|
Should be golden then. I was in the 50's at 1.2 volts
|
#
?
Jul 14, 2014 22:33
|
|
|
|
| # ? Apr 27, 2024 14:29 |
|
|
So when is the 14nm stuff coming out?
|
|
#
?
Jul 15, 2014 00:50
|
|
|
wargames posted:So when is the 14nm stuff coming out? It may never, to be perfectly honest. 10nm is certainly very much in question right now. Rime fucked around with this message at 00:54 on Jul 15, 2014 |
|
#
?
Jul 15, 2014 00:52
|
|
|
Picked up the CM Glacer 240L, fiddled with it a bit to reverse the direction of the fans to allow me to maintain positive air pressure. Played around with the OC, and this is what I settled on: Cranking it up to 4.8 required a whole lot more voltage, which was something I wasn't willing to accept for daily use (especially since SpeedStep just does not work on this board apparently, even after I updated the BIOS and completely reset it to default settings). Even at this, under full load it gets a bit louder than I'd like (the fan control options on this board leaves something to be desired as well, and PWM control on pump speed does nothing and it stays at the max 3500RPM all the time). Whatever, it's good enough to tide me over to Haswell-E at the very least.
|
|
#
?
Jul 15, 2014 07:39
|
|
|
Rime posted:It may never, to be perfectly honest. 10nm is certainly very much in question right now.
|
|
#
?
Jul 15, 2014 11:02
|
|
|
Since I live like 10,000 miles from a Microcenter I don't really get the big hoo-ha about the Pentium AE since even when paired a cheapest Z87/97 board, the B85 + i5 4590 that costs only a mere 20% more crushes it relevant real world benchmarks. Far less priceworthy than the $180 E6300 Conroes IMO.
|
|
#
?
Jul 15, 2014 12:27
|
|
|
Palladium posted:Since I live like 10,000 miles from a Microcenter I don't really get the big hoo-ha about the Pentium AE since even when paired a cheapest Z87/97 board, the B85 + i5 4590 that costs only a mere 20% more crushes it relevant real world benchmarks. Far less priceworthy than the $180 E6300 Conroes IMO. The Microcenter deal *is* why it's such a great deal - that i5-4590 is $199 by itself on Newegg, or double the cost of the whole bundle. Otherwise, yes, it's not a fantastic deal since it needs a Z87/97 board - just get an i3 instead.
|
|
#
?
Jul 15, 2014 14:41
|
|
|
cstine posted:
Gigabyte and asus h87/b85 boards can overclock now also.
|
|
#
?
Jul 15, 2014 14:47
|
|
|
For a few days there you could score the same deal for about $130 on Newegg too which honestly would be worth it if you had some use for it. I mean it is retarded fast for what you get, in its own wayGokieKS posted:Picked up the CM Glacer 240L, fiddled with it a bit to reverse the direction of the fans to allow me to maintain positive air pressure. Played around with the OC, and this is what I settled on: That is what I eventually settled on too. Did you notice how above 1.30 vcore the voltage would spike in both directions every few seconds by huge amounts? I think the mobo is "holding" it back but ... who cares really at the cost. I didn't have any fan or speed step issue though for sure though. Might be worth exchanging if you care about those 1gnoirents fucked around with this message at 15:20 on Jul 15, 2014 |
|
#
?
Jul 15, 2014 15:17
|
|
|
Welmu posted:Source? Intel has roadmapped both Broadwell and (some) Skylake for 2015 which are both 14nm parts. Derp, I was thinking 10nm & 5nm. 
|
|
#
?
Jul 15, 2014 15:38
|
|
|
Rime posted:Derp, I was thinking 10nm & 5nm. Maybe we'll have 450mm platters by then as well. along with rainbows and unicorns
|
|
#
?
Jul 16, 2014 10:59
|
|
|
Note that the nm of the process names aren't the actual nm measurement of the circuits. http://eandt.theiet.org/blog/blogpost.cfm?threadid=48709&catid=366 http://spectrum.ieee.org/semiconductors/devices/the-status-of-moores-law-its-complicated
|
|
#
?
Jul 16, 2014 13:02
|
|
|
Rime posted:Derp, I was thinking 10nm & 5nm. There is a big possibility Apple's upcoming A8 SoC will be fabbed at 20nm to leapfrog Intel's historical lead in process size, especially since the new Qualcomm modem is already 20nm and Apple's heavy investments into TSMC process.
|
|
#
?
