|
The day this starts supporting GPGPU is the day I'll blow my load. Right now, I'm wanting to re-jig some computational fluid dynamics simulations that I have running on CPU to run on a GPU. It would be so stupidly faster it's unbelievable, but I have the major trouble of bandwidth to/from storage, because GPUs lack the RAM I need. If this supports GPGPU and I can use a fast bus to a 12-24GB RAM box, then my problems vanish to a large extent. This will be a wonderful day indeed.
|
#
¿
Sep 19, 2010 14:55
|
|
|
|
| # ¿ Apr 28, 2024 01:21 |
|
|
Rime posted:Unless you bought an i7 920.  Mine's still going strong. It and a 560Ti seem to do just fine for everything I do.
|
|
#
¿
Feb 17, 2015 15:25
|
|
|
Twerk from Home posted:You know what would be way better than us getting better at using multiple cores? Really fast single cores. I would love to have a single core 17.6GHz Haswell instead of a quad core 4.4 GHz one. The comments on that article are amazing: Allen posted:If 10 GHz is the best that Intel can do by 2011, AMD or somebody else is going to eat their lunch. Intel better pick up the pace if they want to remain dominant. Besides, I want it NOW. What will I do with it. Well, I also want the applications now. I guess I've been spoiled by the industry and expect incredible improvements every year. StickWithApple posted:doesn't matter the speed of intel's chip in 2011 because motorola and ibm will already have chips out by then that are more effective, use energy better, and run applications ready for the new iMac's that are introduced at the 2011 MacWorld. They'll run at somewhere near 7GHz and still be faster than an 11GHz or even 128GHz sh*t that intel puts out.. Sinful posted:Back in grad school I worked on computing with light and transistors that had ten states (0-9 or base 10) rather than 2 (0 and 1 or binary). Anybody who doesn't think that these types of technology won't be commercially available by 2011 is kidding themselves.
|
|
#
¿
Jun 3, 2015 09:56
|
|
|
So I'm still on an i7 920. It's beginning to get a little bit long in the tooth, so I'm wondering what to do. Is it worth actually looking towards Skylake and DDR4 or is it just better to go with a 2nd-hand Haswell era chip and be done with it? I have 18GB total of triple channel RAM, half 10666 and half 12800 so I don't think I can really re-use that as it seems everyone went back to dual channels? I was sort of waiting on Skylake to drop and see how it's panning out now, but by the looks of it, it may just be easier to cycle a 2nd-hand Haswell than actually pay the latest-gen premium. I don't really see a benefit of early-adopting DDR4 because the performance seems to be a near-run and the LGA is just going to change so I can't recycle the mobo in a few years anyway.
|
|
#
¿
Aug 10, 2015 14:34
|
|
|
Twerk from Home posted:What's driving your upgrade, do you need more CPU power and have a highly threaded workload? If so, you want a 5820K and X99. Do you want a more modern chipset, onboard stuff, and peripherals? Either H97 or Skylake should meet your needs unless you are imminently expecting to buy a PCI-E SSD, in which case you want Skylake. It's part gaming (dota and CS are fine on my current setup, but newer games are a bit of a push), part newer platform (SATA3 and USB3 would be lovely). I can wait a bit, but it just means I'll hold off on upgrading to Windows 10 for a few motnhs until Skylake settles down and (hopefully) drives the 2nd hand Haswell chips down.
|
|
#
¿
Aug 11, 2015 09:47
|
|
|
Malloc Voidstar posted:You can't play Fallout 4 at 144FPS anyway, physics and FPS are linked.  How does that decision get made and be considered a remotely acceptable solution?
|
|
#
¿
Feb 1, 2016 08:20
|
|
|
dud root posted:My Asrock Z170 Extreme 4 refused to clock my Gskill 3200 RAM anything above about 2400, until the recent V3.20 BIOS. Now it does 3200 without issue. At a guess the compatibility has been improved a bunch across other boards with bleeding edge BIOS (It should've worked in the first place 6 months ago though) I literally had the opposite with my gaming 4. 1.90 accepts my G.skill 3200 XMP fine but upgrading the bios means it won't even post if it's not run at 2133.
