|
So that's why they were called tech decks..
|
#
?
Apr 9, 2019 13:19
|
|
|
|
| # ? Apr 24, 2024 06:28 |
|
|
lol its a pretty good sign that intel is hitting some hard limits when they get the old pentium d superglue out
|
|
#
?
Apr 9, 2019 13:49
|
|
|
Why not just have 2 sockets?
|
|
#
?
Apr 9, 2019 14:22
|
|
|
Khorne posted:Why not just have 2 sockets? marketing
|
|
#
?
Apr 9, 2019 15:00
|
|
|
Just bought an AMD R3 2200G system to build, if my SSD and case isn't shot from a lot of liquid being spilled on it, then it will have cost ~£250 and do exactly what I need  Was always an AMD man back in the day, Athlon XP 2500+ forever!!
|
|
#
?
Apr 9, 2019 15:26
|
|
|
Lambert posted:Instead of calling it "glued together", Intel settled for calling the process "foveros". foveros is 3d stacking with chiplets, so its def not the new Xeon which is just 2 monolithic dies on one package
|
|
#
?
Apr 9, 2019 17:01
|
|
|
Khorne posted:Why not just have 2 sockets? Number must go up. Always up. If number not go up, find way to make number go up, even if dumb. Number must go up. 
|
|
#
?
Apr 9, 2019 17:08
|
|
|
Khorne posted:Why not just have 2 sockets? I mean, AMD is doing the same thing with the Ryzen and Epycs: Its two dies, technically two CPUs, sharing a single package and socket. It just seems weird that this is Intel's path forward.
|
|
#
?
Apr 9, 2019 17:24
|
|
|
Number19 posted:Number must go up. Always up. If number not go up, find way to make number go up, even if dumb. Number must go up. tell that to microsofts marketing team
|
|
#
?
Apr 9, 2019 18:55
|
|
|
How much is driven by the market's needs being forced by per-socket software licensing?
|
|
#
?
Apr 9, 2019 19:01
|
|
|
Pablo Bluth posted:How much is driven by the market's needs being forced by per-socket software licensing? You mean, "how much is this going to kill per-socket licensing in favor of per-core licensing". A lot. The answer is a lot, and a lot of the names you know about have already moved there. Because you're starting to look at a single socket taking over for four at the most extreme case. That could mean potentially 75% drop in revenues, and that would kill anyone operating on a per-socket license scheme if allowed to become widespread.
|
|
#
?
Apr 9, 2019 19:20
|
|
|
What even uses per socket licensing anymore? Everything I can think of switched to per core a while ago?
|
|
#
?
Apr 9, 2019 19:33
|
|
|
Xae posted:What even uses per socket licensing anymore? Veeam uses per socket, though they might be fine sticking with that since they mostly care about licensing hosts.
|
|
#
?
Apr 9, 2019 19:39
|
|
|
Xae posted:What even uses per socket licensing anymore? vSphere Platinum license kits are still per socket. Veeam is as well. I think Oracle moved to per-core but they have some funky math that you have to do sometimes that varies the cores per license based on processor generation and possibly socket count as well.
|
|
#
?
Apr 9, 2019 19:40
|
|
|
Yeah, Oracle has a weird matrix now where you apply a co-efficient to the core count so if its older generation hardware with less per-core performance they ding you slightly less than newer kit with the same density.
|
|
#
?
Apr 9, 2019 20:02
|
|
|
Per-
|
|
#
?
Apr 9, 2019 20:38
|
|
|
Stickman posted:Per- Pay as you go! Just $0.00000128 per CPU cycle! I wrote this without doing math but I'm pretty sure this is over $1/second to run. I guess this would be sold in "1Ghz/hr" blocks or something to make it easier than just cycles.
|
|
#
?
Apr 9, 2019 20:40
|
|
|
SwissArmyDruid posted:love 2 throw my $MONEY cpu in the trash when my motherboard dies. I had to double-check the date to make sure it wasn't an april fool's joke, with Cutress's gormless grin there. If I was buying such an expensive system that wouldn't be my problem. I contact the support, next day a guy shows up in the datacenter with a new motherboard and CPU.
|
|
#
?
