AMD Phenom II and Dragon Preview Event - Invite Only

Printable View

Show 100 post(s) from this thread on one page

12-10-2008, 04:57 AM
dinos22

charles

spill the beans man spill the beans :D
12-10-2008, 05:20 AM
chew*

For those unfamiliar or familiar with the area there is a best buy we could all park at and carpool from a mile away, I have a feeling parking is going to be a pain knowing the area fairly well.

The one off of frontage rd, Enjoy you most likely know exactly where i'm talking about.

14 Allstate Rd Dorchester, MA 02125
12-10-2008, 05:47 AM
informal

Quote:

Originally Posted by FUGGER

I got a peek at CPU scores in 3Dmark 06 at 6ghz, do you guys really want to know the number?

It's not that hard to guess.It can't be lower than Phenom I per clock,which scores ~4000 in Vista at 3Ghz.In XP,Phenom scores 10% better(~4400pts),Vista kills the CPU score for some reason when Phenom is in question.
So if it was Vista(like i suspect),than the CPU score at 6Ghz should be somewhere between 8200 and 8500 points,assuming 6% better IPC than Agena and 96% efficiency in scaling with clock from 3 to 6GHz.Am i close? :)
12-10-2008, 05:58 AM
EnJoY

Quote:

Originally Posted by chew*

For those unfamiliar or familiar with the area there is a best buy we could all park at and carpool from a mile away, I have a feeling parking is going to be a pain knowing the area fairly well.

The one off of frontage rd, Enjoy you most likely know exactly where i'm talking about.

14 Allstate Rd Dorchester, MA 02125

Want to plan on that then?

Does someone have a large enough vehicle to car pool. I have just a little two door RSX. But that'd be cool to meet up close by and carpool in.
12-10-2008, 06:06 AM
chew*

Quote:

Originally Posted by EnJoY

Want to plan on that then?

Does someone have a large enough vehicle to car pool. I have just a little two door RSX. But that'd be cool to meet up close by and carpool in.

I have access to a gmc express 3500 cargo van, not the most comfortable but its just a quick ride over. LMK
12-10-2008, 06:40 AM
ChristmasGT

Quote:

Originally Posted by informal

It's not that hard to guess.It can't be lower than Phenom I per clock,which scores ~4000 in Vista at 3Ghz.In XP,Phenom scores 10% better(~4400pts),Vista kills the CPU score for some reason when Phenom is in question.
So if it was Vista(like i suspect),than the CPU score at 6Ghz should be somewhere between 8200 and 8500 points,assuming 6% better IPC than Agena and 96% efficiency in scaling with clock from 3 to 6GHz.Am i close? :)

Is it because Vista has the TLB Patch enabled by default?
12-10-2008, 07:28 AM
massman

Quote:

Originally Posted by FUGGER

I got a peek at CPU scores in 3Dmark 06 at 6ghz, do you guys really want to know the number?

Yes please.
12-10-2008, 07:30 AM
ryboto

Quote:

Originally Posted by chew*

For those unfamiliar or familiar with the area there is a best buy we could all park at and carpool from a mile away, I have a feeling parking is going to be a pain knowing the area fairly well.

The one off of frontage rd, Enjoy you most likely know exactly where i'm talking about.

14 Allstate Rd Dorchester, MA 02125

Dang, it's only 40 min from me, but I have to work. Straight shot down I93.
12-10-2008, 07:32 AM
Movieman

Quote:

Originally Posted by chew*

If I win a phenom II when i go I will be holding some form of contest for those not fortunate enough to go as I have no plans to use a ddrII variant unless its on a reference motherboard which at this point in time I do not think I will have access to.

Stay tuned ;)

If you win a Phenom II you don't have to worry at all.
Enjoy will mug you for it on the way back to the car!:rofl:
12-10-2008, 07:40 AM
demonkevy666

Quote:

Originally Posted by ChristmasGT

Is it because Vista has the TLB Patch enabled by default?

is it disabled with service pack 1.

maybe that's why I only score 4,000 cpu points in @3.0ghz
12-10-2008, 07:50 AM
D-Cyph3r

Early Retail PII 940 tests

He's gonna be trying on water later. :D
12-10-2008, 07:56 AM
GenTarkin

if retail phenom on avg all of these are capable of such speeds, this will be the chip to have =) I cant wait!!!!

I just have to figure out what mobo to pair it with.
12-10-2008, 07:57 AM
Movieman

I just sent thru a request that they take one of the air cooled machines and set up BOINC on it and run the WCG HCC project on it.
That would show a very real world comparison to the Intel based systems.
12-10-2008, 07:59 AM
massman

Real world to you, not at all to me ;)
12-10-2008, 08:04 AM
chew*

Quote:

Originally Posted by Movieman

If you win a Phenom II you don't have to worry at all.
Enjoy will mug you for it on the way back to the car!:rofl:

As long as he folds with it that will be ok as that will be part of the criteria for the contest. :D
12-10-2008, 08:09 AM
Movieman

Quote:

Originally Posted by massman

Real world to you, not at all to me ;)

If all the cores are at 100% load and working a problem(WU) thats real world as much as encoding or anything else I can think of.
check and mate.:D
12-10-2008, 08:09 AM
Slovnaft

Quote:

Originally Posted by Movieman

I just sent thru a request that they take one of the air cooled machines and set up BOINC on it and run the WCG HCC project on it.
That would show a very real world comparison to the Intel based systems.

with a side by side comparison maybe :rolleyes:
i could haul up a chip :rolleyes:
12-10-2008, 08:12 AM
Halk

So it's the 20th of December? I asked my shop and they said January...