Jul 16, 2014 13:26
|
|
|
Rastor posted:Note that the nm of the process names aren't the actual nm measurement of the circuits. So it's the bit wars for the modern age 
|
|
#
?
Jul 16, 2014 13:34
|
|
|
Palladium posted:There is a big possibility Apple's upcoming A8 SoC will be fabbed at 20nm to leapfrog Intel's historical lead in process size, especially since the new Qualcomm modem is already 20nm and Apple's heavy investments into TSMC process. Because 20nm < 14nm?
|
|
#
?
Jul 16, 2014 14:32
|
|
|
14nm parts aren't on the market yet, they're all 22nm.
|
|
#
?
Jul 16, 2014 14:42
|
|
|
Rastor posted:Note that the nm of the process names aren't the actual nm measurement of the circuits. That IEEE article is a great summary of how hosed things are getting in foundry land.
|
|
#
?
Jul 16, 2014 15:26
|
|
|
There's no doubt we're reaching the limits of silicon. The future may be the spiritual successor to the vacuum tube.
|
|
#
?
Jul 16, 2014 15:36
|
|
|
Palladium posted:There is a big possibility Apple's upcoming A8 SoC will be fabbed at 20nm to leapfrog Intel's historical lead in process size, especially since the new Qualcomm modem is already 20nm and Apple's heavy investments into TSMC process. Combat Pretzel posted:14nm parts aren't on the market yet, they're all 22nm.
|
|
#
?
Jul 16, 2014 15:59
|
|
|
Also, I have to imagine anyone who follows the GPU market has to be chuckling at the idea of TSMC of all people leapfrogging Intel's process size lead, even with Apple investment.
|
|
#
?
Jul 16, 2014 19:45
|
|
|
You know the thought really is kind of amusing, especially these days. But I'd really love if something came out of nowhere that kicked rear end.
|
|
#
?
Jul 16, 2014 19:48
|
|
|
Rime posted:That IEEE article is a great summary of how hosed things are getting in foundry land. With the important caveat at the end quote:Many people in the industry, who have watched showstopper after showstopper crop up only to be bypassed by a new development, are reluctant to put a hard date on Moore’s Law’s demise. “Every generation, there are people who will say we’re coming to the end of the shrink,” says ASML’s Arnold, and in “every generation various improvements do come about. I haven’t seen the end of the road map.” You can pull up articles from the 90's all the way through today that predict the demise of Moore's law at [current date] plus five years (tops!). And smartypants engineers figure it out every time. The end of process shrinking cadence of 18-24 months will probably come, but I don't think it will be soon. However, EUV patterning with any kind of good yield and volume (as far as I know) is still a HUGE question mark right now. At least it was a year ago when I did a lot of reading about it.
|
|
#
?
Jul 16, 2014 23:19
|
|
|
Can someone explain to me what Intel means by the Pentium G3258 only supporting 1333 mhz ram? Even if the motherboard sets and reports something faster, is it being limited in some other way?
|
|
#
?
Jul 17, 2014 02:02
|
|
|
1gnoirents posted:Can someone explain to me what Intel means by the Pentium G3258 only supporting 1333 mhz ram? Even if the motherboard sets and reports something faster, is it being limited in some other way?
|
|
#
?
Jul 17, 2014 02:11
|
|
|
Cool thanks, I was caught a little off guard by it. After I enabled XMP it immediately went to 2133 mhz so I guess it can handle at least that
|
|
#
?
Jul 17, 2014 02:12
|
|
|
Alereon posted:TSMC's new 20nm planar process is not as good as Intel's highly mature 22nm FinFET process, much less the upcoming Intel 14nm FinFET process. Keep in mind that even pessimistic projections have Intel's 14nm mobile processors in the hands of consumers by the end of the year, and TSMC won't be close to catching up until their 16nm FinFET process, so Intel is almost a full process cycle ahead. Sure, if the market success is entirely ALL about process nodes which clearly isn't. As Intel as much as they like to harp about their process advantages and CPU performance, there is no question that the general consumer chip market has been trending towards <$20 Apple in-house and Qualcomm SoCs, the former having absolute 100% control over chip design and the latter with the best in class modem designs as key advantages over Intel, while both also has more than the enough performance for their target market. Another overlooked about Intel's strategy of aggressive process size reduction and CPU improvements for the past few years has already backfired by having no choice but to provide the PC market with TOO much CPU power in order to sell chips. A $50 SB Pentium chip 3 years ago still is now vastly overkill to 90% of the PC population. If a 2500K gamer doesn't want a Haswell upgrade, why would the average consumer even care, just to save another 10W of power consumption? Palladium fucked around with this message at 11:32 on Jul 17, 2014 |
|
#
?
