. That far back was also before heatpipes started getting put in heatsinks. There was a 2-3 year stretch when watercooling first became a thing, it was plainly superior in both performance and noise. Back when people were using repurposed aquarium pumps and cobbled-together parts and part of the process was going to the home depot plumbing section. I did watercooling back then; I still have my swiftech block and pump in the bottom of my parts bin. I think a lot of the watercooling mystique dates back to those days when only the hyper-nerds were DIYing things and posting build logs on forums. It was cool, you could OC to the moon, and it was a thing to boast about. I got on at just the wrong time, less than a year after I put my watercooling stuff together heatpipe coolers started becoming widely reviewed and were like holy poo poo good. whomp-whomp
|
# ? Apr 1, 2020 20:14 |
|
Methylethylaldehyde posted:I always thought it was because of case windows. An AIO looks a shitload sleeker than a gigantic aluminum cube with a fan on one side of it, especially given all the other poo poo motherboards do these days to look 2fast2furious. Yup, thats my theory too.
|
# ? Apr 1, 2020 21:04 |
|
I went to a custom loop solely for aesthetics. Performance isn't any better or worse than my previous Dark Rock Pro 4. GPU is better though
|
# ? Apr 1, 2020 21:06 |
|
So I guess the moral of the story is AMD is awesome, AIO's are not awesome, and Noctua is awesome.
|
# ? Apr 1, 2020 21:44 |
|
AMD is awesome, the average Asetek AIO is probably dead after three years anyways, Noctua has never not been awesome, it's why they can make fans the color that they cemented their branding on and people will swarm to buy them anyways.
|
# ? Apr 1, 2020 22:00 |
|
I like my AMD Wraith Max cooler in my Plex server
|
# ? Apr 1, 2020 22:24 |
|
redeyes posted:Yeah sorry, about 64C maxed out with OCCT burn test. I am using a Noctua 140mm cooler, so i guess that helps. I ran HWinfo for 2ish days and this is what it shows: yes but what are your temps while folding@home (not a trick to help the goon Folding Team #150) Setset fucked around with this message at 22:36 on Apr 1, 2020 |
# ? Apr 1, 2020 22:31 |
|
|
# ? May 15, 2024 01:10 |
|
Huh, I had no idea aios were considered bad, I always thought they were just a progression from air coolers that everybody moved onto these days. I'm still sporting a big fat heatsink+fan. What's so bad about aios. I thought they were quieter and cooler.
|
# ? Apr 1, 2020 22:59 |