How come, AMD could have made the chip count blistering fast, but it would suck at everything else. Its only a small test of what the cpu capability. Wow it cant count to a million fast. What kind of program/game does that? What us enthusiasts want to know is. How good is it at games?
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
X4 PHENOM taken to 3GHZ by MOFO64
- Thread starter SOLDNER-MOFO64
- Start date
How come, AMD could have made the chip count blistering fast, but it would suck at everything else. Its only a small test of what the cpu capability. Wow it cant count to a million fast. What kind of program/game does that? What us enthusiasts want to know is. How good is it at games?
Not as good as the equivilantly priced C2D/Q
Utterly pathetic chip from AMD given how late it is to the market.
Kudos to the for trying though although I can't help but think the money spent on the amd chip would have been much better spent on a intel core 2 setup.
Also I don't understand your first point about the counting ability. Why would it suck at everything else if AMD made it count fast? Intel managed to design a chip that crunches through super pi like never before yet also trounces the competition at everything else.
Your points are arbitrary to conventional wisdom which shows that with current CPUs a close enough correlation exists between superPi performance and their capabilities in other areas - taking into account the number of cores and multi-thread capabilities - that the results mean more than squat.
Agreed, unless that all you do, i couldn't care less what it does in superpi all i would care is maybe a little encoding even then not a lot really, but the main thing would be games.
At moment the graphics card more important to me maybe in the future that will change not sure.
Associate
- Joined
- 24 Feb 2004
- Posts
- 1,083
- Location
- Leeds/Cyprus
The OP: don't you also post on XS? Didn't know you were also registered here. How come you were banned from XS if you don't mind me asking?
BTW, isn't the retail version of the chip supposed the B2 stepping? Or is that still in the pipeline? Did you get yours as an advance sample or something?
BTW, isn't the retail version of the chip supposed the B2 stepping? Or is that still in the pipeline? Did you get yours as an advance sample or something?
Great work there Mofo, nice to see TicTacs method works on changing the NB. 
Keep it going and keep it real.
It is a shame there are no motherboards ready for Phenom yet, without a good board with a good working bios it will be a while before we all find out how well these chips clock. I am certain they wont match a Q2D in their present stepping but the B3s should be a major improvement.
I do hope everyone realizes had it not been for AMD and its Athlon 64 range none of us would have had a chance to own a C2D or a Q2D, Intel sat on their laurels with the P4 and netburst, and without AMD we would all still be using P4s.
sculptor.
Keep it going and keep it real.
It is a shame there are no motherboards ready for Phenom yet, without a good board with a good working bios it will be a while before we all find out how well these chips clock. I am certain they wont match a Q2D in their present stepping but the B3s should be a major improvement.
I do hope everyone realizes had it not been for AMD and its Athlon 64 range none of us would have had a chance to own a C2D or a Q2D, Intel sat on their laurels with the P4 and netburst, and without AMD we would all still be using P4s.
sculptor.
How though? I can understand comparing CPU performance within the same product line with SuperPi but it clearly doesn't correlate when comparing Core 2 to Athlon 64.
Core 2 Duo E6550: ~21seconds
Athlon 64 X2 6000+: ~28seconds
depending on memory used, that's approximately a 7 second difference. A 33% time difference, which almost certainly doesn't correlate to a 33% performance difference in games, video encoding and other benchmarks. These two CPU's perform the same otherwise. The unified L2 and other factors of the Core architecture make the CPU look far better than it is in general usage. You also need to bare in mind that SuperPi only loads one core, so you can't really bring multicore capabilities into it.
I'm not saying that Core 2 isn't as good as people say it is, that's utter rubbish. They're faster for sure. All I'm saying is, SuperPi isn't exactly the most appropriate way to compare these CPU's.
Apologies, I had no intention of derailing the thread. I almost certainly appreciate what Mofo64 is doing and am looking forward to further progress. Just trying to clear possible misconceptions about SuperPi.
Associate
- Joined
- 12 Jan 2003
- Posts
- 2,045
That is exactly why Super Pi scores mean absolutely nothing in regards to overall chip performance.
His chip would annihilate your Northwood in every way.
Good effort with the overclock. AMD need to sort out some better steppings soon for the enthusiast crowd.
Permabanned
- Joined
- 29 Feb 2004
- Posts
- 1,922
- Location
- fife, Scotland
Could this mediocre SuperPi score be attributed to the supposed cache problem in current Phenom chips?
Anyways, run more tests.
What cache problem? It simply doesn't have enough cache to keep up with a conroe or a kentsfield. If they gave each core 2MB L2 I can guarantee you it would be a good 33% quicker minimum.
Could this mediocre SuperPi score be attributed to the supposed cache problem in current Phenom chips?
Anyways, run more tests.
You mean this cache problem
http://www.theinquirer.net/gb/inquirer/news/2007/11/30/phenoms-feature-infamous-l3
It could be true.
Yer also AMD have basically filled up the whole die with the Phenom core. There iis no more room to add any more L3 Cache.
