Slashdot Mirror
Gigabyte Board Sets Intel X79 Overclocking Record
MojoKid writes "Renowned overclocker 'Hicookie' achieved a new high clock speed on the Intel Core i7 3930K processor by cranking the chip past 5.6GHz using a Gigabyte GA-X79-UD3 motherboard, the first mobo in the world to achieve a mulitplier of 57x. There was a bit of a scandal with Gigabyte recently when a YouTube video showed one of its X79 boards going up in smoke. Gigabyte released a BIOS update for several of its X79 boards to prevent such incidents from happening, and there were outcries that the new F7 BIOS would ... [reduce] overclocking performance; Hicookie's achievement should erase those concerns."
A way to compile Gentoo in a reasonable amount of time.
Obviously you're misinformed, since a gentle overclock, say a few hundred Mhz, is well within the average chip capacity, with no appreciable impact on chip lifespan compared to the gain.
Really, by the time you hit the edge of the lifetime, your chip's performance would be worth replacing anyway.
For some applications 2 cores at 5.6GHz are better than any number of cores at 3GHz.
5.6 GHz should be enough for anybody. also, at 1.21 GW power draw, it will take you back in time.
Overclocking rarely causes damage. If you gradually increase the speed, you'll hit the point where your computer isn't stable. However, this is well before the point where you cause permanent damage. As long as you stay reasonable (don't change voltages) you're getting a good performance gain for free. Why not get a 30% performance boost?
Actually, back in the day, it was the opposite of stupid. I bumped my old K6-III by 25%. I used it for video transcoding and a 25% increase in performance was huge. It could take a full day to do a high quality multi-pass of an hour of video. Shaving 5 or 6 hours off that was kind of a big deal. Sorry you couldn't figure out how to do it right.
8585.05 is is a bigger number than 8320.4. Do you see how that works?
If you do a gentle OC without touching the voltage, you can't damage the chip. Two things damage a CPU: Heat and Voltage. Even if it locks up, it's not going to hurt anything. I squeezed a +10% OC and an 8% under-voltage on my GPU. 10% faster at full load and runs cooler than stock settings.
Just remember, AMD's 8150 FX processor hit 8.5 GHz with two cores.
"The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants." ~Thomas Jefferson
Because some of us would rather pay more to get a 30% performance boost without fiddling about trying to gauge system stability, and others of us are happy enough with the out-of-the-box stable systems that we have by default.
I've done my share of overclocking (having first overclocked a 386SX from 33MHz to 40MHz, a P100 to 120, and then some K6-2s from 300 to 350), but lately I'd rather just have a system that is both reliable and that doesn't need fucking-with.
YMMV.
Kid-proof tablet..
It would be refreshing to see some crazy underclocks too. For example, if you ran Sandy Bridges at 500MHz they would still be quite fast and not require much cooling.
There's not much fiddling about these days. If you're buying Intel stuff new, you really don't have much choice other than to get an unlocked multiplier - and setting a multiplier isn't exactly rocket science. If you stick to the basics (either undervolt at stock frequency or crank up the frequency at stock voltage), it should take about 10 minutes including reboots and stress testing.
AMD on the other hand might be a bit more complicated - I know I never got a stable overclock back during my Socket939 days (X2-3800+, crappy MSI board with a Northbridge fan that died after a year). I wonder if it's similarly easy there now... anyone care to chime in?
Even with older Intel systems it's not too fiddly - overclocked a Pentium Dual Core (cut down Core2Duo) based system from 2.6 to 3.4GHz a few days ago, that only took about 15 minutes. Crank FSB by 10 MHz, see if it boots, run Prime95 for 60 seconds and see where the temps plateau out... reboot, repeat. Eventually the board stopped booting (270MHz FSB) - would no longer find a boot device, so backed off to 260MHz, ran Prime95 for 10 minutes, temps under 50C with stock cooling - done.
