I checked the best Conroe processor computer in the Top 200 Computers and it is the #22 one, delivering presently 2039RAC (that may increase) owned by MSattler. I checked his results and see that it's doing about twice as fast as my D950 - taking half the seconds. I see he's using the $1100 (per Pricewatch) X6800 that runs just short of 3 megahertz. I don't know how much less cheaper Conroes do. Both of us are using Simon, Chicken of Angnor's cruncher. There are two other Conroes several positions down, in the 1650 range.

---

I checked the best Conroe processor computer in the Top 200 Computers and it is the #22 one, delivering presently 2039RAC (that may increase) owned by MSattler. I checked his results and see that it's doing about twice as fast as my D950 - taking half the seconds. I see he's using the $1100 (per Pricewatch) X6800 that runs just short of 3 megahertz. I don't know how much less cheaper Conroes do. Both of us are using Simon, Chicken of Angnor's cruncher. There are two other Conroes several positions down, in the 1650 range.

Yeah, that would be me. I think the RAC has about maxed, but I got stupid last night and I have an E6700 coming. Addicted to Seti. Well, I guess there are worse things.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

I checked the best Conroe processor computer in the Top 200 Computers and it is the #22 one, delivering presently 2039RAC (that may increase) owned by MSattler. I checked his results and see that it's doing about twice as fast as my D950 - taking half the seconds. I see he's using the $1100 (per Pricewatch) X6800 that runs just short of 3 megahertz. I don't know how much less cheaper Conroes do. Both of us are using Simon, Chicken of Angnor's cruncher. There are two other
Conroes several positions down, in the 1650 range.

Can you hear a processor squeak? Sure you can. It's just a gas getting an off-the-shelf processor to perform what the Cray's used to do.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

hmmm, i think life was more simple with the freshly baked pentium Is.... you know.... 100Mhz, weeks to crunch one WU...

---

I checked the best Conroe processor computer in the Top 200 Computers and it is the #22 one, delivering presently 2039RAC (that may increase) owned by MSattler. I checked his results and see that it's doing about twice as fast as my D950 - taking half the seconds. I see he's using the $1100 (per Pricewatch) X6800 that runs just short of 3 megahertz. I don't know how much less cheaper Conroes do. Both of us are using Simon, Chicken of Angnor's cruncher. There are two other Conroes several positions down, in the 1650 range.

Yeah, that would be me. I think the RAC has about maxed, but I got stupid last night and I have an E6700 coming. Addicted to Seti. Well, I guess there are worse things.

Don't feel too sad about your 6700. I just looked at the Top 200 and saw, several places down from your X6800, a 6600 @2.40 GHz, #2585744, of a certain "Anonymous" that's (sorry to say) outdoing yours and is presently in the 1900s RAC. Why do I know it's outdoing yours? I see that it's doing well over twice the speed of mine. That processor sells for about $375 at PriceWatch. He's using Crunch3r's cruncher. Maybe he has it highly overclocked but that is only a guess.

---

Check out my Conroe 6600 installed last Tuesday and RAC going up

http://setiathome.berkeley.edu/show_host_detail.php?hostid=2606440

Wow! It must be a lot faster than I thought. I was looing at some of the results and it finished a workunit in a 1/4 of the time it took a 3Ghz P4!

---

Check out my Conroe 6600 installed last Tuesday and RAC going up

http://setiathome.berkeley.edu/show_host_detail.php?hostid=2606440

Wow! It must be a lot faster than I thought. I was looing at some of the results and it finished a workunit in a 1/4 of the time it took a 3Ghz P4!

Nice stats that machine, eh?

---

Well, right now my X6800 is in the 21st spot at 2064, probably close to it's limit I think. And my E6600 is in about 40th spot at 1591, and still rising. (The chart has not been updated to reflect these numbers as of this writing). I'm gonna be cranking up an E6700 this weekend, we'll see where that takes us....

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Thanks Michael, I have slightly overclocked Conroe 6600 to 2.61G and still using air cooling, runs at about 45c and load is about 160watts, not bad eh! and usually only powered on for about 14hrs/day.

---

Geoff, I spot-checked how many seconds it's taking your machine to crunch workunits and it looks like it's taking about 2/3 as much time (1.5x the speed) as my PD950. For 24 hours mine's doing close to 1200. So your machine might top out at around 1800 per 24 hours if nothing changes. I see you're using Crunch3r's cruncher, too. One can always get a quicker estimate of RAC by averaging daily differences of total for that machine and correcting for hours per day operation.

---

Clyde, thanks for working it out, I think you are about right. The fans are a bit noisy and it isn't far from the bedroom so that's why it is off overnight! I have tried Chicken's on both this 6600 and my P4 2.4 and seem to find Cruncher's slightly faster and on the 6600 I have only Cruncher's SSE2 version. I would try overclocking it a bit more but it is on Warranty. BTW I have always about 600+ total pending credits but this will even the figures out over a period of time and the 6600 has only been crunching for 14 days and total credit is near 18,000 that's 1285/day.

