• 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.

Phenom x4 Agena 9600: Sandra XII Bench

AMD is planning on moving them to 45nm process next year, so it might be worth waiting some more, preliminary will still be interesting though.
 
i keep feeling that this is like the "wait for the 2900's to come out" saga that affected the release of the ATi HD2900 gfx cards...they did finally come out and were crap compared to the GTX...i really hope AMD pull sommin out the bag. a hefty overclocker would be good but the speculation is already there...

wait and see i guess. here's hoping, come on AMD!!
 
J0hnb0y said:
i keep feeling that this is like the "wait for the 2900's to come out" saga that affected the release of the ATi HD2900 gfx cards...they did finally come out and were crap compared to the GTX...i really hope AMD pull sommin out the bag. a hefty overclocker would be good but the speculation is already there...

wait and see i guess. here's hoping, come on AMD!!
Click to expand...

I think AMD's main problem atm is that they just can't manufacture the chips as well as intel. I believe AMD's top chip is the 6000+ which is still on 90nm. That would seem to suggest that the 65nm chips aren't working out too well.

We could look at this both positively or negatively. On one hand when AMD move to 45nm they might see a huge jump. OTOH if they are struggling with 65nm they might be in even more trouble when they get to 45nm and 32nm. Its been well documented for years and years that getting to 45nm and below is when it would start getting difficult to manufacture chips- intel seem to have made the break through, now it is up to AMD, and IBM, to follow suit.
 
Owenb said:
I think AMD's main problem atm is that they just can't manufacture the chips as well as intel. I believe AMD's top chip is the 6000+ which is still on 90nm. That would seem to suggest that the 65nm chips aren't working out too well.

We could look at this both positively or negatively. On one hand when AMD move to 45nm they might see a huge jump. OTOH if they are struggling with 65nm they might be in even more trouble when they get to 45nm and 32nm. Its been well documented for years and years that getting to 45nm and below is when it would start getting difficult to manufacture chips- intel seem to have made the break through, now it is up to AMD, and IBM, to follow suit.
Click to expand...

I see it as a positive, AMD has always maxed out the current nm process, look at what they done with 90nm, you think Intel could match that? I dont think so, but they had to move to a smaller process, I believe that the only way intel can compete with amd is to jump to a smaller nm process every time, their already on 45nm, perhaps their worried about AMD's potential 65nm process?

Remember you can only shrink to a certain size nm before things become impossible. One day Intel may not be able to jump the the next lowest nm process?
 
Hi there

I've got a Phenom 9600 and am just waiting for one of the 790FX board to put it in along with 4x HD 3870's. :D

Any early reports on which 790FX is the best clocker?
 
From what I've seen the Gigabyte board does over 400MHz, so FSB overclocking isn't an issue with the 790FX board. I would suspect that it has more than enough Vcore options as well.

Jokester
 
Kinitawawi said:
I see it as a positive, AMD has always maxed out the current nm process, look at what they done with 90nm, you think Intel could match that? I dont think so, but they had to move to a smaller process, I believe that the only way intel can compete with amd is to jump to a smaller nm process every time, their already on 45nm, perhaps their worried about AMD's potential 65nm process?

Remember you can only shrink to a certain size nm before things become impossible. One day Intel may not be able to jump the the next lowest nm process?
Click to expand...

Why wait when you already have a better process online.... Switching to a smaller process isnt just about performance and features... Profits factor into it in a big way... How many chips per square inch of wafer.. more chips more profits. 65nm has been very profitable to intel, and the financial boys will want to get into 45nm as soon as possible.

When the process development reaches impossible levels, chip makers will either have to fine a completely new technology, or make do with what there is, and accept that chips will become physically larger and learn new ways to control the heat output etc.
 
Gibbo said:
Hi there

I've got a Phenom 9600 and am just waiting for one of the 790FX board to put it in along with 4x HD 3870's. :D

Any early reports on which 790FX is the best clocker?
Click to expand...


the one with the biggest box? or stupidest name, or worst colour scheme is often a good indicator :p

really just commenting on the if they could run at 3Ghz i think they'd sell one thing. thats just, well, tosh, thats never once been the case in the , well, history of all things cpu in the last 5 years at least.

c2d can do 4Ghz, 3.4Ghz with easy, i've not had one that couldn't do 3.6Ghz, but they don't sell them. most of the ath xp's i had would do 2.6Ghz or more, yet they didn't sell them. my quad cores can do 3.6Ghz and 4Ghz, yet they don't sell them. also, amd ARE planning to sell a 3Ghz quad core part within a few months, so they can hit that speed, and are planning to sell it, which to me says they can go significantly higher than that overclocked.

simple fact is, that for 99% of home users, any quad core is overkill, there isn't a game out a 2.4Ghz quad can't handle, there isn't really a game a 2.4Ghz dual core can't handle, and that includes x2's. crysis shows a noticeable increase from 1.66Ghz up to 2.13Ghz, but even thats not massive, and thats dual core, 2.13 up to 3Ghz made barely a dent, and quad core was essentially no faster than dual core. a phenom, kentsfield, penryn doesn't honestly matter unless you do encoding, do intensive 3d rendering, for gaming. if you could run both machines side by side with 100 people trying to decide which is better, honestly, no one could tell the difference. you'd be hard pushed to notice the difference even in sup commander between a 2.4Ghz dual and a 4Ghz quad.

what i have in general liked is amd's mobo price, while the 865 best boards were £100 plus, and the 875's were even more amd's nice like nf4 you could get a fully featured board for £65. If overall the cheaper package is amd, it will be a great option. plus who doesn't want to try overclocking 4 cores separately :p

while i think the whole, turning cores off/massively down for downloading overnight or a ultra silent setup for playing back a film, i doubt i'd do it more than once or twice unless they make the overdrive app so streamlined that you could change between profiles you choose in 2 clicks.
 
Corasik said:
Why wait when you already have a better process online.... Switching to a smaller process isnt just about performance and features... Profits factor into it in a big way... How many chips per square inch of wafer.. more chips more profits. 65nm has been very profitable to intel, and the financial boys will want to get into 45nm as soon as possible.

When the process development reaches impossible levels, chip makers will either have to fine a completely new technology, or make do with what there is, and accept that chips will become physically larger and learn new ways to control the heat output etc.
Click to expand...

i think we could potentially see the use of slots for small pcb's with cpu's on instead of the current things at some point. because eventually we will slow down the nm reduction and increase die numbers. so you want a a way to fit a bigger cpu on the mobo, think either a low riser 2400pro card compared to a 8800ultra size. more cpu power, bigger card, same case/mobo size and more latitude for variation in cpu package size. tbh heat isn't that huge an issue. 99.9% of us could live with a bigger case. the bigger problem isn't heatsink size/dissapation capacity, but the lack of die size. newer chips on a newer process can do the same amount of heat for a lower amount of power used, but that turns to heat, and that ends up being transfered over a smaller die surface so the problem has been keeping the delta in temp as high as possible with bigger sinks. but slightly lower clocked and hotter chips are fine too, but heatsinks get massively more efficient the more heat they have on them, because the temp difference between heatsink and air gets much bigger. but when we've got 128 cores and nothing to use them for, who cares :p

we are getting to a difficult point where, more cpu is nice but, we've got not that much to do with it. when you look at how long crysis took to get out, and how much time went into the design, every single time we increase the complexity and how interactive the games are it takes more time. there is just a limit to how many people you can have working on a game, and how long you really have to get a game out. we've got cpu hardware, but not much to use it on right now. but things aren't looking to change massively. the best engines already take 4-5 years between new versions, we're at a limit of coding complexity and realistic design idea's. i can see this set of quads lasting a heck of a lot longer than you'd think when it comes to how long they aren't utilised much.
 
one problem here and thats the fact that all of those cpu in the Sandra bench are all at their stock clocks, which in theory means that if other cpu's beat at stock it then you might as well get the better cpu's and overclock them. the only way i feel this will be worth it is if its less then £150 which is un-doubtably not going to happen from AMD. tbh i think you'd be better off with a nice and cheap, bog standart Q6600 G0.
 
I don't really see it as a problem, the AMD is also at stock clock and has an overclocking potential (wish we knew what it was though...) That is why I don't understand why the Q6600 isn't there (not a selectable cpu on benchmark?), it would be a far better performance comparison at stock speeds.
Once the Phenoms are released on boards that were made for them I hope to see some decent comparisons, I am itching to buy but want the best VFM!
 
YEs, I have to agree, this is a very unsatisfying benchmark....

Would really like to see how it performs once they are out (overclockign mainly)
 
Kinitawawi said:
Remember you can only shrink to a certain size nm before things become impossible. One day Intel may not be able to jump the the next lowest nm process?
Click to expand...


I don't see how that's relevant. Intel doesn't have to move to 45nm, it's expensive, difficult and is untested. The fact is that they can though, AMD can't atm and the Phenom's are hard to manufacture because they're native quads. Intel's natives will be out end of 2008 in the form of Nehalem.

No-one cares if AMD can maximise 90nm/65nm performance. Big deal, it doesn't matter how something works, it matters how much performance it gives. Intel's already sold over a million quad core chips; it's a business obviously, so they've won anyway.
 
You must log in or register to reply here.
Back
Top Bottom