Apr 9, 2019 21:02
|
|
|
Number19 posted:Number must go up. Always up. If number not go up, find way to make number go up, even if dumb. Number must go up. <impolite reference to one of the products of the company you work for not having numbers that are going up><acknowledgement that you are in no way responsible for that product><also acknowledgement that I still frequently use that product voluntarily> The target for that CPU is a highly specialized compute market that absolutely has a time = money driver, and this is never going to be a CPU that anyone buys individually or would ever have bought individually, it will always have been part of a larger 'solution' from a service provider that included the entire product stack of compute, storage, networking, management, and software. What we think of it isn't relevant at all, and even the people ultimately buying it may never know or care what the CPU model is.
|
|
#
?
Apr 10, 2019 00:59
|
|
|
SwissArmyDruid posted:You mean, "how much is this going to kill per-socket licensing in favor of per-core licensing". its all about per thread licensing.
|
|
#
?
Apr 10, 2019 01:01
|
|
|
EoRaptor posted:<impolite reference to one of the products of the company you work for not having numbers that are going up><acknowledgement that you are in no way responsible for that product><also acknowledgement that I still frequently use that product voluntarily> I doubt anyone other than large service providers will even look at that thing. It’s a specialty part that will serve a couple of niches like you said. It’s in no way a general purpose part and Intel is not expecting to move any kind of volume with them. It’s probably going to require specially tuned code and OS kernels to make proper use of it and I’m willing to bet you there all sorts of trade offs required to optimize for it. e: it’s almost certainly going to have to present multiple NUMA nodes from a single socket so maybe they can take advantage of all the groundwork AMD laid optimizing stuff for that when Epyc/Threadrippers launched 
Number19 fucked around with this message at 01:26 on Apr 10, 2019 |
|
#
?
Apr 10, 2019 01:23
|
|
|
Does Ryzen Master not have options to launch on startup and apply profiles automatically? The Gigabyte B450I Aorus Pro Wifi I bought for some SFF PC has the worst BIOS I've ever seen and won't even let me change 99% of the CPU's settings (I can mess with RAM timings and that's basically it) and I can work around that with Ryzen Master. Or could, if I didn't have to manually launch the program and apply profiles every single time 
|
|
#
?
Apr 10, 2019 22:44
|
|
|
Win+R > shell:startup lets you add a program shortcut to make it start with Windows. But as for automatically applying a profile, I don't know. I use BIOS myself and don't know what the default behaviour is.
|
|
#
?
Apr 10, 2019 22:58
|
|
|
Having seen reviews of that motherboard being OCed, I think the problem with the BIOS might be the user. You either need to read better, or you don’t have the board you think you do. It should be a good OCer, though with a Gigabyte board you want to be conservative with voltage because voltage equipment reveals it delivers more power than the software reports, not less like other manufacturers.
|
|
#
?
Apr 10, 2019 23:10
|
|
|
Craptacular! posted:Having seen reviews of that motherboard being OCed, I think the problem with the BIOS might be the user. You either need to read better, or you don’t have the board you think you do. It should be a good OCer, though with a Gigabyte board you want to be conservative with voltage because voltage equipment reveals it delivers more power than the software reports, not less like other manufacturers. I saw the reviews too, which is why I bought it, but no, it's the right board and all the fields I'd expect to edit for overclocking (rather: underclocking) refuse to be changed (some are straight greyed-out but apparently they do lock a handful of OC options on B450 boards). Since Ryzen Master is no real help (I mean, it helps and has great settings but I'm not manually picking and applying a profile every time I launch the computer) I'll just go through all the options again because it's probably hidden behind one option in some sub menu. E: This person on Reddit apparently had the same problem, I have no idea if they ever fixed it though orcane fucked around with this message at 09:01 on Apr 11, 2019 |
|
#
?
Apr 11, 2019 08:20
|
|
|
I have a gigabyte board and I oc through the bios. Try to downgrade/upgrade your bios, or check the gigabyte motherboard forums for comments about whether that is working in your current bios version.
|
|
#
?
Apr 11, 2019 15:42
|
|
|
Turns out some options can only be changed with the +/- keys, which I didn't even try yesterday since it only applies to very few options and I was navigating the UEFI with my mouse. Most stuff (and everything in AsRock and Asus UEFIs I used for the past ~decade) can be clicked on or toggled through with enter, but most (but not all) of the Gigabyte's OC settings don't do anything if you do either of those. Navigating all voltage increments etc. like that is a pain but at least it works 
|
|
#
?