---

Thanks Michael, I have slightly overclocked Conroe 6600 to 2.61G and still using air cooling, runs at about 45c and load is about 160watts, not bad eh! and usually only powered on for about 14hrs/day.

I get my core 2 duo tomorrow....should have it crunching by tomorrow night or Thursday latest.

---

You are correct that there have been some prelim. indications that that the shared I/O bus reduces potential performance.

No doubt about it.
Intel's new Dual Independant Bus has improved things considerably (along with the new Core 2 Duo based CPUs) as they are now able to match & even beat the Opteron systems in Single & Dual CPU configurations.
Previously with the Xeons the more CPUs you had, the smaller the increase in performance due to the I/O bottleneck; the new Core 2 Duo based CPUs & the DIB alleviate that problem- to a point.
With system that have 4 or more CPUs, the Opterons are still untouchable due to the I/O bottleneck that still remains with the present Intel I/O architecture.

Does this hold with the Xeons released this week? I think they are packaged in groups of 8. It seems stupid to do this if the performance wasn't there.

---

May this Farce be with You

You are correct that there have been some prelim. indications that that the shared I/O bus reduces potential performance.

No doubt about it.
Intel's new Dual Independant Bus has improved things considerably (along with the new Core 2 Duo based CPUs) as they are now able to match & even beat the Opteron systems in Single & Dual CPU configurations.
Previously with the Xeons the more CPUs you had, the smaller the increase in performance due to the I/O bottleneck; the new Core 2 Duo based CPUs & the DIB alleviate that problem- to a point.
With system that have 4 or more CPUs, the Opterons are still untouchable due to the I/O bottleneck that still remains with the present Intel I/O architecture.

Does this hold with the Xeons released this week? I think they are packaged in groups of 8. It seems stupid to do this if the performance wasn't there.

Yes, the new Xeons have a single FSB per socket which helps performance tremendously (instead of each CPU sharing an already clogged FSB).

Incidentally, the name of this technology isn't Dual Indepandant Bus (DIB), as that is a technology that was released during the Pentium II. The DIB was, essentially, the fact that the Pentium IIs had a slower L2 cache that ran at 1/2 clock speed, running on a seperate bus than the main bus, a.k.a. the "Front" Side Bus. This seperate, independant bus was known as the "Backside" bus - hence you had dual independant buses.

The L2 cache in modern Intel processors still sits on this "backside" bus, but typically runs at full CPU speed for best performance.

---

Incidentally, the name of this technology isn't Dual Indepandant Bus (DIB), as that is a technology that was released during the Pentium II.

Ah, thanks for that.
I knew the term sounded familiar, but couldn't figure out why.

---

Grant
Darwin NT

I'm crunching with my new E6400 now.....let'er RIP!

---

Yeah, people have :o)

You may find this review (select other pages from the link) interesting. The link points directly at "Power Usage & Performance Scaling" and contrasts Core Duo, Core 2 Duo and Pentium-D systems.

Regards,
Simon.

Simon,

The big problem here -- the example "low power" machines are nearly double the power consumption of my Athlon XP. I know it's probably because of the video card (I don't need or want a Gamerz video card), but if I go by the review, I'll be paying more, not less.

-- Ned

---

Yeah, people have :o)

You may find this review (select other pages from the link) interesting. The link points directly at "Power Usage & Performance Scaling" and contrasts Core Duo, Core 2 Duo and Pentium-D systems.

Regards,
Simon.

Simon,

The big problem here -- the example "low power" machines are nearly double the power consumption of my Athlon XP. I know it's probably because of the video card (I don't need or want a Gamerz video card), but if I go by the review, I'll be paying more, not less.

-- Ned

Ned you might check these power consumption benchmarks between the new Core2 Duo and the Athlonby machines at Toms Hardware
http://www.tomshardware.com/2006/08/02/thg_tuning_test/page4.html

---

When we finally figure it all out, all the rules will change and we can start all over again.

Yeah, people have :o)

You may find this review (select other pages from the link) interesting. The link points directly at "Power Usage & Performance Scaling" and contrasts Core Duo, Core 2 Duo and Pentium-D systems.

Regards,
Simon.

Simon,

The big problem here -- the example "low power" machines are nearly double the power consumption of my Athlon XP. I know it's probably because of the video card (I don't need or want a Gamerz video card), but if I go by the review, I'll be paying more, not less.

-- Ned

Ned you might check these power consumption benchmarks between the new Core2 Duo and the Athlonby machines at Toms Hardware
http://www.tomshardware.com/2006/08/02/thg_tuning_test/page4.html

Same problem.

My target is in the 60w range for the entire system.

The lowest power on the graphs is 160w.

---

Same problem.

My target is in the 60w range for the entire system.

The lowest power on the graphs is 160w.

Not sure you can get it down to 60W, My Pent M, with 6600 graphics card, indicates 72 to 75 watts as seen by the UPS when crunching and the screen is on. Measured at the wall socket the power drawn only drops about 5 watts when it goes to blank screen.
That was done with monitor on separate wall socket. The stated power consumption of Pent M is 27 watts. The mobo is Aopen i915Mm-HFS which uses mobile chipset.

Andy