@dr. who
mind blowing man! you do put things into perspective very nicely. you guys do great work, i am a big fan :)
keep up the good work!
Printable View
@dr. who
mind blowing man! you do put things into perspective very nicely. you guys do great work, i am a big fan :)
keep up the good work!
Well that's nice to hear, at least we have some form of 'mole/insider' at Intel.Quote:
Originally Posted by Drwho?
Now to play the waiting game....I'll pass on Sandy Bridge but look forward to the whatever revision it brings. A refined 32nm process should make for some cheap hexacores.......eventually ;P
Francois,
Make sure you eat right, exercise and do whatever to keep in shape and be healthy. We need you to work at Intel well passed your retirement age :D
I like the way you think. Intel should never restrict overclocking. After all there is a very small niche of people that even know about it. In my 15 years of IT carrier i have met many professionals and not many even knew about it and even less tried to overclock. I can count those that tried on one hand:rofl:
Keep the good stuff rolling.
Just a side note.Yep, in many cases you can, just have to have the knowledge ,some software and a interface.Quote:
you can't upgrade your car for free.
Many cars with different specced HP on engines, have exactly the same engines, there are some minor differencies but you can extract more power without physically changing anything.
What dr.who says sounds encouraging, however, i hope its not just PR talk (as in no details).
If sandybridge will be overclockable and bulldozer fails, they gonna have a new customer.
I'm really excited to see exactly how far you can overclock the "K" CPUs, the fastest SKU goes to 3.8GHz on turbo and concidering typical 24/7 on todays CPUs lies there at 4~4.2GHz and Intel lab stability != our LinX 3hr+ or whatever stability (not to mention the TDP requirements). Those high factory frequencies really is promissing, if SB overclockability ratio wouldn't become worse than what it's on bloomfield, lynnfield etc. we would be at like 4.7~5GHz on air lol. Ofc things don't always work like that and could be that clock frequency scaling is a lot worse on this architecture (more leakage) due to added complexity but WHAT IF, would be sweet to get like 20% frequency boost on top of whatever IPC increase SB has.
My point is that there is no equivalent to the almost guaranteed, risk free, free overclocking done in the days of FSB / bclock, for both AMD and intel.Quote:
Originally Posted by T_M
But of course, we can argue about my car analogy as much as everyone wants.
I believe what he meant when he was talking about the quartz is that its not going to be a piece of cake to overclock anymore, as in the mainstream user or one with less knowledge or time can't really do much, but that us with our skills or whatever you want to call it, and our motivation mostly will prevail. So that way they get rid of the fact that anyone can follow a template and overclock thier 930 for example to 4ghz. I would think apart from the settings there will be some sort of physical modification or trick or something that involves some dedication. Like for example intel makes the mobo manufacturers make sure you cant unlock over 175blck, but doesn't tell them they can let you easily mod a large resistor or 2 and fix it. 50-70% of the new overclockers wouldn't dare take a soldering iron to their board b/c theyd be too scared and would be limited. the other 30% of us would b/c we have done it before. Or the engineers might do something like put in an extra resistor on the underside of the processor that limits overclocking, but we can remove or replace it to overclock. That would basically make it so that that 50-70% would buy a better processor, limiting the extreme overclockers numbers and regain the market share they lost when overclocking became very popular 3-4 years ago and exponentially rose. Or i might just be 100% wrong.
Very true @ similar rumors about Nehalem. Keeping fingers crossed.Quote:
Originally Posted by thebanik
Eeeeeeeeeeeeeexactly, hence intel removing the 'hidden value'. Not saying OC is dead, but the extra free value is.Quote:
Originally Posted by sin0822
I agree. All this "the more difficult the more rewarding" sounds just like bs to me.Quote:
Originally Posted by cegras
Its not a rumour about intel blocking bus OCing.Anand had the chips, most probably from intel ,and he verified thats true.Quote:
Very true @ similar rumors about Nehalem. Keeping fingers crossed.
Dr. who seems to be verifying it too, albeit in more rainbowish colors, as in "its good to have artificial lock, it lets you be creative in circumventing it".
Lets hope it will be possible without some extreme measures.
wow ... I did not confirm, neither deny ... this is intel PR job :) ... What I am saying is that when computers become more complex, the overclocking become a little harder, the design much harder ... That's all I said.
Don't make me say something I did not
Take it easy ... don't sell the skin of the bear, before you killed it ... (Funny translation from french :) )
Francois
For the fun of it ... this is what I get with Wimax in a cafe store:
http://www.speedtest.net/result/947291502.png
lol no no ... no cable, pure 100% wireless ... no string attached.
Hi Guys,
I've been looking at the news about SB on Techspot and the 'leak' on Netbook Italia. Now I havent read all 12 pages of this thread so I dont know if it has been brought up (probably has a thousand time knowing my luck) , but doesnt the new I3's just look like I5's with disabled cores/cache and with HT enabled?
i3 2100 @ dual core 3.1ghz 3mb cache (is half of) i5 2400 @ quad 3.1ghz 6mb cache no HT
i3 2120 @ dual core 3.3ghz 3mb cache (is half of) i5 2500 @ quad 3.3ghz 6mb cache no HT
So is Intel just disabling cores and enabling HT?? Maybe the i5's will unlock like a lot of current AMD Phenoms!?
Of course I am probably totally wrong and there will be other differences that I dont know about lol:P
James
ARRGGGHHHH, I've been having uberslow net since I moved back to this student appartment, it's supposed to be 5/5Mbit but I'm getting like 0.35 ~ 0.7Mbit speeds most of the time. I'm going nuts, can't even stream youtube videos in lowest quality without buffering. :nuts: Complained but they don't seem to be find where the problem is. It's usually a bit faster during night though so would have to stay up at night and sleep during day.Quote:
Originally Posted by Drwho?
