Slashdot Mirror
P4 2.2GHz Overclocked to 3.5GHz
GraveD sent linkage to a site
explaining how a homemade nitrogen cooling system
overclocked a P4
from 2.2Ghz to an incredible 3.5ghz. There's plenty of stuff
to poke at over there. Update: 01/17 20:42 GMT by T : boaworm writes: "According to this paper, the Finnish geeks have successfully oveclocked a Pentium 4 to 3675 Mhz. They claim it is a new World Record, and it sure looks like they beaten another O/C'd Pentium 4 submitted earlier today on slashdot. (Summary in English in the end)."
Didn't they start this campaign to get 'hacked' ? so they could close some more holes they couldnt find them selves ?
Now i wonder, it worked they all readdy found 7!
Quazion.
If it's "let's attack the binary and see if we can break it", that's potentially harder to catch something like this, but then again, how hard can it be to see if the binary links against the system C library at the known offsets of gets, fgets, sprintf, etc.
What would be lamest of all is if the certification process goes something like, "What's your security engineering process? Oh, sounds secure to us."
The unique thing about software is that it is infinitely clonable. Once you've written a subroutine, you can call it as often as you want. This means that almost everything we do as software developers is something that has never been done before. This is very different than what construction workers do. Herman the Handyman, who just installed a tile floor for me, has probably installed hundreds of tile floors. He has to keep installing tile floors again and again as long as new tile floors are needed. We in the software industry would have long since written a Tile Floor Template Library (TFTL) and generating new tile floors would be trivial.
h tml
from http://www.joelonsoftware.com/news/fog0000000337.
>We try harder." [Avis Car Rental] - Harder than >what? Yesterday?
You're too young, no doubt, to remember the Slogan Wars between Avis and Hertz of the early 60's.
In those days, it was considered taboo for an advertiser to directly mention the competitor's product when making comparisons. In fact, it was quite a shock when, in the mid 1970's we started seeing TV commercials where one brand explicitly stated that their product was better than a specific competitor's product. It's pretty common now, but you never saw it back in the day.
Anyway, some consumer survey gave Hertz marketroids the idea that they were the #1 car rental company (in an unbound domain, with unspecified terms, naturally). Hertz went to town
with this "fact." Worthy of note, the Hertz sign atop the infamous Texas School Book Depository building.
Avis countered Hertz with their own ingenious slogan: various flavors "We're #2, but we try harder."
At the same time, they made yet another marketing innovation -- they designed all their ads so that they could be distinguished at a distance of 40 feet. Thank Helmut Krone for that.
-fb Everything not expressly forbidden is now mandatory.
A 3.5GHz P4 probably would perform like a 2.5GHz Athlon, given the difference in IPC. However, factor in SMT (HyperThreading) into the equation and it gets a lot more interesting. Hammer will have some competition when it comes out, even with a PR rating of 3400+ - the P4 will probably get to 3GHz by the end of this year.
In the end, the consumer is the one to win. But remember, speed in a processor is only good if the rest of the system can keep up with it. Witness i845 (the SDRAM version) as a way of making a fast P4 perform even worse than before.
I am more interested in the upcoming GeForce 4 and R300 chips myself as a way to increase gaming performance - processor power is secondary, as long as it is sufficient. For rendering performance however, I am interested in fast processors, and it looks likely that SMT P4's will rock with Lightwave 7b on a quad CPU board (8 virtual processors!). Not that I could afford one of these anyway, so the point is moot.
I've got a plot showing SPECint2000 vs SPECfp2000 for eight different chips, including the Pentium 4 2.0 GHz.
From the looks of it, overclocking to 3.5 GHz might make the Pentium 4 almost equal in performance to the IBM Power4 running at 1.3 GHz.
"Provided by the management for your protection."
Damn, you were faster. :)
:-)
Here's that english summary from muropaketti:
English summary!
Today we cooled the new Intel Northwood 2,2GHz CPU with liquid nitrogen (LN2 -196C).
The motherboard used in the tests was Asus P4B266 based on the Intel 845 chipset (DDR). There was a voltage modification on the motherboard which allowed the VCore to be raised as high as we needed. The memory module was Crucial PC2100 128MB and memory settings were the fastest possible (CAS 2 2-2-5).
We used a copper bowl on top of the CPU and poured some LN2 into it. It took a while until the CPU temperature started to drop and when it was cold enough, we started the test.
First test was run at 3300MHz (FSB 150MHz) and with no problem at all (VCore 1,9V). The next step was rather high but after raising Vcore to 2,05V Northwood worked stable at 3520MHz (FSB 160MHz). We went on with the tests and finally hit the limit.
We were able to boot to Windows 2000 when the CPU clock frequency was 3675MHz (FSB 167MHz) but we couldn't run any benchmark programs. The highest STABLE CPU clock frequency we were able to reach was 3630MHz (FSB 165MHz). At 3650MHz we were able to run heavy benchmark programs such as SuperPi and Pifast successfully although the VCore was quite high (2,12V). It seems that Pentium 4 can handle it without any conflicts.
Check out the pictures above
I think the 3675MHz Wcpuid-shot we were able to get can be considered as the overclocking world record at this moment (17/01/2002), but I'm pretty sure the Japanese will try to beat it as soon as possible
BTW, Quake 3 Arena was quite fun to play when the CPU was running at 3500MHz! o_O
I got a P4 1.4GHz at work a few weeks ago. I have a Athlon 800MHz at home. The RC5 client from distributed.net runs at 2.9 Mkeys/s on my home system. My machine at work only runs the client at a whopping 2.4 MKeys. So based on my result, a 3.5GHz P4 would be like a 1.8GHz Athlon.
Flaming/joking aside - anybody know why the RC5 client does so poorly on a P4 compared to a much slower Athlon?
But I didn't overclock the processor - I overclocked the ISA bus!
The standard speed for an ISA bus is about 8 MHz, but my motherboard had jumpers for running it at different speeds. I had that baby running at 20MHz, and was lucky enough to find an ISA video card and network card that could run at that speed!
It really helped bump up the FPS when playing doom. <g>
I'm a leaf on the wind. Watch how I soar.