I've just gone and bought a throwaway AM2 CPU so I can get a new board and memory etc set up and on water for the 940 coming out.
12-10-2008, 08:13 AM
Movieman

Quote:

Originally Posted by Slovnaft

with a side by side comparison maybe :rolleyes:
i could haul up a chip :rolleyes:

I could bring my i7-965 system at 3733mhz but at this level these guys would laugh at me. I setup for long term stability not max bench.
12-10-2008, 08:15 AM
massman

Quote:

Originally Posted by Movieman

If all the cores are at 100% load and working a problem(WU) thats real world as much as encoding or anything else I can think of.
check and mate.:D

And of course everyone who buys a new processor is intended to use it at 100% load all the time.

Real world =/= maximum load. Real world = internet, gaming, video editing, ...
12-10-2008, 08:19 AM
Movieman

Quote:

Originally Posted by massman

And of course everyone who buys a new processor is intended to use it at 100% load all the time.

Real world =/= maximum load. Real world = internet, gaming, video editing, ...

That is true and I do agree with you on that point BUT to gage the effectiveness of a processor or shall we say rate it against others the only reasonable way to do so is to have both run the same app and compare them.. As in the benchmarks that we run.
I assume your familar with benchmarks?:D
12-10-2008, 08:20 AM
Slovnaft

Quote:

Originally Posted by massman

And of course everyone who buys a new processor is intended to use it at 100% load all the time.

Real world =/= maximum load. Real world = internet, gaming, video editing, ...

thenagain this demonstration is drawing heavily on a crowd that has a large interest/fixation with DC. a demo in this capacity would go a long way in promoting this launch with the DC community.

Quote:

Originally Posted by Movieman

I could bring my i7-965 system at 3733mhz but at this level these guys would laugh at me. I setup for long term stability not max bench.

i've got 20x200 stable at 1.3v, cpu can also bench at 5.0+.
But i'm not anticipating intel hardware demos being introduced so last minute.
12-10-2008, 08:27 AM
massman

Quote:

Originally Posted by Movieman

That is true and I do agree with you on that point BUT to gage the effectiveness of a processor or shall we say rate it against others the only reasonable way to do so is to have both run the same app and compare them.. As in the benchmarks that we run.
I assume your familar with benchmarks?:D

I didn't say that the benchmark you suggested is not a decent benchmark, I just said that your claim as it being "very real world comparison" isn't always correct.

Next to the obvious fact that everyone has a different opinion regarding what benchmarks should be used to measure performance and how those measurements should be compared, there are much more other factors to take into account if you want a real world comparison: price, heatload, idle power drawing, availability and so on.

I'd say that your BOINC benchmark would give a fairly good indication of what this CPU can do in comparison to an Intel alternative. However, scoring lower doesn't mean it's worse, even in your benchmark. If it's 25% slower, but 50% cheaper, you could buy two and have a 50% faster platform :).
12-10-2008, 08:38 AM
chew*

Me personally I don't think although i could be wrong that deneb is meant to go head to head with i7 clock per clock. This is just based on an educated guess based on suggested retail price and current retail prices of the I920. Etc retail 3.0 deneb suggested ball park $300 I920 $300.

I would love to see how it compares to current intel quads however especially q9550 being priced in its range, If asked I would bring my current rig for a head to head comparison on air.

I'm thinking that Amd most definitely has the edge here with what apears to be a majority of cpu's stable @ 4.0 on air while most struggle for 3.6 on air with intel.
12-10-2008, 08:39 AM
Movieman

Quote:

Originally Posted by massman

I didn't say that the benchmark you suggested is not a decent benchmark, I just said that your claim as it being "very real world comparison" isn't always correct.

Actually the benchmark inside BOINC is a useless thing.
It is the timeframe required to do the work itself that shows the true power of any cpu.
I also can't think of any better way to measure a cpu's effectiveness than researching the ways to cure disease.
I know that I'm in the minority on that point on this forum where there are over 80,000 members and maybe 2000 that do this type of work but thats how I feel.

Quote:

Originally Posted by massman

Next to the obvious fact that everyone has a different opinion regarding what benchmarks should be used to measure performance and how those measurements should be compared, there are much more other factors to take into account if you want a real world comparison: price, heatload, idle power drawing, availability and so on.

That is correct and we agree on this.

Quote:

Originally Posted by massman

I'd say that your BOINC benchmark would give a fairly good indication of what this CPU can do in comparison to an Intel alternative. However, scoring lower doesn't mean it's worse, even in your benchmark. If it's 25% slower, but 50% cheaper, you could buy two and have a 50% faster platform :).

There's more to total cost than the chip itself.
Two system require additional space, a second PSU,hard drive,etc so many times the more expensive cpu can be a lesser total cost when all the factors are included.
This is why I buy server grade equipment not gaming systems.
Many benefits that more than offset the initial cost.
They don't break down like gaming systems do.
They require less maintainence than gaming systems.
I have 2-8 core clovertown machines here that have run at 100% load since December 2006 and with the exception of a 35 hour downtime to electrical loss from an ice storm are at 100% load 24/7..
That's TWO years at 100% load 24/7..
There simply is no gaming machine that will do this that I have seen.

Show 100 post(s) from this thread on one page