|
Arivia posted:it's okay to run operating systems that don't have the gameboy advance as a target platform, you know that right 
|
#
?
Aug 31, 2021 14:17
|
|
|
|
| # ? May 31, 2024 01:53 |
|
|
|
|
#
?
Sep 2, 2021 08:31
|
|
|
BlankSystemDaemon posted:Cheapest Danish retailer is charging the equivalent of $390. Order from mindfactory.de to mailboxde.com When ordering between 00-06 clock I think you get free shipping from mindfactory. You only have to pay postage from mailboxde.
|
|
#
?
Sep 2, 2021 11:33
|
|
|
Ihmemies posted:Order from mindfactory.de to mailboxde.com
|
|
#
?
Sep 2, 2021 18:41
|
|
|
Ffffffffffffffffffffffffffffffffffffffffuck that.
|
|
#
?
Sep 2, 2021 23:12
|
|
|
Hope those prices are a joke because lmao
|
|
#
?
Sep 2, 2021 23:19
|
|
|
as always, its gonna depend on the performance. the price inflation in CPU/GPUs sure seems here to stay.
|
|
#
?
Sep 3, 2021 00:32
|
|
|
I mean if they can somehow match 5950x performance with the top sku, wouldn't be totally unreasonable pricing. I am still super skeptical of heterogeneous design for desktop, but it would be neat to be wrong.
|
|
#
?
Sep 3, 2021 03:11
|
|
|
5950X is 100euro more than the 12900k on that list on the site it’s supposedly from, so that might be an indication of its expected multi core performance then again the 11900k price makes absolutely no sense compared to literally anything on the market so who knows 
|
|
#
?
Sep 3, 2021 03:26
|
|
|
True, it's possible Intel is just completely insane with pricing on the top sku. I wouldn't be surprised to see alder lake p cores beat zen3 cores in performance, and I wouldn't be surprised if they're more efficient as well. But you can't really justify a $700 sku on gaming performance when a $300 one does just as well, you really have to deliver in prosumer workloads that scale well with core count too. I have a hard time imagining a world where an 8p8e design can match 16c zen3 in a well threaded use case. They're gonna be leaning real hard on those efficiency cores.
|
|
#
?
Sep 3, 2021 04:00
|
|
|
VorpalFish posted:I wouldn't be surprised to see alder lake p cores beat zen3 cores in performance, and I wouldn't be surprised if they're more efficient as well. All the leakers are saying they're burning around ~250-300w to get the chip to ~5.2Ghz+, which is where those real nice performance numbers are coming from. It wouldn't surprise me if at lower clocks (3Ghz or less) that they'd be more efficient than Zen3, and should make them great for laptops, but at high clocks they seem to be every bit as bad if not worse than RKL. My WAG is its a process issue and not a design issue due to how bad the power scaling is at those clocks.
|
|
#
?
Sep 3, 2021 04:39
|
|
|
PC LOAD LETTER posted:All the leakers are saying they're burning around ~250-300w to get the chip to ~5.2Ghz+, which is where those real nice performance numbers are coming from. This is almost certainly short duration limits. They should be able to get mid-high 4s all core well inside a 125w envelope and if they actually do get a 19% IPC uplift plus efficiency gains from 10 esf / Intel 7 instead of 14, they could end up quite efficient on a per core basis.
|
|
#
?
Sep 3, 2021 04:51
|
|
|
PC LOAD LETTER posted:All the leakers are saying they're burning around ~250-300w to get the chip to ~5.2Ghz+, which is where those real nice performance numbers are coming from. yeah, somewhat ironically for these "mobile"-ish nodes (at least on the TSMC side), Intel seems to have things very under control at higher power/clock rates, Tiger Lake clocks exceptionally well in the higher power SKUs, but was kinda disappointing in the lower power SKUs, and leaks on Sapphire Rapids (10ESF) and desktop Alder Lake (10ESF) both seem to point to power scaling at lower frequencies still being a problem. it's rather unfortunate for Intel that they are behind the curve on chiplet technologies/architectures to implement them, because they are kinda forced to lead with laptop right now (and eventually server chiplets like Sapphire Rapids) because that's what they can yield, while AMD is building servers out of something that's smaller than a phone SOC. Desktop is still really the right fit for those power/frequency characteristics but they have to yield much bigger chips than AMD.
|
|
#
?
Sep 3, 2021 04:54
|
|
|
PC LOAD LETTER posted:All the leakers are saying they're burning around ~250-300w to get the chip to ~5.2Ghz+, which is where those real nice performance numbers are coming from. Rocket Lake also had leaks with single core scores beating Zen 3 so we'll see how the performance stacks up once it actually launches.
|
|
#
?
