That's also only a few tests out of dozens that should be performed. Remember it's very easy to skew results. Notice it puts up equal performance on some tests.
Shaff, I never said that it was a bad review. It was an inadequate review that was too narrow in scope, that was biased in Intel's favor! Russ
Sam, Is there an article that goes with the information you posted? If so, can you provide a link? Russ
I find this quite entertaining. Nobody really knows the performance of Bulldozer, and we're arguing about it LOL! Patience men, patience My A7M266 Asus board will run a Mobile athlon thoroughbred. Its bios does not recognize it. But it runs. If memory serves, it believed it was running at 800Mhz. And yet it was better than the 1.4 Thunderbird it replaced. In my opinion, we cannot know the actual performance of bulldozer, until all the kinks are worked out, and boards FULLY support the processor
Nobody has argued about Bulldozer at all yet, this all stemmed from previous articles comparing the Phenom II and Core i5/i7. I agree that this test is limited, and you've even overlooked the biggest flaw with it - it's before the NDA is lifted! The website is printed in the corner of the graphs - it's the front page article. Still though, if there is even a grain of truth to these tests, that's scary. The i7 2600K decimates the FX-8150 in some of those tests, and that's the flagship chip.
What is odd to me and I guess I should dig up the site I found what seemed to be good number is on that site they showed Handbrake numbers that showed around 50% performance increase between the Phenom II and the 2600. But as I recall they were looking at processing time and not fps. Like Russ says it would be nice to see their parameters for testing. Kevin, that was one of the marketing tricks they have been doing for years. Some programs/games are optimized for flaws in Intel processors which means they will not perform as well on AMD instruction sets as well. AMD has also done the same thing too.
I'll have to agree with both parties that the graphs shown could easily have been fabricated or manipulated in some way. Also the NDA is still in effect, so if those graphs are legit, someone's in trouble. Also, the rest of the review on that site paints it in a slightly more favorable light. From what I can see it largely competes well except in memory bandwidth and number crunching...
Granted, a fair few of the tests are close to equal. But they shouldn't be, really. Remember this is an 8-core CPU going up against a 4-core CPU. For it to perform equally at best, and far slower at worst, is a serious problem. I'm hoping that the benches are false, and that Bulldozer performs better than that, but the MSRPs caused me to question my expectations of the architecture, and this hasn't helped, spurious as it may be.
I hope they're false as well. But, as long as six of its cores are equal or close to equal to 1090t, I'll be content. Because I'm sure 8150 will smoke my current quad, in X264 encoding. They are expected to release a revised bulldozer shortly after eh? Much like phenom I gave rise to phenom II.
And you're talking stock frequencies. Correct? And since it has proven to be a good overclocker, I'm sure I'll be impressed. If it can hit 4.5Ghz, I'll probably be happy. But I read a while back, that 5 should be reachable on air? Of course I prefer water now
It hasn't proven anything. They got the Phenom IIs to 7Ghz on liquid nitrogen, but they don't overclock very well compared to Sandy Bridge CPUs, which I think only got as far as 6Ghz. There's no guarantee that joe public will be able to overclock Bulldozer to any higher clocks than Phenom 2.
It seems like it was a statement by AMD themselves. Not sure where I read it now :S I realize that just because they broke the overclock record, doesn't mean all the bulldozer chips will be capable of extreme frequencies(but it sure does look like a more stable CPU). I actually read that 5 was was an achievable clock. I guess time will tell.
Definitely the Intels overclock better than the AMDs right now. and they should be stock testing although sometimes they do OC tests, but they'll still do factory tests too. If AMD is saying 10% increase that isn't very much and is still short of the Sandy, the way AMD was talking a while back I thought for sure they were going to have a contender if not exceed Intels products.
Perhaps they're biding their time, to see what intel is going to do :S Wow, I forgot how hot the earlier Athlons ran. I replaced 1.4 thunderbird with a Palomino 1.67ghz a while back. Just turned it on, to see how the ol girl is doing. 51C at idle! Apparently it can handle fairly high load temps though. I think I'm more nervous about the northbridge though. Very hot to the touch. probably pushing 120 - 130 F. I'm planning on disassembling all the parts and listing them on ebay. I ran a puppy live cd, it seems to be running good. Sure wish Asus made die hard boards like that still...
Geese! Apparently LOL! Well, the duct isn't currently routed to pull heat away from the cpu. So I'd probably help it a little, by placing the plastic duct back on there Pretty old though. Probably not too many people interested in old technology :S I do have 2Gb worth of PC2100. That'd probably be scooped up.
I think the cooler I bought for it was just crap really, but still, that's mega hot. Talking of mega hot, really need to re-seat my i5's cooler. Had to turn the overclock off to stop it from overheating!
Yah, I probably need to re-seat mine as well. It really should be running better than it is. I'm almost certain I applied too much compound. From now on, I'm gonna seriously inspect the heatsink for imperfections. E.g. warps. If there are, lapping may be a good idea. But since my upgrade isn't too far away, I'm not too concerned.