I don't know myself. I doubt the lowest timings are being used. . I know the M/B will be optimized to run better. I am sure by the time we see yonah there will be better timings for DDR2 667 . If it is infact running at 667FSB we don't know. Their not showing us. If I was Intel. No I wouldn't show yet. Still Feb before the release. I think you can plan on the no . system from intel X50 to be equal to the x2 4800. Its looking like its going to be very close anyway. We know more today than yesterday which is a step forward.
Why would you possibly expect a better performance from a retail board and CPU than a hand-tuned system Intel is displaying???
Why would a Yonah be an AMD-killer when it performs slower than an X2 3800 (which is far cheaper if current pricing strategy holds)???
Intellia...I don't think this you're Kansas anymore...
Hay i don't care if its yonah or sossaman but we haven't seen all the no. yet it well be a while. Its really looking very good for a cpu running at 30 watts. Now that Intel is backing yonah on the desktop its really going to get interesting. I am guessing but I would guess the yonah was run with DDR2 400 on a 600FSB.
Ya its will max out at maybe 2.26- But that won't stop people from running them @ 3GHz and you got to admit it will be a killer O/C. I would like to see 1 @ 2.4 2.6 2.8 on air . Ya gamers are going to be going after these things.
Things to remember...
1. Sossaman is a server chip that won't even do 64bit...
2. Sossaman uses the Lindenhurst chipset, and that maxes out at 32GB of Ram anyway
3. The low-power dualcore Opterons run at or below the Sossaman's power level, and they don't have these limitations...
4. The Opteron platform has already been validated (usually takes a year after release) while Sossaman has not
1. Sossaman is a server chip that won't even do 64bit...
2. Sossaman uses the Lindenhurst chipset, and that maxes out at 32GB of Ram anyway
3. The low-power dualcore Opterons run at or below the Sossaman's power level, and they don't have these limitations...
4. The Opteron platform has already been validated (usually takes a year after release) while Sossaman has not
What you don't think nvidia isn't going to sli these . come on . Ati will . Dfi will build a gamer board for it . Yes I expect the chipset to get some more tuning its a mute point. The m/b that intel is showing it on now is the supreme M/B lol. And until proved differant i am going with the idea that it was running on ddr2 400 and a 600fsb. I have no reason to believe anything else. Did you notice the * behind the 2.GHz sassaman and the 2GHz yonah . they were both running lineare. Which tells me the 1.5 GHz weren't . so there you have it. Since we know the sassaman was running asyn. and DDR2 400 with a 600FSB we can see that is the same way the yonah was set up .All because of those 2 little *
Ummm...you think DFI will build a gamer board for a mobile chip and a server chip?
you think Nvidia will build an SLI chipset for a server or a laptop?
sigh...
Very well, I will grant you that. We dont have enough information about the systems to draw a final conclusion (we also dont know the specs of the AMD systems, by the way). Yes, yonah will be great for overclockers (if they dont care about x86-64. Personally I wouldnt buy a CPU without it but that's just me) IF desktop motherboards are available at DECENT prices and if the dual-core cpus themselves are not insanely overpriced (which they might be, since intel plans on going forward with single-core versions as well).
EDIT: Now, let me just say this--whenever you demonstrate a new product you want to demonstrate at it's best, always (I'm guessing that's why we werent told at what clockspeed that merom laptop was running, I'd guess they're still working on that). That's just good business sense. You NEVER let people think that your product is inferior than it really is because it's very hard to make them dismiss that first impression later on. Crippling Yonah's performance would make no sense unless there was a problem getting it to run @ FSB667 with DDR2 667, which I dont think there is.
EDIT: Now, let me just say this--whenever you demonstrate a new product you want to demonstrate at it's best, always (I'm guessing that's why we werent told at what clockspeed that merom laptop was running, I'd guess they're still working on that). That's just good business sense. You NEVER let people think that your product is inferior than it really is because it's very hard to make them dismiss that first impression later on. Crippling Yonah's performance would make no sense unless there was a problem getting it to run @ FSB667 with DDR2 667, which I dont think there is.
You can say whatever you want, Intelia, I don't believe you.
But if it's true, if Intel regains the crown, I'll buy that.
But if it's true, if Intel regains the crown, I'll buy that.
Their going to be expensive but i here the yields are great and the fact that their on 65nm. which is suppose to reduce cost we shall wait and see.
Heh, the shrink will reduce cost for Intel, these cost savings are almost never passed onto the consumer (since they do have to pay for all the retooling done to their fabs, the R&D, etc). Same with yields. They'll charge as much as they can charge until they get some competition, or overstock.