Of course, you can try to get past the FSB block by lowering ratios for RAM speed and PCIe clock (although I'm not sure which clock ratio would correspond to my missing boot device problem), but when you stick to the basics and don't try to push the envelope all too far, overclocking really is pretty simple. And 15 minutes (including that last 10 minute run of Prime95) for a 30% performance boost (I mostly use the machine for encoding video so that I don't tie up my laptop for hours on end) is pretty decent...
Ya know, maybe its just me but I never did get the whole 'Must rule teh benches!" crap. I had a customer who would pay me to strip down his Windows installs just to get a few more FPS in the benches and i swear this guy went through so much hardware his grandma was on a skulltrail because that was the weakest hand me down he had!
But all the benchmark bragging kinda ignores what the chip makers really don't want to talk about and that is chips have been "good enough" for the vast majority for quite some time now. Thanks to the long tail of the consoles even the gamers have been able to stay with hardware for years and there just hasn't been any real "killer apps" come along in quite some time that needed a huge hardware increase which is why you see ATI pushing Eyefinity and Nvidia pushing CUDA because there just hasn't been any major need...well except for those wanting to win benches and stuff like TFA, but really how big of a niche is that?
I know I ended up getting rid of my full size notebook for a EEE E-350 netbook because i found when i'm out and about it was more than good enough for the tasks i had, and its taken nearly 5 years for the games to finally use enough CPU I'm gonna have to build the kids new triples or quads to replace their duals, so seriously what real good would you get from an aggressive OC like that besides winning some score? And are they bragging about less than 6GHz? Didn't AMD hit a record of 7 something last year at one of those OC contests? Seems a bit low to be bragging about.
Personally i'm glad things have slowed down, its nice being able to game on a $60 HD4850 and still have plenty of purty on my 1600x900 LCD, its nice not having to spend but around $100 each to upgrade the kids from Pentium Ds to AMD triples (may go with a quad, does anybody know if a 3.2GHz triple or a 2.2GHz quad would be the better route? both are the same price and I'm thinking clock will matter more, am I wrong?) and they can keep their HD4850s probably another year. So if that's your thing have fun, everybody needs a hobby, but it just seems a little pointless to me when i haven't seen a consumer app be CPU bound in ages.
ACs don't waste your time replying, your posts are never seen by me.
Too bad AMD still can't deliver on real-life processor performance. Bulldozer was nothing short of an embarrassment.
Such a small overclock will also have no appreciable impact on performance, either, outside of benchmarks.
The reason you can overclock without raising the voltage is that there is a voltage "guard band," which is like a safety margin. Some of my research has been about finding ways of reducing that guard band, because it's wasted energy. But that guard band is there for a good reason. Typically the critical paths in the chip (those with the longest propagation delay, which limit the safe clock speed) are a bit faster than the clock period. But that's only true whe the voltage is stable. If the voltage droops, then the propagation delay of those paths will increase, possibly too much, and you get incorrect computation. Voltage droops occur when circuits suddenly start switching a lot, demanding more current, or in other words, the effective impedence of the circuit drops, and by V=IR, for the current being supplied by the voltage regulator at that instant, the voltage inside the chip will drop. The regulator cannot respond instantly, so a guard band is provided so that the maximum droop never brings the instantaneous voltage below a certain margin. If you overclock without raising voltage, then your CPU will work fine most of the time, but certain workloads will cause wide swings in current demand, and if you execute one of those, you may crash your system.
This is why memory tests are worthless for stability testing, because due to cache miss latency, the current demand is relatively low and stable. Prime number generators are also not so great, because their current demand is relatively high and stable. I know that some of the SPEC and PARSEC benchmarks have some wild behavior, like FFT, for instance, or anything that has a lot of barrier synchronization. For the regular user, what's likely going to happen is that you'll get random such events where variation in cache hits and vector computation phases will cause significant spikes in current, and your game will crash.