Apr 11, 2019 16:01
|
|
|
orcane posted:Turns out some options can only be changed with the +/- keys, which I didn't even try yesterday since it only applies to very few options and I was navigating the UEFI with my mouse. Most stuff (and everything in AsRock and Asus UEFIs I used for the past ~decade) can be clicked on or toggled through with enter, but most (but not all) of the Gigabyte's OC settings don't do anything if you do either of those. Navigating all voltage increments etc. like that is a pain but at least it works Gigabyte's janky rear end BIOS settings are one reason why I didn't go with an Aorus mobo for my build.
|
|
#
?
Apr 11, 2019 16:53
|
|
|
orcane posted:Turns out some options can only be changed with the +/- keys, which I didn't even try yesterday since it only applies to very few options and I was navigating the UEFI with my mouse. Most stuff (and everything in AsRock and Asus UEFIs I used for the past ~decade) can be clicked on or toggled through with enter, but most (but not all) of the Gigabyte's OC settings don't do anything if you do either of those. Navigating all voltage increments etc. like that is a pain but at least it works You can just type in the values (at least on the Gigabyte boards I've used), you don't have to cycle through the increments.
|
|
#
?
Apr 11, 2019 18:55
|
|
|
I have an aorus and typed mine  F/e: that is on my Intel PC though not my AMD one
|
|
#
?
Apr 11, 2019 21:46
|
|
|
+1 for just typing in values on Gigabyte boards. I have exclusively purchased Gigabyte boards for since the Socket A days, I only own 2 ASUS boards because one was given to me and the other was second hand and a surprise purchase (wrong board mailed). I feel ASUS is a good brand, but a little too gamery, like one board has a button on the board to boot directly into the BIOS settings. I guess that is kinda cool but I am certain I will never actually use that button since I rarely change my BIOS settings once everything is all setup. If typing in at the BIOS is not enough, I can use Gigabyte software to key in the values from within Windows. Or with ASUS I can use a special double ended USB cable to run ROG connect on another PC to tweak BIOS settings. . . Do people seriously do this?
|
|
#
?
Apr 12, 2019 20:32
|
|
|
I think Haswell was the last generation where every mobo wasn't drowned in XTREME GAMING labels outside of low-end. I may be wrong here regarding Gigabyte's dual-bios but I feel its just a gimmick that creates more bugs than providing actual convenience.
|
|
#
?
Apr 13, 2019 03:14
|
|
|
.
sincx fucked around with this message at 05:50 on Mar 23, 2021 |
|
#
?
Apr 13, 2019 03:36
|
|
|
I have never actually used the dual BIOS, but it has never caused a problem for me either.
|
|
#
?
Apr 13, 2019 04:14
|
|
|
Palladium posted:I think Haswell was the last generation where every mobo wasn't drowned in XTREME GAMING labels outside of low-end. Don't forget about LED lightshows. My MSI mobo was about as low-key as it got without losing the RGB LEDs.
|
|
#
?
Apr 13, 2019 05:07
|
|
|
Palladium posted:I may be wrong here regarding Gigabyte's dual-bios but I feel its just a gimmick that creates more bugs than providing actual convenience. Same thing with BIOS debug LED's on mobos. Not really used often by most but they're incredibly useful for when you do need them.
|
|
#
?
Apr 13, 2019 05:53
|
|
|
It seems like it'd mostly only be useful for making incremental changes that might cause boot failure. Every board should have some way to clear the CMOS that'll fix that - it'd just be more work to get your stable settings back.
|
|
#
?
Apr 13, 2019 08:52
|
|
|
Crotch Fruit posted:a little too gamery, like one board has a button on the board to boot directly into the BIOS settings How do people use that with gaming? Is this like the days of having custom config.sys/autoexec.bat boot disks for different games, but for RAM timings? Please tell me it is, that would be so amazing.
|
|
#
?
Apr 13, 2019 08:59
|
|
|
I remember being terrified while growing up about a bios/firmware update going wrong and bricking xyz device but I have never had anything go other than flawlessly (saying that, it will happen tomorrow  )
|
|
#
?
Apr 13, 2019 10:28
|
|
|
|
| # ? Apr 24, 2024 06:28 |
|
|
Palladium posted:I may be wrong here regarding Gigabyte's dual-bios but I feel its just a gimmick that creates more bugs than providing actual convenience. The variant with dual bios but no physical bios switch that switches automatically on bluescreens etc, yes, absolutely. It's terrible. The ones with a physical switch are fine.
|
|
#
?
Apr 13, 2019 14:12
|
|