Hey stupid-ass!!!
How is yonah an X2 killer when at 2ghz it scores lower in the test versus the 2ghz 512kb cache 3800+??? NO answer...Then STFU!!!!
It is
Oh by the way I score in the 700's now.....
How is yonah an X2 killer when at 2ghz it scores lower in the test versus the 2ghz 512kb cache 3800+??? NO answer...Then STFU!!!!
It is
Oh by the way I score in the 700's now.....
Again we see things differantly . No matter what I am impressed with the performance of this cheap just the way were seeing it . If its at its best I am happy with it . lets not forget it is meant for the note book.
Duvie if you look at the setups they weren't running @ 667 FSB both the sossaman and the Yonah were running lineare as shown by the *. That means the 1.5ghz were running Asyn. We know the sassamon was running ddr2 400mgz. the 1.5 GHz sassaman was running a 2.5 multiplier. so it was running @ 600FSB not 667FSB. The yonah appears to be setup the same way.
So what GHz you got those puppies at.
Intelia, believe me, the CPUz screen shot is wrong. Just think about it: have you ever seen any other CPU with a 2.5 multipler or something even close? You have a pentium 4, correct? check its FSB in CPUz and you'll notice it's 200 at stock for 800MHz FSB cpus. The clock steps would be huge if they actually used a 667MHz clock generator. I'm not going to try to convince you that it was running at 667, though I believe it was, just letting you know you cant trust CPU-z all the time, especially with unreleased CPUs. Hell, sometimes CPU-Z shows 3+ voltages on pentium Ds, which would probably melt the whole computer, and these are not new CPUs.
Now about the "lineare Hochrechnung der Prozessorleistung", this means that they PREDICTED the processor's performance at the higher clocks in a linear fashion, they only had 1.5GHz parts. Want proof aside from the fact that that's what the words actually mean? Take the Sossaman's dual socket scores. 884/663 = 1.3333 = 2000Mhz/1500Mhz. See? Linear progression. It seems they overstated the Yonah 2.0's performance a bit, but not too badly.
The K8 is a 1999 design btw. AMD's next gen stuff (k9/k10) is due out 2007.
Sick of these threads........Wait for the f*cking thing before you start saying' NAH INTEL AND DOTHANS OFFSPRING IS DE TEM TIMES FASTER THEN AMD64 AND ITS NOT GAY".
Itanium
Netburst
Prescott
Tejas
Nehalem..........To name a few Intel projects that didnt work out.
Sick of these threads........Wait for the f*cking thing before you start saying' NAH INTEL AND DOTHANS OFFSPRING IS DE TEM TIMES FASTER THEN AMD64 AND ITS NOT GAY".
Itanium
Netburst
Prescott
Tejas
Nehalem..........To name a few Intel projects that didnt work out.
Cerb
Elite Member
- Aug 26, 2000
- 17,484
- 33
- 86
This is all pointless speculation. Intel or AMD could make a CPU that's 10x as powerful as the ones we have--but who would buy it?
When the benchmarks have per-unit prices next to them, then start arguing over which one is best. Until there are prices to compare, this is even more pointless than neffing.
When the benchmarks have per-unit prices next to them, then start arguing over which one is best. Until there are prices to compare, this is even more pointless than neffing.
You missed that these chips are going to become the new desktop processors from Intel, didn't you? Intel KNOWS that the P4 ran out of room to grow, so they went to the only other processor design they had ALMOST ready for the desktop.
Things we can expect from Intel:
Somewhat better performance than we have seen since they havn't released any of these chips yet. Engineering samples are to get the motherboard and perhaps chipset manufacturers something they can test their development samples on.
A transition period where the application performance will be worse than previous processors as has been stated by many others. Different CPU architecture means compilers need to be updated, and old apps probably won't be recompiled so may run slower than from the current P4 and AMD64 based processors on the market.
A scramble by Intel to cover up the performance disparity between the "new" processors and older processors. They are showing the performance/watt stuff to try to cover it up. Performance/watt is only critical in notebooks. The only thing that may make this a logical approach is if there are new government regulations set to go into effect that will require government agencies to reduce the amount of power they use. At that point, Intel will be able to say, "Buy our stuff, it runs at half the speed of AMD processors, but uses a quarter of the power AMD needs". When it comes to government, they may use three times as many systems to get things done and cover it up by saying that each computer uses less power so it's acceptable.
