|
New Zealand can eat me posted:Does he say which firmware he's running on/which ones hes tried? IIRC ASUS has pushed multiple beta and maybe even a release firmware update out with hilariously unsafe defaults (like 1.8v SOC!!!) They? or maybe another company also recently pushed out older fw with a new name.
|
#
?
Feb 3, 2018 18:55
|
|
|
|
| # ? May 15, 2024 22:07 |
|
|
My last Asus board, an Z170 ITX, auto set the voltages super high, setting a 4.5GHz overclock on my Skylake cooked the VRMs. Also offset voltage settings applied in the bios just didn't work.
|
|
#
?
Feb 3, 2018 19:00
|
|
|
My ASUS Z97A 3.1 also doesn’t apply offset voltages for poo poo, I had to set a constant voltage and turn off some of the power management features to get a stable overclock. Very disappointing, since my old ASUS P67 and Z77 boards work perfectly with offsets.
|
|
#
?
Feb 3, 2018 19:24
|
|
|
I had an ASUS Sabertooth wind up DOA twice, and the third one still likes to change the clock in the BIOS once every 24 hours at random, like I can no poo poo just be using the computer and it's suddenly 4-6 hours into the future and it's not the OS. I do not like ASUS at all, and have been hardcore ASRock since.
|
|
#
?
Feb 3, 2018 20:34
|
|
|
GRINDCORE MEGGIDO posted:Also offset voltage settings applied in the bios just didn't work. Yeah I've had a Z170 Pro Gaming and Z370-E Strix lately and both have had weird stuff with the voltage offset. I think other settings are overruling the voltage but the bios doesnt actually tell you what is doing it, and its got like 40,000 voltage and OC settings so its real hard to diagnose
|
|
#
?
Feb 3, 2018 20:36
|
|
|
Ya. My old Asus p67 was rock solid, really good. But I think I'm staying with Asrock now until I hear some really great things for a long time about Asus. GRINDCORE MEGGIDO fucked around with this message at 00:41 on Feb 5, 2018 |
|
#
?
Feb 3, 2018 20:56
|
|
|
ufarn posted:I think he reviews hardware and writes about tech so I'd assume he'd at least have updated firmware and such. It's pretty crazy that ASUS firmware is bricking CPUs like that anyway; I thought they were the safest choice for AM4, but I might give ASRock another look now. I'm not sure he's justified in blaming ASUS. The 2nd fw he installed was a beta so I would say ultimately that's on him. You can find a lot of posts like this , mentioning "Elmor from ASUS" sharing fw updates via Mediafire and no checksum (lol) coupled with bad advice from enthusiasts like a 1.25v SOC. If you dig up ASUS's original Ryzen Overclocking guide, they specifically say not to go past 1.2v SOC for this reason, as does everyone else. That's not specific to BIOS 3008, just an example of the critical instructions from an employee vs forums telephone game snake oil vs established safe limits at work here. Who does he write for? I'm totally down with a "launch checklist" to confirm norms when I flash a beta bios, but having to read through forum posts to figure out which voltages to lock in to prevent harmful auto settings sounds like a really lovely way to gamble for a few percent perf increase. The specific bios he installed doesn't even offer any tangible benefits to him, the AGESA update was for APU compatability! That AMD still RMA'd it a second time is pretty cool, IMO that is somewhat of an admission that perhaps they did provide an outdated/incorrect reference resulting in various voltage fuckups. I know ASUS wasn't alone in this, but I do think they were the most popular because they had the best early x370 availability, so it's the squeakiest wheel. All that said, ASUS also murdered my B350 Prime or whatever with a bad firmware update around launch. Wasn't even a matter of checking voltages, it ate itself before it could finish flashing. They sent me a second board without even asking for the first one back! That was the biggest motivator behind me getting the Gigabyte Gaming 5, if it continues to fail to post after trying a cmos reset it will kick itself over to try the 2nd bios automatically. New Zealand can eat me fucked around with this message at 18:38 on Feb 4, 2018 |
|
#
?
Feb 4, 2018 00:02
|
|
|
https://wccftech.com/gpus-cpus-cryptomining-amd-ryzen-threadripper/ Apparently the Threadripper mining story is gaining traction. Eugh.
|
|
#
?