Sep 3, 2021 08:35
|
|
|
In a year's time getting *any* CPU will likely make everyone pine for the darkest days of 6-8 months ago. My guess is Intel knows they didn't have to beat AMD this time around. They know they'll sell everything they make eventually.
|
|
#
?
Sep 3, 2021 08:42
|
|
|
CPU supply is not in the same place as GPU supply. Dies are smaller and they don't have crypto to drive demand. Even zen3 seems to be available at msrp with no effort now. (Maybe this is my US bias). Current pricing on the 11400, 11600k and 11700k plus the fact that they unlocked faster memory on the lower end chipsets suggests someone at Intel understands that if they aren't ahead at 1c, multicore, or efficiency then they have to compete on value and price competitively. The 11900k of course is absurdly priced for what it is but everything else is in line with their position in the market. If they're able to retake the lead in gaming, they might crank prices to parity or above AMD again, but a $700 sku that isn't good in heavily threaded workloads is in a very bad position. Edit: sorry, meant the 11700 non k. 11700k is also overpriced. VorpalFish fucked around with this message at 14:28 on Sep 3, 2021 |
|
#
?
Sep 3, 2021 14:24
|
|
|
Both have numbers that go up, so clearly they're the same, according to every benchmark ever.
|
|
#
?
Sep 3, 2021 15:26
|
|
|
I'd like to think the i5-12400 is going to be relatively cheap, but it's gonna suck if the rumors are true and the non-K chips are big cores only and you need the i5-12600K to get the 6 big / 4 little set-up
|
|
#
?
Sep 3, 2021 15:32
|
|
|
risc-v fans gets hyped https://www.tomshardware.com/news/apple-looking-for-risc-v-programmers quote:Apple is currently looking for experienced programmers with detailed knowledge of the RISC-V Instruction Set Architecture (ISA) and Arm's Neon vector ISA for its Vector and Numerics Group (VaNG) within its Core Operating Systems group. Apple's VaNG is responsible for developing and improving various embedded subsystems running on iOS, macOS, watchOS, and tvOS.
|
|
#
?
Sep 3, 2021 15:46
|
|
|
It'll be pretty funny if by the time there's major ARM on desktop adoption that Apple may already move to RISC-V. They'll probably do a decade of development again if they ever move over to it, but it's still pretty exciting.
|
|
#
?
Sep 3, 2021 16:10
|
|
|
obligatory “RISC......RISC is good”
|
|
#
?
Sep 3, 2021 17:00
|
|
|
Sounds like Alder Lake S will get announced on October 27-28th, with mass availability (along with Z690 and DDR5) on November 19th.
|
|
#
?
Sep 3, 2021 23:28
|
|
|
the apple cpu tech stuff is interesting in an abstract sense but their ipads having m1 chips kinda summarizes my issues with them you have extremely powerful hardware which you can't do much with. and it applies to their "desktop" hardware as well since you can't game or do most mainstream compute tasks on mac os
|
|
#
?
Sep 4, 2021 00:34
|
|
|
It can run WebKit browsers and outlook fine, that’s what “mainstream” tasks have been for the past what 10 years.
|
|
#
?
Sep 4, 2021 00:49
|
|
|
it's unfortunate that apple and nvidia are enemies or whatever having nvidia gpu cards paired with m1 chips would probably have led to better (un)official gaming / GPU compute support
|
|
#
?
Sep 4, 2021 00:52
|
|
|
shrike82 posted:the apple cpu tech stuff is interesting in an abstract sense but their ipads having m1 chips kinda summarizes my issues with them I intentionally chose a MacBook Air as my development machine this time around, and it's incredibly fast at compiling, testing, and using heavyweight IDEs like IntelliJ and CLion. Easily noticeably faster than the 6 core 35W Intel processor I had before, and the Air is fanless to boot. I'm not going to pretend that it's all roses though, for software development now I have to think about and choose if I want to build and run things as ARM or x86, and the answer is not always ARM even if ARM code and binaries are available.
|
|
#
?
Sep 4, 2021 00:55
|
|
|
I'm looking forward to picking up the M1 iPad Pro - my old one was great, and you can do plenty of CAD and design on there. Sounds like some reviewers have gotten video editing out of it, but I can't really speak to that. It's nice having a mono tasking device, especially when I'm writing. Literally not sure why anyone else would use it that way, but maybe there's a enough brokebrains like me out there to form a market segment
|
|
#
?
Sep 4, 2021 01:04
|
|
|
shrike82 posted:the apple cpu tech stuff is interesting in an abstract sense but their ipads having m1 chips kinda summarizes my issues with them aside from running Windows games (which the M1 can do surprisingly well via poo poo like CrossOver or even a VM) what “mainstream compute tasks” can’t MacOS do?
|
|
#
?
