AMD FX-Series Lineup Revealed in CPU Support List

[alyon]

Arent the bulldozer chips have better single threaded performance solely because when they run one core it is using the all the shared resources in the module while the other core idles? In a multithreaded situation the performance per core would be much lower due to having the share the resources?

 

[CaedenV]

[citation][nom]Dyslogik[/nom]Wait a minute... this article talks about 4-8 cores and not 4-8 modules which is a more suitable term for the Bulldozer architecture...So my question is: since each module has two cores (2xcores) does that means the FX-8150 has 4 modules? I heard that each module is almost as faster as a normal dual-core because of the 2x int schedulers, if so this means that only something like a 6-core i7 processor can be on pair with this or close to this. Since the rumors says the FX-8150 is 50% faster than the Core i7 950 I wouldn't doubt it looking at this architecture. But like everyone has said, we need benchmarks to be completely sure.[/citation]
Each module has 2 cores, the lineup is 2 modules for the crap proc, 3 modules for the middle ground, and then 4 modules and increasing speeds for the high end. As most of their proc are identical most people will buy the bottom line procs and unlock modules, and overclock to save money. Unless they join Intel's game of locking down the overclocking for all but the top processors of each family (but this is doubtful as AMD is awesome that way). As to if each module will be faster than one of intel's core+hyperthread clock-per-clock it is likely, but intel will be able to hit much higher clocks at lower temps. Still we wont know for sure until we see benchmarks, which will be all sorts of fun!
[citation][nom]LORD_ORION[/nom]Evolution huh? It's not like we haven't had organisms stuck in various layers of sediment as time went on to verify evolution. Please pick up a layman scientific book on some topic you are interested in and read.[/citation]
Both evolution and creationism/intelligent design have huge monstrous holes in them. To 'scientifically' believe either theory is a terrible mistake. To believe one or the other on 'faith' or because it fits your particular philosophy is a different matter, but still completely off topic, and should not be on a site where intel and AMD fan-boys should be practicing their religion of bashing one another (in the name of love of course )

 

[wiyosaya]

As others have expressed, I am getting tired of the hype, and want parts. With any luck, BD will live up to the hype. If not, then the absolute best AMD can hope for is to continue to appeal to the "value" market, and I expect there will be a lot of people out there, myself included, who will be disappointed. That said, I will almost certainly not switch to Intel as their parts are overpriced, IMHO.

One consequence that might be expected if BD does outperform Intel's best parts is that AMD may once again become the pricy part. (Remember the days when AMD had no competition in Intel's "Core" parts?) If that happens, I might not buy into the coming generation of parts; rather, I might just wait to upgrade until BD's successor hits the market and buy the BD generation of parts.

Also, I really hope that AMD has all the embarrassing hardware bugs worked out. If the delay in releasing the parts has resolved the hardware bugs before the part hits the market, then Kudos to AMD. If not, well, I suppose AMD will have to sport another black eye.

 

[noblerabbit]

going to replace my AMD Tricore with a Bulldozer CPU and upcoming raden 7xxx just in time for Diablo 3.

 

[noblerabbit]

[citation][nom]cypeq[/nom]And the source for that statement is ?[/citation]

Allow me to source: Intel is going down **puts on sunglasses* the street to deposit 15 Billion dollars of hard cash this Q4

 

[silverblue]

wiyosaya

All CPU designs have bugs (and plenty of them), some serious, some not. The TLB bug was embarrassing, sure, but a complete non-issue for the vast majority of people. You'd expect that the endless tinkering with the Bulldozer design will have reduced the likelihood of something scary happening but until the chip is in the wild, we won't know for sure (unless AMD finds something and admits to it pre-release).

 

[Guest]

@caedenv

actually if i understand the architecture correctly (and i maybe wrong seeing as im not a hardware engineer) overclocking will only be required by the die hards and LN nuts

the chip is designed to maximize thermal headroom, if at any time a module is under utilize it will turn that module off and use the extra thermal headroom to overclock the remaining modules, from what i understand 3.6 is just the base frequency for all modules, the system will decide the actually speed of each module based upon workload balancing. As for module unlocking, this is a much more complicated beast then normal cores, each module consist of 2 cores (for lack of better word to describe them) a module may have been disabled (and downgraded to a lower module count chip) due to a defect in one or two of the cores, this significantly decreases the chances of a successful unlock, though there are rumors that AMD is working on a platform that allows the use of half a module (i.e only utilizing a single core of a module) which begs the question would a 4 half module chip out perform a 2 module chip or even a 3 module chip

for enthusiast who like to fiddle with stuff i think bulldozer will be a dream, there are a multitudes of configurations that may yield better results for different applications even more so when the GPGPU gets thrown into the equation (id imagine a 2 module 2 half module chip with a decent enough GPGPU would be entertaining to play with, where by the two modules are used for normal computing and the two half modules are used to load up the GPGPU, id imagine gaming might also profit form that configuration)

k id admit im really pumped by the idea of bulldozer now, please for the love of all that is good dont be a disappointment....

 

[JamesSneed]

I know what Ivy bridge is. I was saying that the E class chips can be used in socket 2011 and 1155 I believe. Just like the older i7s could be placed in 1156 or 1366.

Except you are incorrect. The 2011 in socket 2011 is the number of pin wholes in the socket. So 2011 pins do not fit into a socket with 1155 pin wholes. You are also wrong the older i7's could not be placed into 1156 or 1366.

This chart will show you which CPU series fit in which socket: http://en.wikipedia.org/wiki/CPU_socket

Please stop saying this as someone that has no clue may actually believe it.

 

[captaincharisma]

FX series? so they going to have NVIDIA chips integrated into AMD CPU's now

 

[sonofliberty08]

i have a good feelings about this

 

[someonewhoknowsalittle]

Do AMD best new CPUs let me play Solitaire or Hearts and let me email or send some newfangled thing called a chat or instant message? Should I trust AMD?I watch Sally on TV. She no talk about AMD. Never saw a commercial, an ad or anything that said AMD. Is AMD as fast as a grapefruit, a snail, a Lada, a two toed toad or is AMD's CEO a fruit? Not that there's anything wrong with that. Predator? My dentures have more bite and my anal sphincter provides more gas than CEO of the month Rory.

 

[Guest]

Sounds good but will the 990FX chipset support PCI-E 3.0?

 

[ern88]

I WANT BENCHMARK LEAKS NOW 😀

 

[clonazepam]

I just want to predict that these processors will surpass the clock for clock performance of any current Intel processor... in any software bench or application that isn't or hasn't been optimized for intel cpu's. I think these chips will solidly outperform its intel counterpart in many softwares that are open license or freeware. Just like Intel, some of the line-up will be complete overkill for gaming. ~Cheers - jus' having fun because I'm excited, been waiting a long time for these.

 

[silverblue]

The 8120 is a C0 stepping processor from what I've heard. I'm wondering how Gigabyte have tested one or two processors not slated to appear until Q1 2012...

 

[tajisi]

It's rather sad that this information is the most revealed in weeks (despite supposedly being ready for launch soon) and it comes from a third party.

AMD, numbers. Please. Stop being afraid over the first Phenom fiasco. Don't overhype things and you'll be fine. If it sucks, you'll still have plenty of diehards buy it regardless just due to the logo on it. If it's great you have nothing to lose since people are tired of waiting and going with Intel. Just let us known one way or the other. Thanks.

 

[msgun98]

If AMD's processors are 85% as fast as the intel Sandy Bridge counterparts but are much cheaper, I'll probably go AMD.

 

[killerb255]

I'm usually not too picky, but..."artiture"?

 

[verbalizer]

I'm usually not too picky, but..."artiture"?

you need to use spell check..

 

[dickcheney]

[citation][nom]Dyslogik[/nom]Wait a minute... this article talks about 4-8 cores and not 4-8 modules which is a more suitable term for the Bulldozer architecture...So my question is: since each module has two cores (2xcores) does that means the FX-8150 has 4 modules? I heard that each module is almost as faster as a normal dual-core because of the 2x int schedulers, if so this means that only something like a 6-core i7 processor can be on pair with this or close to this. Since the rumors says the FX-8150 is 50% faster than the Core i7 950 I wouldn't doubt it looking at this architecture. But like everyone has said, we need benchmarks to be completely sure.[/citation]

Nice to see that the pipe dreaming is still going strong.

 

[Nnymrod]

My Crosshair V is waiting...

 

[verbalizer]

getting this constant notification on thread...
hope this will fix it.

 

[dragonsqrrl]

[citation][nom]msgun98[/nom]If AMD's processors are 85% as fast as the intel Sandy Bridge counterparts but are much cheaper, I'll probably go AMD.[/citation]
sigh... I know a lot of AMD fanboys like to reassure themselves and their fellow AMD diehards of their tunnel vision buying decisions by constantly repeating this to themselves, but in many benchmarks this is completely untrue, especially in productivity and content creation.

If you're fine with the performance you're getting with a Phenom II, then that's great. But please don't try to deceive yourself and others with these cookie cutter figures. In many cases there is far more then a 15% performance gap between an i7-2600k (or i7-980) and a Phenom II X6. This, and the argument that there's never a need for more processing power then what AMD products currently offer are getting really old and need to stop.

 

[silverblue]

Funny, because "just enough" seems to be a rather new idea. In any case, the biggest wins for Intel seem to be in synthetic benchmarks, and until SB came out, was there really any point in buying a 6-core Intel over AMD's Thuban? The price difference was shocking... it still is.

 

[Guest]

Lmao at "someone needs to leave a FX-8150 at a bar". I'm sure that iphone5 that was left was traced back to the employee that had it. But I'm still hoping an AMD employee goes out to celebrate the BD launch with his buddies, gets wasted, (and just so happened to have a finished FX-8150 with him to show his friends) and leaves that little gem for one of us to find!