Feb 4, 2018 18:50
|
|
|
NewFatMike posted:https://wccftech.com/gpus-cpus-cryptomining-amd-ryzen-threadripper/
|
|
#
?
Feb 4, 2018 20:29
|
|
|
NewFatMike posted:https://wccftech.com/gpus-cpus-cryptomining-amd-ryzen-threadripper/ Well, there goes me ever being able to get one of the future generation Threadrippers when I need to upgrade my professional workstation. EDIT: From that article: quote:“The way to go at this point is to try to explain to end users how they can also get the benefits of mining,” Triolet says, “because if you should get a high-end desktop with a high-end GPU today and you’re not mining, you’re kind of wasting money. If they are just sitting there doing nothing, why not make some nice profit?”
|
|
#
?
Feb 4, 2018 21:51
|
|
|
My understanding of motherboards right now is... ASUS: Okay until they bricked boards, I guess. MSI: Never updates anything. Gigabyte: Melted systems at first but has really good Newegg ratings so assume they're not going to repeat that incident. ASRock: Nice but nowhere near the value they used to be. Going to say people make it sound like B350 is only $10 more than A320 but it's really more like $20-30. If it was $10 I would get one. I may go for ASRock A320 but will try to spring for Gigabyte B350 if I can. I have a personal affinity for MSI after owning one of their GPUs but people seem to hate them, and the APU models will require an upgrade they seem intent to not bother with. Craptacular! fucked around with this message at 22:08 on Feb 4, 2018 |
|
#
?
Feb 4, 2018 22:06
|
|
|
Craptacular! posted:My understanding of motherboards right now is... It's true about Asrock value. They were the absolute bomb for value at one point (Z77, maybe newer) and also good. I'm just going to stick with Asrock now for a few years.
|
|
#
?
Feb 4, 2018 23:38
|
|
|
I've settled on ASRock stuff. And yeah Z77 was my first one and that was an insane deal, came with 8GB of DDR3 for free. No problems with a Z270 Taichi.
|
|
#
?
Feb 4, 2018 23:48
|
|
|
That rep of Gigabyte replacing components in their products after the initial launch with positive reviews, is that still the case?
|
|
#
?
Feb 5, 2018 00:23
|
|
|
I actually really like my MSI B350M board. Yeah, they haven’t updated it but nothing is broken and needs to be updated. I’d like a future BIOS update to update the AGESA code once it gets more stable, and definitely an update to disable the AMD PSP. Aside from those two complaints, the board itself works great. Edit: wait, it does have 1 big problem: if you overclock with CoolNQuiet enabled it locks the processor to a lower than expected frequency. A lot of people have reported this bug to MSI and they haven’t done poo poo about it yet.
|
|
#
?
Feb 5, 2018 01:13
|
|
|
NewFatMike posted:https://wccftech.com/gpus-cpus-cryptomining-amd-ryzen-threadripper/ quote:HardOCP suggests Uh huh quote:that a powerful multi-core processor is capable of paying for itself in under a year while mining crypto Like ETH? quote:For mining Monero Ohhhhhhhhhhhhhhhhhhhhhhh The upcoming 2200G and 2400G are more at-risk than TR, since their GPU is a revised RX 550, and 10Mh/s might still be enough for some people
|
|
#
?
Feb 5, 2018 03:13
|
|
|
lmao if buttmining causes shortage of AMD motherboards
|
|
#
?
Feb 5, 2018 03:21
|
|
|
Anime Schoolgirl posted:lmao if buttmining causes shortage of AMD motherboards The 1700x also sold out but that seems unrelated, because they were selling them for $265 or less and they aren't really good for any flavor of butts. Khorne fucked around with this message at 03:47 on Feb 5, 2018 |
|
#
?
Feb 5, 2018 03:44
|
|
|
Setzer Gabbiani posted:Uh huh RX 550 isn't bad at all for iGPU, hopefully those 25%+ overclocking figures hold for final silicon. Is it Zen+ that's going on a non-mobile process?
|
|
#
?
