|
gradenko_2000 posted:I used an A10-7860K for about a week while sorting out some troubleshooting woes and you could really feel it chug just opening browsers and such. buddy there's enough security endpoint crap to choke even the fastest client hardware these days
|
# ? Jul 12, 2023 13:12 |
|
|
# ? Jun 9, 2024 06:31 |
|
Klyith posted:I still use an A9-something craptop. And get this, most of the time I limit the CPU speed to 70%. I absolutely believe it. Last night I got YouTube running on Ubuntu and it’s amazing how hard those two cores work to just sit on a webpage and play 720p video. If you haven’t installed h264ify, that will probably help on the IGP, but I haven’t exhaustively verified that. It skins my nose a little that this thing is running single-channel memory, but I happen to have two 4GB SO-DIMMs left over from an old project. Later today I’ll swap RAM and see what a difference it makes. I’ll verify that it took at least 20 minutes to compile GZDoom, stock.
|
# ? Jul 12, 2023 13:57 |
|
I had a Llano A8 from the first iteration of AMD's "APU," idea. I even upgraded the system with a fanless Radeon HD 6670, which allowed you to crossfire between the iGPU and the card for an extra ~20% peak fps performance and frequently dropped frames as DDR3 1866 and GDDR5 shared the framebuffer. Bad decisions all around. Intel was just better in every respect back in that era.
|
# ? Jul 12, 2023 14:22 |
|
Hasturtium posted:I’ll verify that it took at least 20 minutes to compile GZDoom, stock. I would love to see how long it takes to compile binutils and gcc on this thing
|
# ? Jul 12, 2023 16:32 |
|
Some rumors about the core counts of Raptor Lake Refresh are going around, its from RedGaming so... you know.. 8+8 on the 600 sku is p neat if it comes to pass.
|
# ? Jul 12, 2023 22:55 |
|
Cygni posted:8+8 on the 600 sku is p neat if it comes to pass. I'm really skeptical of that because meteor lake only goes up to 6 P cakes, I'd bet whatever comes after it is still 8 P cores on a module. If this comes true, this is a hell of a refresh value wise.
|
# ? Jul 12, 2023 23:48 |
|
The i3 tier (I'm guessing the bottom of the chart has a typo) going up to 6 cores is tasty as hell.
|
# ? Jul 12, 2023 23:53 |
|
Twerk from Home posted:I'm really skeptical of that because meteor lake only goes up to 6 P cakes, I'd bet whatever comes after it is still 8 P cores on a module. If this comes true, this is a hell of a refresh value wise. Meteor Lake is only going to be on mobile though, and they've already split the core counts and naming on mobile vs desktop as it is. Dunno if what Meteor Lake is doing in laptops will impact the desktop SKU positioning much.
|
# ? Jul 13, 2023 00:18 |
|
gradenko_2000 posted:The i3 tier (I'm guessing the bottom of the chart has a typo) going up to 6 cores is tasty as hell. I thought it's something they should have done last gen. Puts them in a much better spot to compete with the 5600 for a low end platform.
|
# ? Jul 13, 2023 00:49 |
|
Kazinsal posted:I would love to see how long it takes to compile binutils and gcc on this thing It’s pokey on vanilla Ubuntu, though it picked up a smidge switching to Xubuntu + lightdm. I fear how pokey this would be trying to lug Windows 10 around - just having a few Firefox tabs open makes it chug. I am… tempted by your proposal. Let me kick it around a bit first. Also, genuinely impressed that dual channel memory doesn’t seem to have made much difference.
|
# ? Jul 13, 2023 01:50 |
|
Hasturtium posted:It’s pokey on vanilla Ubuntu, though it picked up a smidge switching to Xubuntu + lightdm. I fear how pokey this would be trying to lug Windows 10 around - just having a few Firefox tabs open makes it chug. I am… tempted by your proposal. Let me kick it around a bit first. Also, genuinely impressed that dual channel memory doesn’t seem to have made much difference. If you want something fairly reproducible for comparison then a buddy of mine has a script set that can produce gcc cross toolchains easily: https://github.com/travisg/toolchains Each arch passed to -a causes a whole new compile/link cycle for the whole suite so you’d want to time eg. just x86_64.
|
# ? Jul 13, 2023 02:02 |
|
Kazinsal posted:If you want something fairly reproducible for comparison then a buddy of mine has a script set that can produce gcc cross toolchains easily: https://github.com/travisg/toolchains Y’know, this would be riotously funny to run and compare on my eight core Power9 versus this thing. 32 threads of screaming ppc64le versus… this.
|
# ? Jul 13, 2023 02:07 |
|
Hasturtium posted:Y’know, this would be riotously funny to run and compare on my eight core Power9 versus this thing. 32 threads of screaming ppc64le versus… this. That’d be brilliant. The script can only guess at 2-way SMT so for 4-way SMT you’ll have to pass -j32 as well.
