|
Alereon posted:A New Kind of Overclocking The change to overclocking is interesting. I assume that most retail processors will be sold as "K" models since 10% is a relatively small price to pay for the extra performance a knowledgable person can get out of it. When I first read this paragraph I thought Intel would limit unlocked multipliers and overclocking in general to the completely unreasonably priced Extreme Editions. Phew.
|
# ¿ Sep 14, 2010 14:32 |
|
|
# ¿ Apr 25, 2024 22:08 |
|
Somebody please open a OC nostalgia thread. That picture of the Slot A Athlon with the little debug module (gold finger something?) can only be topped by a Celeron 300A with peltier/air cooler. Content: I think Im going to wait for the Sandy Bridge update of Apples MBP line until I replace this 17" C2D. Should arrive roughly one year from now, no?
|
# ¿ Sep 23, 2010 16:44 |
|
Haswell is shaping up to be very interesting, but Im really curious what Broadwell will bring. One CPU architecture spanning from 4W phones to 85W+ desktops sounds crazy, but at this rate anything is possible. And I really dont find Intels 17W IVB SKUs slow at all. Quite the opposite, a 2012 Macbook Air has more processing power than 95% of all non-gaming consumers need. eames fucked around with this message at 21:19 on Jan 4, 2013 |
# ¿ Jan 4, 2013 21:14 |
|
canyoneer posted:http://techreport.com/news/24162/intel-reveals-7w-ivy-bridge-cpus-convertible-haswell-ultrabook Note that this is 7W SDP (Scenario Design Power). The TDP of these CPUs is 13W, the rest is just marketing fluff. Heres an article on this (which I admittedly didnt read yet): http://hothardware.com/News/Intel-Confirms-New-7W-Ivy-Bridge-Chips-Haswell-Parts-To-Follow/
|
# ¿ Jan 9, 2013 11:33 |
|
incoherent posted:I really, really want that 5200 in a macbook air. So do I, but in reality Apple will probably drop the dGPU in the 15" rMBP because the HD5000 is now fast enough for smooth facebook scrolling.
|
# ¿ Mar 28, 2013 10:18 |
|
Factory Factory posted:Link also has a picture of a delidded GT3e SoC part, with one multi-chip package with the CPU and eDRAM and one PCH package, so click through for that. Interesting. I read that the integrated VRMs on Haswells are in a separate die on the chip, so I expected GT3e to have three dies.
|
# ¿ Apr 13, 2013 08:16 |
|
Install Windows posted:An ARM version of a Macbook Air would be a lovely laptop that performs poorly and doesn't have the benefits of either the iOS or the OS X ecosystems, and would only have the dubious benefit of marginally better battery life. Warning, wild offtopic speculation ahead I think the more likely route is a Macbook Air that can be separated at the hinge, with a fully functional but even lighter iPad Air as the display and a powerful CPU/Battery/IO/Keyboard for heavy work as the (optional) "base" part. Itll take a few years until we get there but that seems to be the obvious solution to me, although I have no idea how it would work on the software/OS side of things. I find it unlikely that they would mix two different architectures in one device, though.
|
# ¿ Nov 1, 2013 21:16 |
|
Intel open sourced their Broadwell GPU drivers. I expect they will be able to keep their +50% GPU performance per generation pace up.quote:The changes are massive and it's looking like the Broadwell graphics improvements will be astonishing and provide significant improvements over Haswell and earlier generations of Intel graphics. Cant wait to see some benchmarks of this chip compared to the respective Maxwell parts. http://www.phoronix.com/scan.php?page=article&item=intel_broadwell_linux&num=2
|
# ¿ Nov 4, 2013 07:16 |
|
PC Watch posted this: Looks like high end iGPUs (GT4e) died a fiery death. That would also explain the massive delays across Apple's mac product line. Imagine Mr. Ive sitting on his finalized design for a new Retina Macbook Pro thin enough to slice cheese and then learning that he'll have to make room for another 45W chip (dGPU) because that planned GT4e CPU is not going to happen. Coffee Lake looks to be another 14nm design (tock #4?) eames fucked around with this message at 22:08 on Jul 21, 2016 |
# ¿ Jul 21, 2016 22:02 |
|
Looks like somebody in Hong Kong managed to get his hands on a retail version of Kaby Lake (i5-7600K) for the Desktop. http://translate.google.com/translate?u=http%3A//www.facebookhk.com/&hl=en&langpair=auto%7Cen&tbb=1&ie=UTF-8 The chip seems to be stable at 5.1 Ghz with 1.55V with air cooling but I have no idea how safe that voltage is/will be. Will these be widely available before christmas?