|
|
#
¿
Jun 22, 2016 04:43
|
|
|
canyoneer posted:What a world where someone can use a 10 year old PC and not hate every moment of it. I only upgraded from an i7-920 to a 6700K 2 years ago. I game, too, though it's largely CSGO and Dota. Honestly the thing that taxed it most was Anno 1404, it sent the CPU fan into high gear. Your comparison between a P3 450 to an '09 PC is hardly apt. Usage requirements really plateaued around 2010-2015 depending where on the curve you fell. In particular, the move from single to dual core was the big step, since one thread didn't suddenly tie up everything else. Add in hyperthreading on dual cores and you rapidly lose any desktop advantage after that point. For example, web browsing is no faster on a 6700K vs a 920, really. And more people are playing non-taxing games and doing just fine with them. My dad now uses the 920, all he does is browse the net, watch TV on his digital tuner and gently caress around with linux. None of that requires anything more than a 920.
|
|
#
¿
Jul 11, 2018 09:07
|
|
|
Not gonna lie, watching Intel squirm like this is kinda amusing, even after all this time. Can’t wait for the I7-9999XE where it crashes under any benchmark level loading but is totally stable and fine for daily desktop use so long as that only involves email and word processing and Intel only sells it under NDA to the admin secretary market.
|
|
#
¿
Mar 4, 2019 00:43
|
|
|
TheFluff posted:Steve isn't pulling any punches: This is a murder. I'm watching a murder on YouTube.
|
|
#
¿
Mar 31, 2021 03:15
|
|
|
|
| # ¿ Apr 28, 2024 01:21 |
|
|
Got another 16GB (2x8GB DIMMS) for my 6700k / z170 system after hitting some limits on my existing 16GB. Throw it in, system boots, but firefox crashes a couple of times and hten it freezes on a memory error. gently caress, faulty DIMMs I guess. Time to do some memtest and juggling DIMMs and slots to work out the bad combo. So I run memtest on the whole lot and get a bunch of errors as expected. I shut down, remove the old set and start testing the new set in each of the 2 sets of slots. It boots, so I memtest and it passes fine. I swap the DIMMs to the other set of slots, boots again fine and passes fine. What the gently caress, the new set is fine in both sets of slots, and the old set never gave me issues, what's going on? I throw all 4 DIMMs back in and try to boot. It now tries multiple times before showing a message that it failed to boot, and to hit DEL to go into my BIOS to change settings but otherwise it'll boot. So it boots, and looks all good. I run a memtest and it passes fine. At this stage I'm confused as gently caress. I reboot, and it does the exact same pattern - fans spin up and down a few times, then the BIOS complains it couldn't boot and to go into it to change settings but continues to boot afterwards. Then the penny drops - when it's showing that message, it's telling me that it's changed some settings in order to boot. Lo an behold, it was undong my XMP 3200 profile back to the stock 2133 and booting fine under that. However, it wouldn't permanently disable XMP, so the next reset it'd go through the process of failing several times before showing a relatively non-descriptive message and proceeding to boot fine. Suddenly it all makes sense. Skylake only supports 2133, but can generally do 3200 fine. However, it appears my combo of CPU + mobo can do 2 DIMMs at 3200, but not 4 DIMMs, under the XMP profile that lifts the RAM voltage to 1.35V. That's ok, easy fix is to manually give it some more juice. So I set it to 1.36V and it boots fine. Hurrah! Time to find the balance point. Set it back to 1.355V, boots fine. Cool, let's go to 1.35V and work up from there. Set it manually to 1.35V, boots fine. What the gently caress. Set it to auto, so it uses the XMP profile. Boots fine. What. The gently caress. It's 1:30am and I had an 8:30 meeting so at that stage I went to bed, figuring it'd poo poo the bed in the morning. But nope, it's now booting with absolutely auto XMP profiles. I have no idea what has changed to make it work. All I can put it down to is little nano-beavers built up a dam across a gate somewhere in the RAM and upping the voltage washed away whatever blockage they'd built and now at 1.35V the electrons can flow freely again. Seriously though, clearly it's right on the edge of what's stable such that something changed to make it pass but it will probably start playing up again. Those sort of intermittant issues are a bitch to try to diagnose. I'll run memtest on it overnight tonight to see whether it's stable at 1.35V. If so, I'm going to write this off as a bizarre mystery. Otherwise, at least I have a good way to start resolving the issue.
|
|
#
¿
May 7, 2021 02:38
|
|