http://img171.imageshack.us/img171/9...roadmapxf9.jpg
http://translate.google.com/translat...hl=en&ie=UTF-8
Printable View
I mean, that's great, but do we really need a new product for a 100mhz speed boost? They couldn't have gone to 3.2ghz? Considering the 940 is really a sub-80W part, I imagine 3.2ghz should be possible using the 125W TDP bracket.
100mhz doesn't seem enough to be happy about, was looking for a 200/333mhz boost to cheer about.
Already posted -----> http://www.xtremesystems.org/forums/...d.php?t=214860
The new 710-720 x3 cores, Are they defective deneb quads or are they actual tri cores, And why 7.5mb cache?
What does this say about deneb yields? Did it not take over a year to get enough defective agenas to launch x3.
tri cores = fail
less wasted silicon more for profit.
Tri cores get higher IPC since they have more l3 cache to use for all three cores, same with dual core ones with l3 cache.
I hate when companies tech is lacking that they have to keep increasing clock speed to gain numbers. Its like where AMD was a year or two ago. They should just bust out a 4Ghz chip and save themselves the extra time and silicon wasted.
There's no way they could launch a 4GHz chip. You play with a Phenom II yet? Most require 1.5vCPU or greater and the TDP shoots through the damn roof. I liken Phenom II to Netburst, it scales well but needs cooling.
Why are you guys against tri-cores?
AMD knows that even against wolfdale clock for clock they can't do much. So they throw you a bone and you get an extra core.
And with most games only being dual core optimized, that leaves you a free core to deal with the OS/background processes.
Seems decent enough to me. As long as the price is comparable to wolfdale, I think AMD board owners will enjoy the tri-cores.
Can anyone disable 1 core in Ph2 920 and rise HT to 2.2Ghz?
That should be the tricore.
Amazing. Tick, tick, tick, tock.
I was wondering the same...
Either they are committing good quad-cores to sell as tri (stupid idea) or yields are worse than on 65nm (most likely). This or other way consumer wins, because tri-cores will be cheaper and less energy hungry. Perfect for HTPC/gaming machines, not so great for workstation/servers.
On the other thought 1st quarter volume of Agena/Barcelona was definitely MUCH lower than it is on newer 45nm stuff. Looking from this angle is hard to say anything about yield of AMD 45nm process.
3.1ghz is completely unnecessary? 3.2 or 3.266ghz would be good. Much of the point of the top end processor from both INTEL and AMD is that they have an unlock multiplier, so the actual frequency portion is not as important anymore. That being said, you need a bigger megahertz difference to make that part relevant.
Tricores are not a bad idea actually. They help AMD diversify the deneb lineup a bit and lets AMD discontinue obsolete processors. AMD needs to discontinue the athlon x2 altogether. Its just a really old architecture at this point and will help AMD image.
Ph2 core is smaller. yields must be way better otherwise there weren't so much Ph2 in the market.
And in 1 month 6 new cores come to market in volume. yields can not be a problem with actual roadmap.
Btw these new tri core are way better. more cache, more clocks, higher ipc, better overclocking. Now these tri-cores can make the hole difference. The Ph1 tri-core was very weak.
What's with this stock 333 & 266 jumps.. some of you either have never owned AMD CPU or have forgotten. :)
i was wanting no clock increase i want an am3 BE as cheap as i can get it at launch of a good amd board