AMD knows that their current processors and architecture is more than a match for the "new" Intel desktop and server processors. As a result, they will continue to tweak the performance, release higher clocked processors over time, and work on more innovative changes to their processors. It's speculated that more COMMON functions will be added to AMD processors in the future. We have the onboard memory controller right now, but what if a PPU were added, or something like that?
Intel has had a history of finding performance parity by increasing the amount of on-die cache on their processors. AMD has gotten their performance by finding ways to make the system as a whole run better.
Benchmarks will come out that obviously favor Intel, but when it comes to real-world performance, that's what matters to consumers. Intel has had an advantage in some benchmarks in the past because of the higher clock-rate on the processor. Going to the new processors will hurt that advantage.
And for those stuck on the whole clock rate MHz per MHz when it comes to performance, remember that any company could come out with a processor that only does PUSH/POP operations and clock it to 8GHz on air without a problem. How far you can push a CPU design is based on a lot of factors almost all of us have NO clue about.
The P3 ran out of room to grow based on the current technology of the time once Intel hit around the 1.1GHz mark. They learned many things back then, but what they didn't learn it seems is that a NEW design will give a lot more room for growth than going with an older design with new ways to make it go faster. AMD came out with the Athlon in those days, and it solidly kicked Intel's butt, forcing Intel to switch to NetBurst(P4). The design of the Athlon was better and allowed AMD to scale the chip well. Intel is returning to the P3 with updates, but how well can it scale after they were already forced to give it up for something else?
AMD also should be looking for an all new design for CPUs if they havn't already begun work on one. They know that they can probably scale the Athlon design to 3.8GHz or beyond, but there IS a limit. Perhaps they are working on a way to take single threads and split the instructions to make multi-core run single-threaded applications faster. It's possible to do this but is VERY complex and would take time. It WOULD give an advantage to the first company to implement it.
What's really going on in development at this point is known by a small number of people who are directly addressing the issue of "where do we go from here". The designers are doing things that the execs at both companies probably have NO clue about, and are doing the hard-core research and development, testing new ideas, throwing them out if critical flaws are found, etc. Not all good ideas are practical, and so, it takes a while for these things to happen. IDF is a function of marketing, and hype. If they had something that really beat AMD, we would have been hearing about it for weeks now.
Conroe doesnt have an on-die controller Intelia. You are thinking of Perynn. In addition Sossaman has a 14 stage pipeline and therefore can never be as fast per clock as current A64s since they use a 12 stage pipeline. Period
Hacp
Lifer
- Jun 8, 2005
- 13,923
- 2
- 81
Originally posted by: Intelia
Very nice for early M/B and cpu . It looks like the yonah clock for clock is going to be all over the AMD 64's . It kinda looks like it could be real one sided. Merom and conroe should really be fast. Just imagine a Conroe @3.0 GHz than in 2007 ondie memory controller this is going to be good'
Cinebench 2003 (32 bit)
2x Dual Sossaman 2,0 GHz* 884
2x Dual Sossaman 1,5 GHz 663
1x Dual Athlon 64 X2 4800+ 638
1x Dual Pentium XE 3,2 GHz 610
1x Dual Athlon 64 X2 3800+ 536
1x Dual Pentium D 3,2 GHz 528
%%%%%%%%%%%%%%%
1x Dual Yonah 2,0 GHz* 521AMD X-2 killer
%%%%%%%%%%%%%%%
1x Dual Yonah 1,5 GHz 383
1x Single Dothan 2,13 GHz 286
1x Single Pentium XE 3,2 GHz 284
1x Single Dothan 1,73 GHz 228
1x Single Sossaman 1,5 GHz 206
http://www.computerbase.de/news/hardwar.../august/idf_benchmarks_sossaman_yonah/
Repost. Why should this thread exist when another one was made earlier about the SAME topic?
http://forums.anandtech.com/messageview...atid=28&threadid=1672630&enterthread=y
I think everyone is missing the real point of this thread. Intelia is comparing AMD current offerings with Intels chips that are not even out yet. That is like saying "Hey my X2 destorys my sisters old pent III chip in speed" The new Intel chip better have better performance then what AMD has out now. If not, Intel better pack up and go home. Intelia if you want to make any logical argument wait until intel puts this chip out and compare it to what AMD has out THEN ... not what is offered now. Intel has no excuse to fall this far behind in performace, but they did. AMD beats what Intel has NOW..
Your argument that some future Intel chip is better then the current AMD chip is just crazy.
PKing
Your argument that some future Intel chip is better then the current AMD chip is just crazy.
PKing
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter