Tuesday, July 21st 2009
First Intel Clarkdale Core i3 Low-Voltage Overclocking Feat Yields 4 GHz at 0.832 V
Intel's upcoming dual-core derivatives of the Nehalem/Westmere architecture, codenamed "Clarkdale" seems to have some interesting electrical characteristics. The CPU component of the chip is built on Intel's brand new 32 nanometre process that facilitates higher transistor densities, and in the process, intends to bring down TDP. An overclocking feat by Coolaler.com seems to suggest one of two things: either these chips have naturally low vCore voltages, or that the overlocking headroom at low-voltages is exceptional. Coolaler used a pre-release engineering sample of the Core i3 Clarkdale processor on a compatible platform, and achieved 4 GHz of clock speed with the vCore at 0.832 V. The frequency multiplier of the CPU was set at 25.0x, and a bus speed of 160 MHz used. Intel will be ready with these processors by the end of this year.
Source:
Coolaler.com
121 Comments on First Intel Clarkdale Core i3 Low-Voltage Overclocking Feat Yields 4 GHz at 0.832 V
I want to see benches with HT off...
4GHz is this retail speed or just some high speed prototype ?
5GHz or 6GHz dual core not bad you will get the same speed like 3GHz Quad core but with lower power consumption.
However, when Intel makes an affordable true 8-core 32nm chip, and if AMD doesn't have anything close that'll match it, I'll buy one from Intel, just because it's neat.Nope -- I have a feeling it's all going according to plan.Because they were trying to regain the performance crown (you want to make it as easy for people to adopt/upgrade to sell as many chips as possible).That's exactly it.AMD did it when they were ahead, performance-wise, and now Intel is doing the same (forcing you to buy/choose from tons of sockets/chipsets). It's a neverending circle. When/if AMD gets back ahead, they'll start doing the same thing again, probably.
early 775 boards cant run pentium D, later ones cant run core 2 duo, then theres ones that cant take quads, and then theres still ones that couldnt take 45nm.
There was many, many revisions of 775 - how the hell is having two main desktop sockets different?
With an integrated memory controller CPU controls memory support, so it makes sense that if you want different memory, use different sockets. AM2 vs AM3, 1366 vs 1156
AMD has similar issues, although they only started with AM2. AM2, AM2+, high wattage chips that blew out average boards. It sure as hell got confusing for average joe who wanted to buy "the best"
1. socket 1366: Triple channel ram, high end CPU's. more cores, more threads speed comes before anything else, chipsets based around 2 video cards, or more.
2. Socket 1156: dual channel ram, midrange CPU's, dual to quad core (cool running, power efficient chips for the average user - efficiency over speed). Mobo chipsets designed for 1 video card, two at most.
3. There is another, lower socket. Its designed for ITX platforms and such, with integrated graphics. It'll be just like the atom, perhaps even a replacement. No one whines that atom isnt 775, do they?
2, 4, 8, and 16 respectively.
They could have a native quad core be an i5, as well as a dual core with HT - i cant confirm it, but thats how it seems to be going.
As these sockets progress through time, I can guarantee you there will be more chipsets developed. You can't expect these to last anywhere near as long as 775 and stick stick with the original few chipsets that were released, it just isn't going to happen.Yes, but none of that is solved by a dual(or triple) socket system. As new processors are released, having multiple sockets doesn't help guarantee that they will be compatible with older boards. Most of the reasons older 775 boards were not compatible with newer processor either came down to the chipsets not supporting the processors, or the electrical specifications changing. Having multiple sockets does not solve this issue, we can't say that the X58 or P55 chipset is going to support processors released 4 or 5 years down the road, and we also can't say that Intel isn't going to make an electrical change on newer processor rendering the current two sockets useless.
And of course no one whines that Atom isn't 775, just like they don't whine that the mobile processors use a different socket than the desktop processor. People don't buy an Atom machine expecting to slap a core 2 in it. However, you better believe there will be people buying an i7/i5/i3 system, and expecting to be able to put a better i7 in it, only to find out the best i7's they were looking to use won't work because they use a different socket...
How many topics a week do you think we will get on this?
It doesnt solve many problems, but it doesnt ADD any either - its no different to how AMD had socket 939 and socket 940, one being mainstream and one being server.
All intel have done, is made a second series which costs more money for the niche market of entuhsiasts and small business/home servers. as far as 90% of the world cares, 1366 may as well not exist - they'll never see it, and they'll never buy it. its like the intel extreme CPUs.
lolololol:roll:
So if I may, from what I understand the situation is something like this:
LGA775 -nada lowend i3
LGA1155 - Dual channel DDR3 memory controler, integrated graphics, (?) no pci-e controler lowend/middle i5
LGA1156 - Dual channel DDR3 memory controler, pci-e controler middle/highend i5/i7
LGA1366 - Tripple channel DDR3 memory controler, ... Ultra highend i7/i9
This was a very wothwhile debate :)
If they get chips at 1.0v or lower stock for 1155, we'll be seeing one hell of a leap in CPU power in the mobile arena. single core atom, pssh. lets go dual core w/ HT :D
Yes, the 1366 was a niche market, but only because Intel made it so. If they didn't release 1156, and instead launched i5 and i3 processors for 1366, then it would be mainstream.
The problem now is that people are going to buy an i5 machine, and want to upgrade to i7 or i9 later, and find out they can't because it uses a different socket. It is confusing and dumb. It gets even more confusing when people with i7 processors go to upgrade to another i7 processor...and they can't!:banghead: Imagine that. How idiotic would it be if someone bought a Core 2 desktop only to find out they couldn't upgrade to another Core 2? Well that is exactly what Intel is doing...:banghead:The problem is, it doesn't help with compatibility. Even though most of the chipset is on the processor, that doesn't guarantee compatibility. Even in the 775 days, the main reason for the needing of a new motherboard was generally caused by changes to the power requirements of the processors, not the chipset itself. The chipset supported the processor just fine, but the power setup on the board did not. That is why Pentium D's were not compatible with certainly older boards, and why the 45nm Quads were not compatible with certain boards. It might have even been a reason behind why the Core 2s required new boards also. But take a chipset like the 965P...that goes way back...and it supports virtually every 775 processor in existance when on the right motherboard.
And while the northbridge is on the processor, the southbridge isn't, and it still has to maintain compatibility with the newer processors, which isn't guaranteed either.
And again, compatibility is hampered by dual sockets. Having two different processor in the same generation, in the same series(Core i7) that won't work in the same motherboard doesn't sound like compatibility to me...:banghead:
the core2 COULD HAVE WORKED IN NORMAL 775 BOARDS, thats what it was tested in, BUT intel wanted to sell more chipsets, so they SOFTWARE FUZED the cpu to give an error and refuse to boot if on a pre core2 board.
I have owned many an AMD rig, and have never had a problem finding upgrades even YEARS after the socket came out, 939 was a semi-dark spot, but i avoided 939 as it didnt offer much more perf then 754 and was a good bit more expensive( no 15% in synthetic benches and 7% at best in rare real world apps isnt enough to justify the diff in cost)
I have no intrest in these chips/sockets, sure they will bench well, but fact is that im about 80% gamer, 10% video/audio encoder and 10% webhead, and AMD covers those bases VERY WELL.
slap in a current nvidia card and grab badaboom and you got better encode speed then any intel cpu can offer!!!
then again, superpi = useless benchmark across platforms.