|
Seamonster posted:When 7nm comes around for sure. If nothing else than to put an absolute spike bat to Intel's I'd be more than happy if they could deliver 8c/16t at 5GHz, honestly.
|
#
¿
Jan 26, 2018 06:38
|
|
|
|
| # ¿ Apr 25, 2024 12:37 |
|
|
NIGARS posted:"PC Chips it is." Ah, the classiest of them all, with their fake cache chips 'n' all. But didn't they get bought out by ECS?
|
|
#
¿
Feb 24, 2018 09:12
|
|
|
Fractal Design is the answer to most case-related questions, honestly.
|
|
#
¿
Feb 24, 2018 22:22
|
|
|
gourdcaptain posted:I'd agree if mine didn't have a useless door on the front that I can't shut without blocking all the fans I don't know which case you have, but it most likely has intakes on the sides of the front bezel where air is drawn in. The door is to remove a direct path for the sound, the foam to dampen higher frequencies HalloKitty fucked around with this message at 22:42 on Feb 24, 2018 |
|
#
¿
Feb 24, 2018 22:38
|
|
|
crazypenguin posted:Do these new 10GbE devices also support that new 2.5/5 Gbps standard? That really is a bigger selling point to me. 5Gbit really does work over Cat5e up to reasonable lengths.
|
|
#
¿
Mar 3, 2018 17:40
|
|
|
Jago posted:I think it's old news that AMD is seriously binning the 8C/16T stuff... still weird to me that the 8/16 is at a higher clock than the 6/12. I guess Intel is doing the same thing right now. I feel like historically there would be a lower core count, higher clock-speed sort of trade-off. Maybe I'm imagining it. AMD did it with Threadripper too - higher core count, higher clocks, the best dies...
|
|
#
¿
Mar 6, 2018 09:03
|
|
|
repiv posted:CTS Labs, a security consultancy, don't even have HTTPS set up on their main site Yeah, there are no contact details, and the whois page is also fishy. I suspect that this "consultancy" could be some unscrupulous firm thrown a few dollars by Intel. The page only just appeared, too. There was no record of it before January on the wayback machine in its current state. Before that it was just a landing page. vv Yeah, that's just frankly comical. I'm all up for hearing about new exploits, but this is just silly HalloKitty fucked around with this message at 18:22 on Mar 13, 2018 |
|
#
¿
Mar 13, 2018 18:17
|
|
|
KKKLIP ART posted:Other than maybe wanting a higher resolution screen I was expecting 1366×768 after reading this. 1920×1080 is plenty. Sure, not the best aspect ratio, but the resolution is fine.
|
|
#
¿
Apr 11, 2018 16:29
|
|
|
Lungboy posted:Did i miss them or have no benchmarks come out for the new Ryzen chips yet? With the release on thursday, when should we expect them? Thursday? Thursday, yeah, it's when the NDA lifts.
|
|
#
¿
Apr 16, 2018 11:54
|
|
|
A threadripper part at 5GHz is basically the dream, so if they can deliver that....
|
|
#
¿
May 17, 2018 17:26
|
|
|
eames posted:Sony's devs are working on LLVM compiler improvements for the current Zen architecture. This should be good news down the road for all gamers as it seems exceedingly likely that the PS5 will launch with some sort of Ryzen CPU/APU. Piledriver? Nope, Jaguar. A chip designed for the fashionable netbook or tablet of the day.
|
|
#
¿
May 21, 2018 07:18
|
|
|
Arzachel posted:Hahaha, you'd think so but clock for clock A4-5000 is less than 5% below a FX-4300 in single thread Cinebench 11.5. It really only goes to show how much of a dog that whole Bulldozer architecture really was..
|
|
#
¿
May 21, 2018 09:45
|
|
|
wargames posted:is that one of those delete system32 jokes? I doubt it - anandtech found serious performance issues with hpet forced on on intel platforms.
|
|
#
¿
May 29, 2018 06:32
|
|
|
Hmm, I'm not sure I have a reasonable use for 32 cores. I'd rather have 16 at a higher clock speed.
|
|
#
¿
Jun 6, 2018 14:15
|
|
|
NewFatMike posted:I didn't know we were up to USB Type D We already have A and C to insert our plugs into...
|
|
#
¿
Jun 22, 2018 15:36
|
|
|
AMD's fuckin' going for it, no shame. Why the hell not?
