God, let this berserkers rampage persist in these most honorable and historic times.
Printable View
God, let this berserkers rampage persist in these most honorable and historic times.
3.6GHz vs 2.72GHz
And the winner is conroe. The increase in CPU score alone is massive: lower clocks and 1000 points more and those clocks on the x1900 are measly.
Could be the same reason why I can't get a PCMark04 completed succesfully on grammar multi-thread test on my E6600. Dual Prime95 stable for 4Hrs+, PCMark05 stable, dual SPi 32M stable, 3dMark01se loop overnight, 3dmark03/05/06 passed without issues, AQ3 completed, I can throw any benchmark except PCMark04...Quote:
Originally Posted by freecableguy
using non Conroe e6600 CPU, PCMark04 works flawlessly on the same set up :confused: :confused:
AMD does have stuff up their sleeves... Will it be enough? who knows and we won't be seeing it till early 2007 either way. Apparently Intel wants to get back to 80%+ market share... I'm glad to see it... I am plumb tired of netburst. and good work FCG
Intel owns in am3, always have. I was beating guys by 20k with single dothan and 7800gtx256 that were using SLI setups with FX's.Quote:
Originally Posted by ex2cib
AM3 is NOT a good benchmark to compare with....it is so intel favored.
A 780m on air crushes the cpu score of the 3.7ghz fx60 and pretty much matches the conroe cpu score at same mhz:
dothan @ 2700mhz 7900GTX@ def clocks
http://pic20.picturetrail.com/VOL256.../142920102.jpg
http://pic20.picturetrail.com/VOL256.../142920104.jpg
Could be the overhead of SLI/nv drivers on the cpu dragging down the FX60, or it could be that the intel chips just own in am3 period. Lots of other variables there including x1900 vs 7900gtx, single vs dual gpu, etc..
Regardless, conroe does looks like it could do some damage in AM3 in the right hands. Nice prelim benches Kris.
In 06,05, and 03 it is adiff story though. Got any runs to post up using the exact same clocks you ran there in the am3 with the other benches I mentioned? Maybe something with def gpu clocks in xfire? I'll match your clocks with the SLI rig for compare.
For what i know AM3 is very good indication how cpu performs in games. It's 3dmarks that are far more syntetic, your score could vary on many things.Quote:
AM3 is NOT a good benchmark to compare with....it is so intel favored.
A 780m on air crushes the cpu score of the 3.7ghz fx60 and pretty much matches the conroe cpu score at same mhz:
dothan @ 2700mhz 7900GTX@ def clocks
my dothan does much better at quake4 and doom3 then x2 with same card @same clock. I dont see why am3 is Intel favored there use to be times not long ago when a64 could compete at it. Im surre Conroe is not worce then PM:D Looks like someone needs anew setup to play with;)
Games railer? What are those?...LOL. 3d benchmarks are the ultimate game for me. My everday rig is 3.0c/9200se. Can't play much on that though :)Quote:
Originally Posted by railer
Seriously though, from my experiences, dothan/intel does better than AMD/SLI in am3 at same clocks on cpu end. I think I could still spank my best amd/SLI score with dothan and single 7900gtx. It would be really cool to see someone running SLI with those hacked drivers in am3 with a conroe. I am sure that combo is something alot of peeps would love to see....hint hint Kris.
AND I do have a new setup to play with, but it's not conroe/yonah/xfire ;).
How about Cinebench?
:toast: I think exactly the same! It is MUCH BETTER to save money for buying good videocard (may be couple of them) or... physical accelerator on AGEIA PhysX in the near future. This will have much more impact on real gaming experience than shuffling your CPUs. Do not let syntetic benchmarks to waste your money in this nonsensical way.:)Quote:
Originally Posted by Pillo-kun
Just a quick quip -- FCG's chip is an A1 stepping, release version will be B0 or B1 stepping, i.e. one major and perhaps a minor revision before release, expect 5-10% more performance as Intel engineers work out the design bottle necks and target critical process steps better.Quote:
Originally Posted by sabrewolf732
The clocking of these chips are certainly pointing toward a 2.9 or 3.33 GHz version on the horizon.
Yep plus then also add a FULLY supported motherboard and Bios then add another few % performance ;)Quote:
Originally Posted by JumpingJack
WOW...... Dont know much about overall system performance, do you?Quote:
Originally Posted by Jimd
CPU is everything...... I went from a P4 @ 3.6Ghz to a P-M @ 2.8Ghz and the games FPS just sky rocked with the same clocked X850XT PE.....
Alex
Of course! But I just mean that it is completely wrong way to change your present powerful CPU (as Athlon 64 3700+ for ex.) with the any more powerful one. If you care about the gaming experience (not highest marks in CPU benchmarks)!Quote:
Originally Posted by AgonxOC
If you have not-top-end videocard you can spend equivalent amount of money on upgrading your video than you will have smoother picture (without bad slow downs in hot episodes).
Or you can wait till AGEIA accelerator will be on retail and you can get with it really impressive picture that no one conroes or the CPU at all can show to you.
I just see how it works, so I was impressed. And maybe I buy conroe (or whatever AMD will prepare to answer later) but before I will buy some of the physical acceleration stuff.
Cool, look forward to seeing some results..:)Quote:
Originally Posted by k|ngp|n
Quote:
Originally Posted by k|ngp|n
yea, you do have a point about intel owning am3 now.
am3 seems to like CPU efficiency more than anything else
i mean
P4 < A64 < dothan < conroe
it also likes ram timings/bandwidth. ive had my a64 @ 2700mhz before with higher cpu scores than an a64 @ 2850, just because of extra ram bandwidth and timings (/screams at venice for not wanting to clock ram anymore, rip 313 2.5-3-3-7 :()
then you have no dual core support on am3 either.
what makes dual cores helpful are the dual core graphics drivers
ive always like AM3 as a benchmark, since it ran in a minute or so.
the next time someone posts scores, can you add the max and min FPS as well? those are just cool to look at as well:)
good work on the 150k's FCG & KP, makes me sad though, as you doubled my max score:p:
personally, i think it's largely a smaller L1 cache that helps conroe with this benchmark. The smaller cache size means data is spending less time on the core, and latency, at this point, is less.
As driver commands are relatively small, the shorter time spent on the core, rather than it fitting "inside the cache" as someone else posted, as well as the faster transmission to the vital "part"(chipset, gpu, drive controller, audio) as a side-effect of the smaller cache, is why we see such a boost.
This would explain, very easily, the huge growth in 3d-scores, rather than cpu scores, and we can just forget the extra execution unit that Conroe has.
Well not to mention the entire Intel pre-core lineup right? This has got to be the greatest preformance leap in cpu history, Usally its a decrease due to lower starting freq. and systemarchitectural first batch issues on supporting hw. It´s simply earthshattering. :woot: :clap: Way to go Intel!Quote:
Originally Posted by eBoy0