Sep 4, 2021 01:18
|
|
|
I have a 2018 iPad Pro and was looking at the m1 version but by all accounts, you’re not going to notice much of a difference even if you try to game on the thing let alone just browse the web. I don’t think even battery life has benefited significantly
|
|
#
?
Sep 4, 2021 01:20
|
|
|
M1 iPad Pro is proof that Apple would rather hobble iOS than start blurring the line between the iPad and the Mac It could be such a fantastic device but they make it fall so short of what it It literally has the CPU/GPU of an iMac in it, but you’d be hard pressed to tell vs its predecessor unless you’re one of the five youtubers pushing LumaFusion hard enough to notice trilobite terror fucked around with this message at 01:32 on Sep 4, 2021 |
|
#
?
Sep 4, 2021 01:29
|
|
|
hobbesmaster posted:It can run WebKit browsers and outlook fine, that’s what “mainstream” tasks have been for the past what 10 years. The iOS/Mac version of Outlook and Office in general is reason enough for many people (myself included) to never want to have to use a Mac in the office. (I still use a Mac because almost everything else I do is 'nix based, so it makes a lot of sense, but...still...gently caress the Mac Office design language)
|
|
#
?
Sep 6, 2021 23:50
|
|
|
When I switched jobs I was given a MacBook and was all “wait no yum or apt-get…what the gently caress is brew”
|
|
#
?
Sep 7, 2021 00:05
|
|
|
Just run VMware Fusion. Except the whole issue with Big Sur and Fusion's NAT being pretty broken currently requiring a bunch of hacking around with sysctl still a year after Big Sur's release. So... Parallels works without issue at least.
|
|
#
?
Sep 7, 2021 15:37
|
|
|
Intels Rocket Lake Xeon W’s launched, which means it’s time to see what weird poo poo Asrock Rack is up to! [Img]https://i.imgur.com/kGdUeCf.jpg[[/Img] 4x SO-DIMM (ECC or non) 2x X550 10gbe ports BMC + out of band management port 2x OCuLInk (4x SATA or 4x PCIe 3 lanes each) 12VO Like we talked about earlier, the 12VO portion of the board takes up a pretty big chunk of the board even with only 1 SATA breakout, and necessitates the OCuLInk. Can’t see them bothering with that cost on future consumer boards, but who knows when/if 12VO will get real momentum in the consumer side.
|
|
#
?
Sep 9, 2021 07:47
|
|
|
Cygni posted:Like we talked about earlier, the 12VO portion of the board takes up a pretty big chunk of the board even with only 1 SATA breakout, and necessitates the OCuLInk. Can’t see them bothering with that cost on future consumer boards, but who knows when/if 12VO will get real momentum in the consumer side.  Please, mspaint that up and annotate where you think 8x sata connectors plus a 24-pin molex minifit jr can fit after removing the 12V power connectors and some buck converters. The reason why that board has 2xOculink for breakout isn't 12VO, it's that ASRock stuffed a shitload of features and RAM sockets into a mini-ITX board.(btw, apparently that's not even proper ATX 12VO, that line of boards seems to predate it.)
|
|
#
?
Sep 9, 2021 10:57
|
|
|
I wasn’t implying that they were going to put 8 SATA plugs on the board homie relax
|
|
#
?
Sep 9, 2021 16:58
|
|
|
I use a lot of oculink cables in my work and let me tell you: I really hate oculink connectors. The retention is such a bastard sometime and the surface mount connectors without through hole ground pins are fragile as gently caress. I’m glad it’s probably going to die out.
|
|
#
?
Sep 9, 2021 18:03
|
|
|
Cygni posted:I wasn’t implying that they were going to put 8 SATA plugs on the board homie relax Still a valid question, though. A 24-pin PSU connector is only slightly smaller than one of those SO-DIMM slots. Similarly, you could swap out the OCLink slot with maaaaybe 2 SATA ports. So even if they went "traditional" design, there'd be no way to get more than like 4 SATA ports on that board without getting super obnoxious, and simply no way to get another PCIe connection on there. OCLink was a foregone conclusion even before looking at 12VO, just based on the ports they wanted to shove in there. That said, I agree I wouldn't expect to see similar things in consumer boards, but moreso because consumers sure as poo poo aren't a great market for OCLink in the first place than anything to do with 12VO.
|
|
#
?
Sep 9, 2021 18:08
|
|
|
Now do another in mATX, cowards!!!
|
|
#
?
Sep 9, 2021 18:10
|
|
|
|
| # ? May 31, 2024 01:53 |
|
|
movax posted:Now do another in mATX, cowards!!!
|
|
#
?
Sep 9, 2021 18:23
|
|