Sign in with
Sign up | Sign in
Your question

Penryn and Nehalem; Can AMD catch up? - Page 3

Last response: in CPUs
Share
April 3, 2007 5:42:48 PM

Quote:
Benchmarks? Benchmarks? Where are the Barcy benchmarks? Did I miss them? Was I sleeping? :lol: 


Oops hadn't realised they still wern't published. Sorry can't give any more specfics but they are good.

lol, thats all, just lol
April 4, 2007 10:39:10 AM

Hay JumpingJack.

You always been 85% to 100% Right. The 15% anyone can make Mistakes miss information and so on. Which says alot about you. It always funny when someone trys to prove you wrong and they fail. I have to say Thumbs Up to you. This Opterondo seems to worship a website Wikipedia. Which anyone can edit on the World wide Web.

JumpingJack you know your stuff and your skilled and one of the best on Tomshardware forum.

I bet most of the people on this forum will agree with me on this.
April 4, 2007 11:42:09 AM

Quote:
Hay JumpingJack.

You always been 85% to 100% Right. The 15% anyone can make Mistakes miss information and so on. Which says alot about you. It always funny when someone trys to prove you wrong and they fail. I have to say Thumbs Up to you. This Opterondo seems to worship a website Wikipedia. Which anyone can edit on the World wide Web.

JumpingJack you know your stuff and your skilled and one of the best on Tomshardware forum.

I bet most of the people on this forum will agree with me on this.


I was lurker for quite some time before I hit 'Register', and I always listened to what JumpingJack had to say. He's very knowledgeable and handles trolls/fanboys in a noble fashion. Well done, Jack.

This Opterondo fellow has been spewing AMD PR garbage and his won garbage a lot, quite recently.
Related resources
April 4, 2007 12:56:20 PM

,Still, while AMD is talking some interesting ideas, I don't see a whole lot of doing. They diffently need to saddle up, and rethink some strategies (if they have any)


Oh I see what your sayin, Intels only REAL competitor has no strategy, or needs to rethink on the one they have. LOL maybe they should consult with you.

Oh and this whole AMD being beaten by intel thing is going to Kill AMD. LOL

Look here kids, Just because Intel released a chip slightly better than the old K8, and has something slightly better than Conroe on the way, doesn't mean it's the end of AMD.

Just keep in mind the fact that when you see the Benchies, its the 5-year old K8 VS a brand new Tech. When it comes to FP calc's the K8's are right up conroes A$$.

So can someone please tell me why all the FANBOIS cant believe K10 will eat woodcrest and conroe?

Conroe = Updated K8 as for K10 and penryn, we shall see

To answer the thread Question I would simply reverse it. K10 and AGEA; can Intel catch up?

No im not a fanboi, but I don't like the ignorance&Speculation combo
April 4, 2007 1:12:29 PM

Operando has quite illustrated his idiocy, wonder where he is lurking now...

AMD is up against a monster with well laid out plans and technology for now and the upcoming future that looks quite promising and powerful... while still conserving energy... if only the GPU manu's would catch up...

Regardless, no it is not the end for AMD as a company, but it is the end of the era where they were fastest and most desired... for now and up coming for quite a while, AMD will be playing catch up, and while doing so will be restricted to the mid to low end market for income, which they are under a great deal of debt... that makes things even tougher...

While Intel is liquid and remains profitable... a much better position, especially when combined with the facts that they have the best processor and looks to also have the next one too, and are outlining the ones to come... the road ahead for them looks promising.

AMD is tight lipped with thier next processor, and ones after that... and the only reason I can make sense of for being tight lipped is that what they have seen so far internally, is not as competitive nor as cheap to manufacture... so it seems like a tough road ahead.
April 4, 2007 1:42:09 PM

He'll be back with a new Wiki posting to back up his claim *waits*
April 4, 2007 5:54:37 PM

Quote:

Oh I see what your sayin, Intels only REAL competitor has no strategy, or needs to rethink on the one they have. LOL maybe they should consult with you.

Oh and this whole AMD being beaten by intel thing is going to Kill AMD. LOL

Look here kids, Just because Intel released a chip slightly better than the old K8, and has something slightly better than Conroe on the way, doesn't mean it's the end of AMD.

Just keep in mind the fact that when you see the Benchies, its the 5-year old K8 VS a brand new Tech. When it comes to FP calc's the K8's are right up conroes A$$.

So can someone please tell me why all the FANBOIS cant believe K10 will eat woodcrest and conroe?