|
# ¿ Oct 17, 2016 10:50 |
|
GRINDCORE MEGGIDO posted:Any posters in here delidded their Skylake chips? No but a friend of mine has. He asked me to 3D print the tool for him. Looks very but he said it worked well. https://www.youtube.com/watch?v=4o-rDGHZH4s I can ask him if you have any specific questions.
|
# ¿ Dec 20, 2016 19:55 |
|
PerrineClostermann posted:I was looking at them for my home storage solution Select Haswell home server models are fairly cheap now as they approach EOL. I bought a few new Dell Poweredge T20s with Xeon e3-1225v3 (quad 3.2 Ghz with ECC) for 200 each. They idle at at 12-15W and have no problem transcoding 4k content. Lenovo TS140 are similar. eames fucked around with this message at 21:11 on Jan 10, 2017 |
# ¿ Jan 10, 2017 21:09 |
|
re AVR54, apparently the fix is "a minor silicone tweak". I set up a DS1515+ connected to a pfsense firewall and VPN into that with another pfsense box last December. As far as I can tell all use the same C-2000 series CPUs so the race is on. The 1515+ seems to have a 3 year warranty but the Netgate pfsense boxes only come with 1 year (which already expired). There is probably going to be a lot of bad blood if they just leave their customers SOL. On the other hand every single device they make is affected and they aren't making billions from other products like Cisco, so what else can they do. Wouldn't want to be in their shoes!
|
# ¿ Feb 6, 2017 21:29 |
|
Vendors are now reacting to the 18-month timebomb errata. None of them are allowed to mention the component or company, Synology even had to pull a statement because they mentioned Intel. Pfsense/Netgate vowed to replace all affected units within 3 years of purchase which seems fair. https://blog.pfsense.org/?p=2297 Still, having a ticking timebomb as a firewall which is often a single point of failure feels bad. edit: better link: https://www.servethehome.com/intel-atom-c2000-series-bug-quiet/ servethehome posted:Our educated guess is that Intel may have tied access to those reserve funds to signing an NDA for not discussing the issue. eames fucked around with this message at 00:03 on Feb 8, 2017 |
# ¿ Feb 7, 2017 21:08 |
|
Most companies are rolling out board level fixes/workarounds which allow the buggy processors to be used without the issue occurring. I'm not even sure if there's a newer stepping available at the moment?pfsense posted:A board level workaround has been identified for the existing production stepping of the component which resolves the issue. This workaround is being cut into production as soon as possible after Chinese New Year. Additionally, some of our products are able to be reworked post-production to resolve the issue. C2000 series based products should be safe to buy in a few weeks, assuming you don't end up with old stock.
|
# ¿ Feb 8, 2017 15:54 |
|
If those supposedly leaked Ryzen benchmarks are correct then Intel will need a lot more than "high performance TIM", heh.
|
# ¿ Feb 11, 2017 23:28 |
|
Ihmemies posted:From the amd thread: Their top stock cooler is also rated for 140W TDP and insiders have dropped info that the 1800X hits over 90°C around 4.0 Ghz on air so I'm expecting launch models >200W overclocked. Chances are that this will improve as GloFo becomes more experienced with the chips/process, just like the RX480 went from >150W stock to <100W factory OC over the course of a few months.
|
# ¿ Feb 22, 2017 20:54 |
|
So that just happened... article posted:Intel Core i7-6950X ($1599 US) $300 Price Cut
|
# ¿ Feb 25, 2017 18:27 |
|
the comments on the site say that these are just their normal retail prices but the reductions look substantial enough to make me think that this is the Ryzen response... we'll find it soon enough.microcenter posted:Several CPUs also come with a further $30 savings promotion when bundled with a compatible motherboard. 6700K for $229.99 when bundled with a mainboard isn't too shabby.
|
# ¿ Feb 25, 2017 18:39 |
|
Cannonlake to be a stacked/3D modular architecture? On-die coprocessors for native ARM code? http://seekingalpha.com/article/4051541-suerte-de-capote-intels-cannonlake-leaked-patent Take this with a large amount of but if it is true then AMD is in for a big surprise. Interesting timing on the article of course.