Associate
- Joined
- 26 Sep 2006
- Posts
- 1,945
- Location
- West Bridgford
Soldato
- Joined
- 15 Oct 2003
- Posts
- 16,098
- Location
- Chengdu
lol @ some people, it's like they buy a cpu purely based on superpi.
Q: Hey, fancy playing some games on your new high end rig? Or how about something useful?
A: *SCOFF* You fool, I have to run 1m through 32m, all day! Don't you see the importance??
We've seen how SP varies wildly depending on memory speeds/timings and cache on AMD systems. Still, it's THE MOST IMPORTANT BENCHMARK IN THE WORLD!!!!
Q: Hey, fancy playing some games on your new high end rig? Or how about something useful?
A: *SCOFF* You fool, I have to run 1m through 32m, all day! Don't you see the importance??
We've seen how SP varies wildly depending on memory speeds/timings and cache on AMD systems. Still, it's THE MOST IMPORTANT BENCHMARK IN THE WORLD!!!!
Are you people complete idiots?
NO ONE here is STUPID enough to JUST run super Pi on a chip and then say its good/bad based on that ALONE...
But super Pi does give an indication of how fast a CPU can do certain ops and is a good indication of how well its ALU works... concerns of CPU cheating aside, any one who has done much CPU benchmarking will understand enough to be able to extrapolate the data against what they know of modern CPUs and use it as an INDICATION of POTENTIAL overall performance... they will also be FULLY aware that a CPU CAN perform in other areas quite differently to what the super Pi result could indicate, but usually its not so stunningly different that the super Pi result is invalid as an indication.
um? it has 4 moderatly faster cores, compared to the 1 older core with HT... thats like comparing a bugatti veyron to a vw golf and saying the golf is gonna get murdered on the 1/4 mile... no one is even going to think otherwise...
NO ONE here is STUPID enough to JUST run super Pi on a chip and then say its good/bad based on that ALONE...
But super Pi does give an indication of how fast a CPU can do certain ops and is a good indication of how well its ALU works... concerns of CPU cheating aside, any one who has done much CPU benchmarking will understand enough to be able to extrapolate the data against what they know of modern CPUs and use it as an INDICATION of POTENTIAL overall performance... they will also be FULLY aware that a CPU CAN perform in other areas quite differently to what the super Pi result could indicate, but usually its not so stunningly different that the super Pi result is invalid as an indication.
um? it has 4 moderatly faster cores, compared to the 1 older core with HT... thats like comparing a bugatti veyron to a vw golf and saying the golf is gonna get murdered on the 1/4 mile... no one is even going to think otherwise...
...but according to you the performance difference shouldn't be that massive as in superpi it is "only moderately better than my aging P4C Northwood at 3gig"
lol...
no I never said that... I said it doesn't show as big gains over my ageing P4 as you'd expect for new technology compared to the core 2 which as it happens is no longer new technology...
What would be interesting is to see the geekbench results for a P4, Phenom and core 2 all at 3gig but with the geekbench affinity set to just one core.
no I never said that... I said it doesn't show as big gains over my ageing P4 as you'd expect for new technology compared to the core 2 which as it happens is no longer new technology...
What would be interesting is to see the geekbench results for a P4, Phenom and core 2 all at 3gig but with the geekbench affinity set to just one core.
Associate
- Joined
- 28 Dec 2006
- Posts
- 587
- Location
- Epsom Downs
Can i ask if that 3ghz is stable ???
Cause i know that the review i read they couldn't get it stable at that speed.
Also it's all about working smarter not fast in CPUs now... i think this is where AMD and fallen over
Cause i know that the review i read they couldn't get it stable at that speed.
Also it's all about working smarter not fast in CPUs now... i think this is where AMD and fallen over
Associate
- Joined
- 24 Dec 2004
- Posts
- 458
I am afraid that Mofo64, the PO has disappeared. I hope he will get back to let us know how it is going
He's probably off announcing something to himself in the third person
"After two hours of straining, MOFO64 is proud to announce that he has been to the toilet, and he has installed the resulting poo for a better overclock than the Phenom"
Oddly enough - before Core2Duo I seem to recall that AMD Athlon 64's wiped the floor with Intel CPUs at SuperPi, the only things Intels were good for was video encoding. SuperPi is just one benchmark, and I'm sure there will be benchmarks where the Phenom shines, but overall, this isn't going to change anyone's buying decisions.
It's obvious though, is that the current AMD motherboards being used by MOFO64 and Jokester aren't that great for clocking, so maybe we need to see a comparison with decent motherboards next?
"After two hours of straining, MOFO64 is proud to announce that he has been to the toilet, and he has installed the resulting poo for a better overclock than the Phenom"
Oddly enough - before Core2Duo I seem to recall that AMD Athlon 64's wiped the floor with Intel CPUs at SuperPi, the only things Intels were good for was video encoding. SuperPi is just one benchmark, and I'm sure there will be benchmarks where the Phenom shines, but overall, this isn't going to change anyone's buying decisions.
It's obvious though, is that the current AMD motherboards being used by MOFO64 and Jokester aren't that great for clocking, so maybe we need to see a comparison with decent motherboards next?