Err, bulldozer is supposed to be considerably faster than K10.5, with 8 cores :p: Unless it only overclocks to 3 Ghz or something (possible after seeing K8) I don't see it completely failing.Quote:
Originally Posted by 'RaV[666
this thread was dead for like a month, why do people feel the need lately to resurrect old threads?
I dunno. I'm looking forward to seeing what sandy bridge does...
It's usually a result of the search function.
I'm acutally looking forward to more complexity, just not 1000 dollar processors with 400 dollar motherboards, like it is now, unless its a 500 dollar processor can be pair with a 200 dollar motherboard, and that be the highend, which is what sandybrigde seems like in general b/c you guys are taking everything off the board and putting it on the chip, soon enough the cpu will be the actual cpu unit and im cool with that as long as it doesn't cost me an arm and a leg.Quote:
Originally Posted by Drwho?
well... if its that, yes, thats something i actually like... making it trickier to overclock... but trickier and impossible are two things... and with everything being integrated on die, there are more things you just cant, ever, possibly, modify externally...Quote:
Originally Posted by cegras
if bclock overclocking on sb would be really tricky, great... but if the best you can get is a 5% overclock thats not tricky, thats broken... and if the only alternative we end up with to overclock is multipliers... well that makes overclocking even easier and more boring, and makes it less overclocking to begin with and more "tuning" and "tweaking" which is cool as well, but not really the same...
so from a positive side: overclocking will be more challenging, unlocked cpus for low prices, yay! :D
from a negative side: only 5% bclock overclocking, maybe 10% with tweaking... the only way to really "overclock" is paying extra for a K cpu and adjusting the multiplier in bios... lame! :/
i think 1155 sb will usher in a renaissance of tweaking, cause most people will be stuck at the same multiplier and the same bclock... so then the only way to make a difference is to tweak in software, memory timings etc...
francois, you make it sound as if overclocking is oh so important for intel, and that there is no way they could cut down on overclocking let alone try to limit it... yet, correct me if im wrong, thats exactly what this bclock limit is about... i dont think it was done on purpose, like i said many times... it was done cause most of intel doesnt care about overclocking and they didnt want to have extra transistors in their budget and extra logic to debug... im just guessing here... but thats what it looks like to me...
so if intel would REALLY care so much about overclocking and the people that push it would REALLY be powerful enough to make sure overclocking stays in place and wont be limited or blocked, then how come they couldnt get a bclock to dmi clock divider into 1155? its not that many transistors and not that much debugging...
but im sure its enough to increase the cost of sb by enough to make the penny counters at intel want to save money instead... and the only way the good guys at intel could push for the divider is by going for a low volume highend platform... again, thats great on one side, but disappointing on the other, as the penny counters at intel seem to be too strong and the majority of intel seems to not understand the importance of overclocking...
anyways, thanks for all you and the other guys have done to "fight the machine" :D
The goal of overclocking is Intel's same goal: increasing performance while maintaining stability. At some point, Intel's efforts were bound to encroach on that of overclockers. Bringing more things on die increases the general performance; having the clock generators on die means that the signal degradation is significantly less, allowing for more internal components to stay in sync with less room for error. We continue to expect major performance increase with each new architecture, but want Intel to maintain the elements that allow us to change the internal workings on a whim. It's fun, and it has a value. Being limited to "tweeking" means that we can continue to have the performance enhancements we enjoy. A lot of it simply comes down to the architectural path. The steps made nehalem, which has some pretty amazing overclocks, led to those in sandy bridge. The writing was on the wall as it were. We all enjoy overclocking, but I doubt we want to be the "old goats" that put up a fight when things change. I say let things roll and enjoy the ride. :up:Quote:
...makes it less overclocking to begin with and more "tuning" and "tweaking" which is cool as well, but not really the same...
Why not try that again.....Quote:
Originally Posted by Drwho?
.....on LTE at a café in Stockholm :)
LTE? well, the specs are not even finished :rofl: :ROTF:Quote:
Originally Posted by Kej
SandyB + Wimax = Wroooooooooooooooooom ;-)
no, we want them to allow us to change it... at all... it doesnt have to be easy, id actually prefer it to be tricky like i said, and i think every real overclocker thinks the same :DQuote:
Originally Posted by xVeinx
your contradicting yourself... we are limited to X but will continue the same perf enhancements... if that was the case then X wouldnt be a limit now would it? :DQuote:
Originally Posted by xVeinx
huh? it started overclocking worse than the previous gen and then caught up with it :p:Quote:
Originally Posted by xVeinx
sure!... its not like we have an option anyways heh :DQuote:
Originally Posted by xVeinx
its great unless theres a wall and the wimax signal goes BOOOOOOM and crashes into the wall ;)Quote:
Originally Posted by Drwho?
i tried wimax in taipei and its a joke... great signal, enter any building and its very weak or even dissapears...
theres a reason intel spent billions to get on board of lte ;)
well, in Tokyo, there is no problem, neither in USA ... usually when you lose signal with Wimax, you will lose it with LTE too, because it is usually the steel bars in the concrete that are doing a F-cage and block the signal ...Quote:
Originally Posted by saaya
A lot of people speak about LTE ... Where is it ? The issue of Wimax deployment is even worst for LTE ... There is just no LTE network. there are only few little points world wide ... and Wimax is open on the patent side, not hostage of Quadcomm patents ... just saying ... ;-) :shrug: (This is my personal opinion)
Francois
And what can u say us about SB to LGA 2011?:) Its for 3Q or 4Q 2011?Quote:
Originally Posted by Drwho?