Jul 17, 2014 11:30
|
|
|
Palladium posted:If a 2500K gamer doesn't want a Haswell upgrade, why would the average consumer even care, just to save another 10W of power consumption?
|
|
#
?
Jul 17, 2014 14:11
|
|
|
At least the high end PC market is growing vs shrinking now though
|
|
#
?
Jul 17, 2014 14:57
|
|
|
Speaking of process difficulties: Intel 14nm desktop processors delayed again to Q3 2015.
|
|
#
?
Jul 17, 2014 16:34
|
|
|
Rastor posted:Speaking of process difficulties: Intel 14nm desktop processors delayed again to Q3 2015. Glad i went ahead and got a devils rear end in a top hat instead of waiting
|
|
#
?
Jul 17, 2014 16:40
|
|
|
Does Chip Loco have any reputation or is this all likely BS? E5 2600 and 1600 v3 lineups: http://www.chiploco.com/haswell-ep-e5-2600-v3-specs-35055/ http://www.chiploco.com/intel-haswell-ep-e5-1600-v3-35072/
|
|
#
?
Jul 17, 2014 19:15
|
|
|
Palladium posted:Sure, if the market success is entirely ALL about process nodes which clearly isn't. As Intel as much as they like to harp about their process advantages and CPU performance, there is no question that the general consumer chip market has been trending towards <$20 Apple in-house and Qualcomm SoCs, the former having absolute 100% control over chip design and the latter with the best in class modem designs as key advantages over Intel, while both also has more than the enough performance for their target market.
|
|
#
?
Jul 17, 2014 20:20
|
|
|
japtor posted:Does Chip Loco have any reputation or is this all likely BS? E5 2600 and 1600 v3 lineups: It certainly looks plausible; it follows the same basic Xeon pricing structure and matches core growth we saw SB->IB. The top-end 16/18 core parts probably look similar to the 2695/2697 v2s.
|
|
#
?
Jul 17, 2014 21:05
|
|
|
japtor posted:Does Chip Loco have any reputation or is this all likely BS? E5 2600 and 1600 v3 lineups: 18 cores! 
|
|
#
?
Jul 17, 2014 21:05
|
|
|
Palladium posted:Sure, if the market success is entirely ALL about process nodes which clearly isn't. As Intel as much as they like to harp about their process advantages and CPU performance, there is no question that the general consumer chip market has been trending towards <$20 Apple in-house and Qualcomm SoCs, the former having absolute 100% control over chip design and the latter with the best in class modem designs as key advantages over Intel, while both also has more than the enough performance for their target market. But us with i7 860s want to upgrade, and want quicksync for streaming.
|
|
#
?
Jul 18, 2014 00:19
|
|
|
Don Lapre posted:Glad i went ahead and got a devils rear end in a top hat instead of waiting Ok that's it, I'm picking up devil's oval office too whenever nvidia gets of their asses with the proper Maxwell chips now.
|
|
#
?
Jul 18, 2014 00:47
|
|
|
joke's on you guys, i'm getting haswell-e and totally screwing myself when intel suddenly decides to move broadwell-e off of x99 to a new chipset after all
|
|
#
?
Jul 18, 2014 01:16
|
|
|
I just hope there's an M-ATX X99 board. They've demo'd an ASRock one so surely it will hit production, right? 
Rime fucked around with this message at 02:25 on Jul 18, 2014 |
|
#
?
Jul 18, 2014 02:18
|
|
|
Are there technical reasons for Intel to abandon 2011?
|
|
#
?
Jul 18, 2014 02:45
|
|
|
|
| # ? Apr 27, 2024 14:29 |
|
|
Rime posted:I just hope there's an M-ATX X99 board. They've demo'd an ASRock one so surely it will hit production, right? Don Lapre posted:Are there technical reasons for Intel to abandon 2011?
|
|
#
?
Jul 18, 2014 05:41
|
|