|
|
#
¿
Jun 26, 2018 19:04
|
|
|
AnandTech interviews Jim Keller, thought it might be of interest to some here vv D'oh HalloKitty fucked around with this message at 18:42 on Jul 16, 2018 |
|
#
¿
Jul 16, 2018 17:56
|
|
|
It gets people talking. Look, it's working right now
|
|
#
¿
Aug 2, 2018 19:42
|
|
|
Cygni posted:The full 32C 4-die 2990WX has a 3ghz base and is $1800 (ouch) Ouch? $1800 for a 32 core, 64 thread CPU with high clocks? That's extremely cheap if you compare it to Intel's offerings, which only seems fair. To even get up to the largest Intel sells (28 core, 56 thread) you're talking around $10,000, and the clock speeds don't even compare to AMD's. Even if you compare it to Epyc, it offers killer value. You've basically halved the price, and bumped the clocks significantly. I'm liking the look of the 2950X personally, keeping the heat and power in check, whilst offering the highest clock speeds; but for someone who can make use of all 32 cores, the 2990WX is a bargain. HalloKitty fucked around with this message at 09:51 on Aug 4, 2018 |
|
#
¿
Aug 4, 2018 09:37
|
|
|
Truga posted:I hope I can talk my boss into buying a 2990 for the renderbox we really need at work, that's absolutely hilarious value. Exactly.
|
|
#
¿
Aug 4, 2018 12:34
|
|
|
Cygni posted:Also lol that cooler: Noctua NH-D15 does this as well, the NH-D15S was designed to avoid the problem. Not to defend AMD or Cooler Master as such, but in any case, enormous air coolers blocking the first slot is nothing new.
|
|
#
¿
Aug 6, 2018 18:07
|
|
|
Seamonster posted:I want to see that that 4.4 turbo on the 2950x as a preview for the minimum acceptable clocks on 7nm. Is that even XFR turbo? It could be that XFR turbo is even higher.
|
|
#
¿
Aug 6, 2018 20:27
|
|
|
Harik posted:How the hell did they go from a dominant market position to such a tire fire? Not any chance, honestly; Intel has SO much more capital
|
|
#
¿
Sep 8, 2018 20:18
|
|
|
Mr Shiny Pants posted:What do you mean? I never installed any Asrock stuff, machine runs fine it seems. I'm assume he's alluding to the lack of hysteresis in the fan settings, but I have no experience of that board or software
|
|
#
¿
Sep 15, 2018 11:02
|
|
|
An incredibly important development for Threadrippers larger than 16 cores: Dynamic Local Mode. Advanced Micro Devices posted:Dynamic Local Mode is implemented as a Windows® 10 background service that measures how much CPU time each thread on the system is consuming. These threads are then ranked from most to least demanding, and the top threads are automatically pushed to the CPU cores that contain direct memory access. Once these cores are consumed by work, additional threads are scheduled and executed on the next available CPU core. This process is continuous while the service is running, ensuring the most demanding threads always get preferential time on cores with local memory. (As a corollary, insignificant threads are pushed to other dies.) Essentially, demanding threads automatically get scheduled onto the CCXs with direct memory access. It should help significantly with mitigating any performance regression going from the 2950X to the 2990WX. When is this goodness available? Advanced Micro Devices posted:Dynamic Local Mode available starting October 29th HalloKitty fucked around with this message at 13:45 on Oct 7, 2018 |
|
#
¿
Oct 7, 2018 13:36
|
|
|
Mr Shiny Pants posted:You would think that this would be a scheduler thing, not a thing that would run in a service. I'm guessing they'd prefer Microsoft would roll it into Windows, just like HyperThreading awareness is a part of Windows, but maybe it just takes time.
|
|
#
¿
Oct 7, 2018 14:46
|
|
|
EmpyreanFlux posted:Rome wasn't built on one die. Has to be the new thread subtitle HalloKitty fucked around with this message at 20:29 on Nov 7, 2018 |
|
#
¿
Nov 7, 2018 17:59
|
|
|
The benefit of pcie4 isn't any one specific thing, like storage or gpu, but rather that they can use fewer lanes to achieve the same amount of bandwidth. That means more devices, better links to pcie switches, chipsets and so on.
|
|
#
¿
Nov 14, 2018 09:33
|
|
|
Cygni posted:Microcenter couldn't less this stand. The 1700X is now $129 in store only, with a $30 motherboard credit. Lordy. Jesus H Christ that's cheap for that much power. The absolute madmen.