|
# ¿ Mar 3, 2017 00:04 |
|
Netgate sent out an email regarding the Atom C2000 bug today. Looks like they were able to fix it with a BIOS update because their appliances don't use certain buses. The update can be done remotely via SSH but requires somebody to plug it out and back in at the end of the process. That sure beats having firewalls suddenly die after a few months. Synology on the other hand seems suspiciously quiet...
|
# ¿ Mar 20, 2017 17:57 |
|
Tokamak posted:
http://wccftech.com/intel-kaby-lake-g-hbm2-gpu-multi-die/
|
# ¿ Apr 4, 2017 19:52 |
|
Thanks Ryzen?article posted:Intel will unveil its Basin Falls platform, i.e. Skylake-X, Kaby Lake-X processors and X299 chipset, at Computex 2017 in Taipei during May 30-June 3 two months earlier than originally scheduled, and will bring forward the launch of Coffee Lake microarchitecture based on a 14nm process node from January 2018 originally to August 2017, to cope with increasing competition from AMD's Ryzen 7 and Ryzen 5 processors, according to Taiwan-based PC vendors. http://www.digitimes.com/news/a20170419PD207.html
|
# ¿ Apr 19, 2017 20:35 |
|
Paul MaudDib posted:I can actually see a lot of potential for Coffee Lake to flop. Kaby's lead mostly derives from its exceptional clocks. They seem fairly confident article posted:14nm++ is claimed to offer 25% greater performance at a given power level than the unoptimized 14nm process first used to produce Broadwell and Skylake chips, or as much as 52% less power consumption for the same level of performance. In fact, Intel's projections show that the transistor performance of 14nm++ will actually exceed that of its first generation of 10-nm products. Expect to see 14nm++ underpin Intel's rumored Coffee Lake CPUs later this year. I wonder if the Basin Falls line (Skylake-X/Kaby Lake-X) is going to use the same process. http://techreport.com/review/31660/intel-defends-its-process-technology-leadership-at-14nm-and-10nm
|
# ¿ Apr 21, 2017 07:32 |
|
Paul MaudDib posted:Pretty obviously a TIM problem. There's been a background murmur about this since launch but it's blown up today for some reason. In most cases it gets worse over time, presumably as the TIM migrates out of contact with the surfaces or whatever. Heat often makes various things less viscous so the heavy overclocking that many 7700Ks undergo might be a factor as well. First time I've heard of this problem but it should be easy to check by observing the behavior, delidding, applying new TIM and checking if it stops. My 2013 Crystalwell rMBP had significant thermal issues with massive throttling (50% of nominal clockspeed on all four cores under load). I replaced the TIM which was a was a dry and crumbly mess and clock frequencies went back to 130% of nominal clockspeed at full load on all four cores. As you mentioned the constant heat does deteriorate thermal interface materials over time and OEM pastes are often much worse than what you can get for aftermarket OC purposes. The case of my rMBP has a whiff off planned obsolescence about it but that's -material. Did I mention that the newest MBPs have their chips and heatsinks attached to the rear of the PCB and repasting requires the whole logic board to be removed?
|
# ¿ May 6, 2017 13:11 |
|
JnnyThndrs posted:I thought the TIM wasn't a factor, since people who delidded still had the temp-spike problem, according to the article? Huh yeah, I missed that part. Based on some user reports it spikes much faster than it should be possible given the specific heat capacity of a CPU cooler. If it isn't a TIM issue then perhaps a thermal sensor issue. (this is with watercooling. source) eames fucked around with this message at 14:03 on May 6, 2017 |
# ¿ May 6, 2017 14:00 |
|
Combat Pretzel posted:An 8-core Skylake-X at 600 bux or less would be nice, but I don't see it happening. with ECC support but... Combat Pretzel posted:Too much hubris over at Intel.
|
# ¿ May 7, 2017 22:58 |
|
Place your Core i9 7920X price bets now! I'd say $2299.
