LOL!!! How am I threatened?
What kind of tenuous grasp on reality do you have?
Is there something particularly special about the AMD section that I need to know about?
One thing I have noticed though in this section is a number of people holding quite irrationally optimistic views on Deneb's likely prospects.
You do realize that AMD as said and yes even Dirk, that there ahve been a lot of improvements along with this die shrink. It will be better the 65nm. They said on the record that it has better then expected improvements ion Power Consumtion, and other tweaks. I don't expect it to beat i7 at 8 threads LOL.. but it will be a faster chip then current Phenom's. Which is still better...it is not going to be the same or worse. and it isn't to be priced high. And it is the first 3.0ghz quad from AMD, even 3.0ghz 65nm chips compete comparable in a lot of cases, if this is faster, then it is still better and a desired change. I have AM2+ boards all over my house..so it is a direct fit, no need to spend more then a chip upgrade. Did i ramble off pace here...LOL I am tired and hungry and work is driving me crazy....
~1~
AMD Ryzen 9 3900X
GigaByte X570 AORUS LITE
Trident-Z 3200 CL14 16GB
AMD Radeon VII
~2~
AMD Ryzen ThreadRipper 2950x
Asus Prime X399-A
GSkill Flare-X 3200mhz, CAS14, 64GB
AMD RX 5700 XT
AMD's credibility is in tatters in respect of at least their CPU's, so I have relatively low expectations for them, hopefully they can restore my faith, I'm just not counting on it.
Sure Deneb will be better than Phenom, but I very much doubt by the 15 to 20% margins many are claiming, except in niche applications.
You are right, the memory controller is no longer a ring based controller its point2point how most memory controllers work. Not to mention its TMU increase and SP tweaks and increases.
The R700 is a vastly improved R600 major changes being what I mentioned and some other low level tweaks in there that I cant remember off top of my head.
wasn't r700 a redsign of shaders ? (because of poor AA and AF on both 2900 Xt and 3870)
http://images.bit-tech.net/content_i...rv670-flow.jpg (3870)
http://prohardver.hu/dl/rev/2008-06/...v770_block.jpg (4870)
up util this point, i was sort of going along with what you were saying, this statement however is nonsense.
here is the math,
2.8 gig = current max offiical clock from AMD
3.0 - 3.2 gig = estimated speed from Deneb
= 15% to 20% increase.
considering most Phenoms will hit 3 gig now, this doesnt seem to 'out there' and some of the lucky guys like Hondaman and charged, are hitting 3.5 gig on water with current Phenoms, then.....
it isnt that much of a leap of Imagination to envisage that the new chips will hit 4 gig? is it?
for me a quad core AMD chip at 4 gig is something to shout about.
and as far as i am concerned a die shrink = new manufactueing process = new architecture, even if this resembles the same architecture as before.
its a differant chip, and no amount of 'its rubbish because' will persaude me otherwise, maybe they will listen to you over at the Intel Forums.
not here
I am talking about IPC, I guess I should have been clearer, but I would hardly be shocked if a 3 Ghz Deneb isn't 20% faster than a 2.6 Ghz Phenom.
Isn't 2.6Ghz the fastest available stock speed?here is the math,
2.8 gig = current max offiical clock from AMD
I suspect hitting 4 gig will be notably less common than it is for Penryn Quads.it isnt that much of a leap of Imagination to envisage that the new chips will hit 4 gig? is it?
That is quite a perverse interpretation you got going for yourself there.and as far as i am concerned a die shrink = new manufactueing process = new architecture, even if this resembles the same architecture as before.
Bulldozer will be a new architecture, not Deneb.
The "it's only a die shrink" people remind me of the movie "Life of Brian" about the Romans.
"Deneb is just a shrink. Well okay they did add some cache. Oh yeah and they also increased it to 48-way instead of 32-way. And yes they also probably optimized some of the routines in the processor. And changed cool and quiet for efficiency. Perhaps they tweaked the HT. And perhaps they worked on things we don't know about. But other than those minor things it's just a shrink so what can we expect."
FX-8350, Powercolor ATI R9 290X LCS, OCZ Vertex 4, Crosshair V Forumula-Z, AMD Radeon DDR3-2133 2x8Gb, Corsair HX1000W, Thermaltake Xaser VI, Xonar D2X, Water Cooling 140.3
ok, if you cant get your head round it, here is the 12 year old version,and as far as i am concerned a die shrink = new manufactueing process = new architecture, even if this resembles the same architecture as before.
something made smaller than previous example = new method to make it smaller = a shrinking of previous method, or completely new method.
now even if it looks the same, (Just smaller) doesnt mean this is the same architecture, changes will have to have been made to facilitate the shrinking process
even just shrinking a wire, means this is an architectural change.
plain enough for you?
FX-8350, Powercolor ATI R9 290X LCS, OCZ Vertex 4, Crosshair V Forumula-Z, AMD Radeon DDR3-2133 2x8Gb, Corsair HX1000W, Thermaltake Xaser VI, Xonar D2X, Water Cooling 140.3
Chad no hard feelings here really, you dont seem confrontational at all, hence why i still reply to you, and your not on my ignore list.
but,
for someone that quotes themself to be a 'overclocking newbie' in an intel forum only two months ago, to then come on here and spew IPC performance is sort of grating.
i do truly believe that you have no idea what your talking about now.
so i wish you all the best over at the intel forum, but remind you that you are not the first, and wont be the last intel boy to come over here, and question our logic.
WE are AMD here, and WE know what we are talking about, we niether want to bench against Intel or compare.
i wish you the best for the future, you are now on my ignore list
bye![]()
Bookmarks