|
|
#
¿
Nov 20, 2018 21:08
|
|
|
Combat Pretzel posted:I just hope the Zen 2 based Threadrippers do double the core count for a similar price range of the predecessor models, i.e. my future 3950X having 32 of them for a grand. I'm guessing they probably will. Personally I'd prefer a 16 core with a really high clock speed, but I guess we'll see how things work out. If the 32-core has barely any clock deficit from the 16.. well... 
|
|
#
¿
Nov 29, 2018 17:13
|
|
|
Malcolm XML posted:They use lesser dies right? The io die looks to be what would be crippled or harvested. TR has always sort of been a yield improvement program Nah, the highest-end Threadripper SKUs used the absolute best dies. They've always correspondingly clocked higher out of the box than the consumer Ryzen CPUs. 1950X XFR 4.2, 1800X XFR 4 2950X XFR 4.4, 2700X XFR 4.3
|
|
#
¿
Dec 2, 2018 10:43
|
|
|
NewFatMike posted:If this is real This is wank material for everyone other than Intel. But it can only be taken with an ocean of salt. it's just TOO good.
|
|
#
¿
Dec 3, 2018 11:38
|
|
|
To be fair, it looked and stank of bullshit from day dot
|
|
#
¿
Dec 13, 2018 12:20
|
|
|
LRADIKAL posted:Do you get a measurable improvement in performance compared to just letting the CPU do it's own thing? I wasn't able to. I recently built a 2700X build for my younger brother, and other than updating the bios and installing the chipset drivers, I didn't mess with cpu settings, the attraction of the 2700X being it's pretty much max'd out out-of-the-box. (of course I corrected the ram speed) vv Hm, I'm pretty sure I was seeing 4050MHz all-core with the newest BIOS updates and the "AMD Balanced" or whatever it's called power profile from the newest chipset drivers, but I don't have the machine in front of me to confirm HalloKitty fucked around with this message at 19:02 on Dec 27, 2018 |
|
#
¿
Dec 27, 2018 10:22
|
|
|
Stanley Pain posted:Don't use that power profile anymore. The Microsoft balanced one is the one to use. I just wonder why it installs by default with the chipset drivers then. It holds much higher idle speeds than the built in balanced profile, and I'm going to guess AMD probably has left it like that for a reason. 
|
|
#
¿
Dec 28, 2018 01:49
|
|
|
Sorry, I don't have anything to back this up (although I've been playing around, trying to find evidence - any cached pages (although the cpu list queries a database), links to that f7d bios, etc...), but I thought everyone would want to see this image anyway: I can't really figure out where it originates from, it could be that someone has simply manipulated the image, of course, although it's more likely they'd have just edited the code to give it an authentic look. It's hard to know whether we're seeing more real info, or just an echo of all the other poo poo rattling around. Obviously the current page doesn't look like that. Edit: hmm, is this the first possible leak of the L3 cache size? Double edit: this type of leak is also verifiable after the launch, because it contains very specific information - the stepping and the BIOS version that introduces support. I know a "leak" simply being known information after launch isn't very interesting, but I think that it'd be pretty neat to come back and see if it was valid. Triple edit: for what little it's worth, I think the 3700X is the one to get. 12 cores (more than most people know what to do with... today), out of the box boost to 5GHz, all in the same power envelope as the old 2700X? If it's real, Intel should be very worried. HalloKitty fucked around with this message at 16:19 on Jan 3, 2019 |
|
#
¿
Jan 3, 2019 15:49
|
|
|
That's interesting, H2 stepping matches up exactly with the list I posted on the previous page, making that data more and more likely to be real. My god, AMD might have actually done it.
|
|
#
¿
Jan 4, 2019 17:13
|
|
|
Palladium posted:Well I was recommending 4790K with a H/B mobo over OCing a Haswell i5 with a Z-board back in 2014, there were already a few games where the stock factory overclocked 4.2/4.4GHz 4790K beat OCed 4690K and yet people thought I was committing CPU heresy because "but-but-but OC IS EASY MAN AND WORTH IT" 4790k without bothering to overclock (with h97, for example) became a staple of the build thread, hell, I even built one for a family member
|
|
#
¿
Jan 5, 2019 11:33
|
|
|
fishmech posted:What because you're salty your statements are nonsensical? I can assure you that coding systems well for the triple core symmetric PPC core in the 360 was quite applicable to getting good results in PC multicore programming too. It even had SMT..
|
|
#
¿
Jan 6, 2019 12:30
|
|
|
|
| # ¿ Apr 25, 2024 12:37 |
|
|
BangersInMyKnickers posted:please don't do that Also.. slow fast l4 cache?
|
|
#
¿
Jan 12, 2019 20:28
|
|