Feb 5, 2018 03:53
|
|
|
NewFatMike posted:RX 550 isn't bad at all for iGPU, hopefully those 25%+ overclocking figures hold for final silicon. not really, is just an improve process that is already there.
|
|
#
?
Feb 5, 2018 05:19
|
|
|
Dratssss I'm hurting for Hertz!
|
|
#
?
Feb 5, 2018 06:15
|
|
|
Craptacular! posted:My understanding of motherboards right now is... ASRock is also really slow to release firmware updates. I wouldn't worry too much between brands, IMO the most important thing is the number of VRM phases. There's a decent temperature/stability difference between 8 and 10/12 phase
|
|
#
?
Feb 5, 2018 13:38
|
|
|
Khorne posted:My local microcenter went from 8 1920x+taichi in stock to 0 on saturday between 10am and noon at $913 for the combo. I guess I hopped onto the TR bandwagon just in time 
|
|
#
?
Feb 5, 2018 15:05
|
|
|
Khorne posted:My local microcenter went from 8 1920x+taichi in stock to 0 on saturday between 10am and noon at $913 for the combo. looks like amd marketing finally got the loving hint eh?
|
|
#
?
Feb 5, 2018 15:10
|
|
|
Can't wait for tr3 with 32 cores 
|
|
#
?
Feb 5, 2018 15:15
|
|
|
ufarn posted:That rep of Gigabyte replacing components in their products after the initial launch with positive reviews, is that still the case? The last time of note that this happened (around 2014), it was on a cheapo B85 motherboard and not one of the company's top-end enthusiast boards. Even at the time, the tell was a change in the revision number associated with the product (rev. 1.0 to rev. 2.0), so in theory it would be easy to spot if Gigabyte did make such a change today. A quick survey of all recent Gigabyte Intel boards available through Newegg right now suggests that they're all rev. 1.0 still. To find anything with a revision number other than 1.0 you have to go look at AMD boards and even then the only one with anything different is a bargain-basement 970-chipset board for AM3+ CPUs. There's nothing to say that Gigabyte isn't changing component specifications under the table, but it would be a PR disaster if they did and someone found out. At this point people should just check the revision number on the Gigabyte product they want and make sure that it's still 1.0 before buying.
|
|
#
?
Feb 5, 2018 15:23
|
|
|
Do gigabyte make anything compelling enough that it's worth the effort?
GRINDCORE MEGGIDO fucked around with this message at 16:33 on Feb 5, 2018 |
|
#
?
Feb 5, 2018 16:30
|
|
|
GRINDCORE MEGGIDO posted:Do gigabyte make anything compelling enough that it's worth the effort? They were (are?) a favorite for hackintosh builds.
|
|
#
?
Feb 5, 2018 16:58
|
|
|
GRINDCORE MEGGIDO posted:Do gigabyte make anything compelling enough that it's worth the effort? They make one of the 5 or 6 boards with 10/12 phase VRM and a clock generator (I think it's any of their X370 boards but double check first). Can confirm they own bones for hackintosh Nobody seems to like their X399 stuff, I know I'm not the only one waiting for better offerings
|
|
#
?
Feb 5, 2018 18:32
|
|
|
I don't see much point to buying any X399 except the Fatal1ty. The Fatal1ty is the only board with a 10GbE on the board and if you're going to charge $400 on a motherboard packaging a feature as a PCIe addon card is a loving ripoff. Unless you legitimately will never want 10 GbE, I guess, in which case it might be OK if you can find another board for under $300 (PcPP says the MSI SLI Plus is $280 AR atm).
|
|
#
?
Feb 5, 2018 19:16
|
|
|
Paul MaudDib posted:I don't see much point to buying any X399 except the Fatal1ty. The Fatal1ty is the only board with a 10GbE on the board and if you're going to charge $400 on a motherboard packaging a feature as a PCIe addon card is a loving ripoff. Unless you legitimately will never want 10 GbE, I guess, in which case it might be OK if you can find another board for under $300 (PcPP says the MSI SLI Plus is $280 AR atm). I'm unsure what I will buy. I might try to score used X399 stuff, or buy used a 1700x/1800x but use the x470 chipset with it. Or just buy new. We'll see!! I want a workstation separate from my i7 3770k system (which has better single core performance than any of amd's current offerings, sad).
|
|
#
?