|
# ? Jul 13, 2023 02:15 |
|
Cygni posted:Some rumors about the core counts of Raptor Lake Refresh are going around, its from RedGaming so... you know.. Has the “MAKE MORE CORES” meme completely reversed?
|
# ? Jul 13, 2023 02:29 |
|
So the 14600K is going to have the same core config as the 13700K, which was the same as the 12900K. Both of those had stock power limits of 240W and would happily pull that much under full load. Please tell me the 14600K won't also be like that. What kind of i5 buyer wants to deal with that?
|
# ? Jul 13, 2023 09:35 |
|
Dr. Video Games 0031 posted:So the 14600K is going to have the same core config as the 13700K, which was the same as the 12900K. Both of those had stock power limits of 240W and would happily pull that much under full load. Please tell me the 14600K won't also be like that. What kind of i5 buyer wants to deal with that? It's okay, it'll be on 4nm instead of 7nm, so it'll only pull 200W! I look forward to the 14900KS using more power than an RTX 4080
|
# ? Jul 13, 2023 09:40 |
|
Kazinsal posted:That’d be brilliant. The script can only guess at 2-way SMT so for 4-way SMT you’ll have to pass -j32 as well. Just FYI, in poking through the script the command used to determine -j’s default value accurately returned 32 on the Power9. I’ll let you know numbers after the grinding has concluded, but the A9 basically feels like a Core 2 Duo, and the Power9 is closer to a 3950x.
|
# ? Jul 13, 2023 12:53 |
|
Kazinsal posted:It's okay, it'll be on 4nm instead of 7nm, so it'll only pull 200W! The 14600k will be Raptor Lake again, so it’s still Intel 7. The wildcard might be if Intel activates DLVR, which may significantly help power efficiency.
|
# ? Jul 13, 2023 15:30 |
|
Cygni posted:Meteor Lake is only going to be on mobile though, and they've already split the core counts and naming on mobile vs desktop as it is. Dunno if what Meteor Lake is doing in laptops will impact the desktop SKU positioning much. this is another case of “leakers have thoroughly covered their bases by speculating every reasonable possibility and will draw the bullseye around the one that hits”. Some rumors actually have lower tier desktop products as being meteor lake, possibly in the first half of next year. Honestly the relative lack of clarity for something that’s supposed to come out in this quarter probably points to it not being meteor lake (whole new processor tends to produce enough smoke at the partners to be noticed) and perhaps argues against DLVR (one would think you’d need some voltage stuff patched into the bios that would potentially be noticed). On the other hand they gotta be getting the power savings to bump core count from somewhere, unless they’re just going Full Inferno 300W power draw. I’d put it maybe 25% there’s a meteor lake partial release 1H next year, maybe 60/40 if Raptor Lake Refresh has the DLVR.
|
# ? Jul 13, 2023 20:54 |
|
Meteor Lake S was supposed to use LGA1851 as well, and if the board partners are really about to launch a whole slew of new boards with a new socket, they have certainly done a much better job than usual keeping that under wraps. Also Intel's own (simplified) public roadmaps don't have any Meteor Lake-S on em, but you never know.
|
# ? Jul 13, 2023 20:59 |
|
Kazinsal posted:That’d be brilliant. The script can only guess at 2-way SMT so for 4-way SMT you’ll have to pass -j32 as well. All right, numbers have been run. The time needed to compile the x86_64 version of the GCC 13.1.0 + binutils toolchain with the helpful link you posted: IBM Power9, 8 cores 32 threads (-j 32), 32GB DDR4-2666 dual-channel, Samsung 870 256GB SATA SSD real 9m45.233s user 124m55.872s sys 3m58.117s AMD A9 9400, 2 cores(ish) 2 threads (-j 4 - don't ask), 8GB DDR4-1600 dual-channel, 128GB generic SATA M.2 real 80m30.175s user 127m45.480s sys 17m51.778s If anything, I'm a little surprised the A9 didn't take even longer.
|
# ? Jul 13, 2023 21:21 |
|
gradenko_2000 posted:The i3 tier (I'm guessing the bottom of the chart has a typo) going up to 6 cores is tasty as hell. will they even work with the cheap OG LGA1700 mobos
|
# ? Jul 14, 2023 12:03 |
|
Palladium posted:will they even work with the cheap OG LGA1700 mobos Yes, most LGA1700 boards have already been updated for 14th gen. All Giga boards have already been updated, for example. (with the asterisk about DLVR support, which is still unconfirmed if it will be turned on or how it will work)
|
# ? Jul 14, 2023 15:35 |
|
https://twitter.com/ServeTheHome/status/1681445261762236416 ASUS taking over the manufacturing/support for current NUCs, standing up a business unit to take the brand moving forward.