|
# ¿ May 13, 2017 11:19 |
|
http://cpu.userbenchmark.com/SpeedTest/278103/Genuine-IntelR-CPU-0000-- http://cpu.userbenchmark.com/SpeedTest/233971/Genuine-IntelR-CPU-0000-- http://www.tweaktown.com/news/57568/intel-core-i9-benchmarks-i7-7920x-12c-24t/index.html eames fucked around with this message at 09:04 on May 16, 2017 |
# ¿ May 16, 2017 09:02 |
|
Yeah, looking at the benchmarks the 7920X ran on a single 8GB stick of DDR4-2113 (9GB/s) and the 7900X benchmark is 3 months old.
|
# ¿ May 16, 2017 10:34 |
|
It's a slow day at work so I overlaid the two cache latency graphs. This is pretty deep into territory but whatever, maybe one of you smart guys can explain what we are looking at. orange = 10C 7900X blue = 12C 7920X
|
# ¿ May 16, 2017 13:37 |
|
I want this but I'm pretty sure it'll cost a fortune.
|
# ¿ May 22, 2017 00:26 |
|
New leaked Skylake-X benchmarks https://d1rktuf34l9h2g.cloudfront.net/2/25/2513876a_7900X.jpeg https://d1rktuf34l9h2g.cloudfront.net/c/cb/cb0313a2_6950X.jpeg copypaste: i7-7900X 10C (@4.0 Ghz)= 1386.94Mpix/s i7-6950X 10C (@4.50 Ghz)= 897.28Mpix/s i7-6950X 10C (@3.50 Ghz)= 746.64Mpix/s https://www.techpowerup.com/233667/new-details-on-intels-upcoming-10-core-skylake-x-i9-7900x-surface eames fucked around with this message at 23:19 on May 25, 2017 |
# ¿ May 25, 2017 22:57 |
|
I'm kind of losing track of all the architectures and codenames. How will the entry level "HEDT" Kaby Lake X CPUs compare to "Desktop" Coffee Lake? Aren't those supposed to launch at roughly the core count (4-6C) at the same time?
|
# ¿ May 26, 2017 15:51 |
|
Paul MaudDib posted:Coffee Lake will supposedly introduce 6C to the consumer lineup, but it's limited to the standard consumer featureset (16 lanes, dual-channel RAM, etc). Also, 6-core non-hyperthreaded for some reason. Ok thanks, that clears things up. KBL singlecore performance paired with 6C/12T and a halfway decent upgrade path to 10 cores/quadchannel sounds nice. I hope that X299 idle power consumption won't be off the charts and perhaps for some comparable, unlocked and only slightly more expensive Skylake-W CPUs for ECC but who am I kidding...
|
# ¿ May 26, 2017 16:51 |
|
|
# ¿ May 27, 2017 14:26 |
|
It gets worse
|
# ¿ May 27, 2017 14:48 |
|
WAR DOGS OF SOCHI posted:It's hard for me to believe that windowed cases are still a thing, but it looks like they're bigger than ever? It is surprisingly hard to find a good case that offers a no-window option. I personally would like to buy a Phantex Enthoo Evolv ATX without a window but they don't even offer a solid side panel as a replacement part.
|
# ¿ May 27, 2017 17:15 |
|
DrDork posted:I wonder if the bling-fest is more accepted/popular in other countries, like Korea? Might explain some of it, especially from manufacturers with strong Asian or Euro presences. Yeah that's the only explanation. Same deal with monitors, keyboards, mice... any gaming related computer component really. I simply refuse to buy this blingy stuff figuring that it'll eventually stop because there can't be that many teenagers with the disposable income for these higher end parts. It'd make sense to me if the low end budget parts had (dirt cheap) colorful LEDs and the higher end stuff looked more subdued, like Asus' old workstation mainboards, but it's the other way around.
|
# ¿ May 27, 2017 22:36 |
|
|
# ¿ Apr 25, 2024 22:08 |
|
Intel is preparing more Skylake-X CPUs to counter Threadripper, now with a rumored Core i9-7980XE with 18C/36T. PCGH claims to know from a reliable source that the heatspreader on Skylake-X is not soldered, so the countdown to the first $2500+ Extreme Edition i9 killed during delidding starts now. http://translate.google.com/transla...n&langpair=auto At the risk of sounding like a broken record, just how big is the market for desktop users requiring 18 Cores at ~3 Ghz and without ECC RAM? At this point it looks like AMD and Intel are just rebadging CPUs that they had to make for the server market anyway. I'd be much more interested in a highly binned 6-8C SKU at 5 Ghz (not that I'd buy one).
|
# ¿ May 29, 2017 12:27 |