Feb 5, 2018 19:24
|
|
|
So I have an ASROCK x370 board and the Gigabyte x399 board and have to say that both have been REALLY rock solid. Previous built as an i5-2500k/i7-3770k on a gigabyte board that ran super OC'd for years without a glitch. Gigabyte x399 board is running my Threadripper @ 4.1 GHz with just 1.3vcore so I'm pretty impressed with that. From a feature perspective I think the ASROCK board has a better UEFI Bios, but the Gigabyte board has better Audio + Network stack for gaming (SB Studio vs 720) (5GbE vs KILLER NIC).
|
|
#
?
Feb 5, 2018 19:29
|
|
|
Do you actually use the KILLER NIC? I found it to be an endless source of issues, disabled it and only use the Intel port now.
|
|
#
?
Feb 5, 2018 21:34
|
|
|
Khorne posted:I thought about this, but the truth is most people don't need 10GBE and by the time switches/routers are affordable you will be able to get consumer cards for $40-$60 tops. Even now, you can get consumer cards for around $100. It's not really about price, it's about wasting slots and x16 lanes and screwing up your airflow for something that can be integrated onboard via the PCH. Tons of slots+lanes is one of the big selling points of X399 and you're screwing that up right off the bat. Minor differences in cost don't matter all that much on a $1300 build - even if you could buy a 10 GbE NIC for $50 today, I'd rather drop the extra $50 on the motherboard and have it integrated. The fact that the Zenith Extreme is more expensive than the Fatal1ty is the final nail in the coffin IMO. And yeah if you don't care about 10 GbE and will never care about 10 GbE and can get a board for <$300 (again, like the MSI X399 SLI Plus) then by all means go for it, but most boards cluster around $350-400 anyway, and at that point the Fatal1ty is a no-brainer for $20-50 more IMO, it's easy future-proofing. Paul MaudDib fucked around with this message at 21:50 on Feb 5, 2018 |
|
#
?
Feb 5, 2018 21:45
|
|
|
Paul MaudDib posted:It's not really about price, it's about wasting slots and x16 lanes and screwing up your airflow for something that can be integrated onboard via the PCH. Tons of slots+lanes is one of the big selling points of X399 and you're screwing that up right off the bat. Minor differences in cost don't matter all that much on a $1300 build - even if you could buy a 10 GbE NIC for $50 today, I'd rather drop the extra $50 on the motherboard and have it integrated. The fact that the Zenith Extreme is more expensive than the Fatal1ty is the final nail in the coffin IMO. But then I'd have a motherboard named "Fatal1ty"
|
|
#
?
Feb 5, 2018 22:04
|
|
|
VostokProgram posted:But then I'd have a motherboard named "Fatal1ty" It's the reason I got the Taichi instead. Plus I'm saving my lunch money for some QDR infiniband off ebay if I find a good deal.
|
|
#
?
Feb 5, 2018 22:06
|
|
|
VostokProgram posted:But then I'd have a motherboard named "Fatal1ty"
|
|
#
?
Feb 5, 2018 23:34
|
|
|
GRINDCORE MEGGIDO posted:Do gigabyte make anything compelling enough that it's worth the effort? The Z370 Gaming 7 is fantastic, it regularly drops to under $200 and it's a great deal even at MSRP. It's probably the nicest Z370 overall.
|
|
#
?
Feb 6, 2018 00:22
|
|
|
New Zealand can eat me posted:Do you actually use the KILLER NIC? I found it to be an endless source of issues, disabled it and only use the Intel port now. I've not had any problems with it. I didn't bother with the Killer Nic software though. It's probably garbage anyway.
|
|
#
?
Feb 6, 2018 01:48
|
|
|
|
| # ? May 15, 2024 22:07 |
|
|
Stanley Pain posted:I've not had any problems with it. I didn't bother with the Killer Nic software though. It's probably garbage anyway. Mine was only ever good for like 300mbps, I wouldn't have noticed if it wasn't choking the gigabit internet out on speed tests
|
|
#
?
Feb 6, 2018 05:52
|
|