|
# ? Jul 19, 2023 00:39 |
|
USB4 and occulink connectors have made AMD mini PCs pretty viable these days
|
# ? Jul 19, 2023 00:47 |
|
Sᴀɴᴛᴀ Cʟᴀʀᴀ (Bloomberg Business Wire) Intel today moved to sell its unprofitable "Intel" brand of CPUs, graphics chips, and semiconductor fabs. Going forward the company will concentrate on the remaining profitable sectors: bunnysuit plushies, youtube 90s tech nostalgia channels, and royalties from the valuable "Moore's Law" trademark.
|
# ? Jul 19, 2023 01:02 |
|
shrike82 posted:USB4 and occulink connectors have made AMD mini PCs pretty viable these days a combination oculink and thunderbolt egpu dock hit the market, and after benchmarks, it's not even funny how much bandwidth thunderbolt overhead consumes, which makes one wonder, why consumers even let it get to market dominance in the first place.
|
# ? Jul 19, 2023 01:55 |
|
SwissArmyDruid posted:a combination oculink and thunderbolt egpu dock hit the market, and after benchmarks, it's not even funny how much bandwidth thunderbolt overhead consumes, which makes one wonder, why consumers even let it get to market dominance in the first place. I have some bad news about USB
|
# ? Jul 19, 2023 02:04 |
|
SwissArmyDruid posted:a combination oculink and thunderbolt egpu dock hit the market, and after benchmarks, it's not even funny how much bandwidth thunderbolt overhead consumes, which makes one wonder, why consumers even let it get to market dominance in the first place. I don't think consumers had much say in the matter. I'd also somewhat question TB having "dominance". Apart from being widely used in a corporate setting for laptop docks, it really doesn't have much penetration. Consumer usage of TB devices has to be extremely minimal.
|
# ? Jul 19, 2023 04:55 |
|
I had always assumed that thunderbolt just passed the pcie lanes through but I guess there is more to it than that. Also lol oculink, rest in piss
|
# ? Jul 19, 2023 05:56 |
|
USB wouldn’t suck if you decided not to have a protocol that allows companies to sell the cheapest rear end garbage tier gizmos that can talk to a computer. But that kind of defeats the purpose of it WhyteRyce fucked around with this message at 07:05 on Jul 19, 2023 |
# ? Jul 19, 2023 07:01 |
Cygni posted:https://twitter.com/ServeTheHome/status/1681445261762236416
|
|
# ? Jul 19, 2023 09:56 |
|
do not pursue NUC BU
|
# ? Jul 19, 2023 11:06 |
|
Rock Solid. NUC Touching.
|
# ? Jul 19, 2023 12:09 |
|
WhyteRyce posted:USB wouldn’t suck if you decided not to have a protocol that allows companies to sell the cheapest rear end garbage tier gizmos that can talk to a computer. But that kind of defeats the purpose of it I like using cdc-ecm as an example of what issues crop up with usb protocols, Wikipedia has a good article on Ethernet over USB which includes: quote:Of these protocols, ECM could be classified the simplest—frames are simply sent and received without modification one at a time. This was a satisfactory strategy for USB 1.1 systems (current when the protocol was issued) with 64 byte packets but not for USB 2.0 systems which use 512 byte packets. USB: the worst interface besides all the others
|
# ? Jul 19, 2023 17:28 |
Just give me a serial interface that I can connect mpd to via netgraph(4).
|
|
# ? Jul 19, 2023 17:44 |
|
USB2 being a broadcast bus has a lot to do with some of the protocol jank (periods of expected idle as example). It being a broadcast bus I think also stems from being designed for cheap rear end poo poo
|
# ? Jul 19, 2023 18:07 |
|
BlankSystemDaemon posted:Just give me a serial interface that I can connect mpd to via netgraph(4). CDC-ACM has the same issues
|
# ? Jul 19, 2023 18:09 |
|
SwissArmyDruid posted:a combination oculink and thunderbolt egpu dock hit the market, and after benchmarks, it's not even funny how much bandwidth thunderbolt overhead consumes, which makes one wonder, why consumers even let it get to market dominance in the first place. it was faster than usb3 and could run pcie devices nobody else was selling a pcie breakout solution less janky than running a loose flex out the side of your pc as far as anyone knew
|
# ? Jul 21, 2023 09:20 |
|
|
# ? Jun 9, 2024 06:31 |
|
https://tenstorrent.com/research/the-ojo-yoshi-report-jim-kellers-journey-from-cpus-to-ceo/quote:Keller’s next stop was Intel, where he had 10,000 underlings — another huge change.
|
# ? Jul 21, 2023 18:10 |