Conroe = Updated K8 as for K10 and penryn, we shall see

To answer the thread Question I would simply reverse it. K10 and AGEA; can Intel catch up?

No im not a fanboi, but I don't like the ignorance&Speculation combo

* sigh *

Neither I, Jumping Jack, or most of the others are claiming AMD is dead or dying. But you can't deny that AMD is in a very rough place and, for the most part, they did it to themselves. AMD isn't stupid, but they should have seen this coming. For the last 2 years AMD has become more and more sporadic/late with their releases, and more and more tight-lipped/vague with their info. It's become the flip-flop of 3 years ago, when intel first released NetBurst, months after the K8 line, and still couldn't quite catch-up. Now AMD is on the other side, and they're not proving themselves to be much better (so far).

BTW, the Conroe/CTD line is not just slightly better than K8, it's pretty signficant. Less heat, better power management, and some of the most significant OC'ing (headroom) in any CPU line in a long, long time. Someone was claiming earlier in the thread that the 6000+X2 was the equal of the E6600 except for a couple of things:

1). The E6600 is clocked at 2.4 Ghz, to the X2's 3.0 Ghz.
2). The E6600, when OC'ed to 3.0 Ghz (easy), will mop up that same X2, including FP calc's.
3). At stock speed, the E6600 is only using 65-80W. The X2 is using 90-125W (load values).These are significant power values, and also significant heat disapation differences.

BTW, Penryn will introduce the new 'MG+HK' transistor technology, which, along with the die shrink, is expected to significantly improve the power and heat management even more, while increasing (transistor)switching speeds even further.

As, far as Barcelona 'eating' Woodcrest, Intel is setting up Clovertown, a QuadCore server version of the Penryn to compete head-to-head with AMD. And from early indicatications (info), it certainly won't be a slouch either. Plus Gesher, will be primereing next year, at the newer 32nm. fab.

And as far as Agea goes.. Nehalem will be starting to roll out aproximately this time next year, with IMC, a new p-p serial interface (replacing the FSB), and up to 8 cores, and possibly even including an on-die IGP, all at 45nm. By years end, it will start transitioning to 32nm.
April 4, 2007 11:07:29 PM

I see your point , but this guy is on FIRE. He's talkin bout AMD needing to rethink there strategy... come on. Hasn't AMD always been tight lipped?
They don't need much to take on CORE arch. It will be interesting to see what they have in store for integer perf. , even I dout K10 will beat core there. As far as Fp's I will bet K10 gets near 50% over core2. That's right , I said it. It had to be said. Here we go... now you guys can rip me apart

Non-Bias
April 4, 2007 11:31:25 PM

You are comparing a mid-higher range core2 vs the fastest K8. not known for its efficiency. you are also quoting Intels TDP witch is not accurate. What is the northbridge doin?

If you can read the benchmaks instead of just seeing how far the bar extends, you would see the two are not far apart.

OK , so basically Core2 has k8 by an average of 650mhz (this is what ive concluded so correct me if im wrong).

Remember when k8 had that net thing by 1000mhz ?

so Intel made a HUGE improvement over the p4 , and if you look at it the same way I do then maybe you would come up with this number: 1600mhz
That is one Kick ass improvement, AMD needs 600 to tie.

Before I Get jumped , I know this is no real way to gauge performance. just a really ruff :idea:
April 5, 2007 7:31:31 AM

Energy efficency does come into play if you're trying to 'catch-up', just like Intel tried to do with the Pentium D 'Extreme Edition'. Yeah, it could barely manage to catch the AMD mid to top range, put it used an unholy amount of power (225W!!) and put off enough heat to keep your whole neighborhood nice and toasty during the month of Febuary. Now AMD is in the same position, where basically the choices are to ramp up the Ghz (and power, and heat) and try to 'brute force' your way to the top (which is exactly what the FX-7x 4x4 series try to do), or you go back to the engineering board and start a new CPU design (which Intel just did).

BTW, the Ghz to Ghz comparison is a really bad comparison. You have to take into account the whole package; power efficency, heat, instructional latency, memory latency, etc. Simply coming out with a 3.6 GHz FX-66 still will not win (almost)anyone over (of course there are always a few idiots looking for the worst deal in town). It would probably require over 150W (probably closer to 200W), it would definitely require some rather specialized cooling (like the 4x4's), and it would cost a small fortune (probably around $1000, due to low batch availablity). So, when you figure it all out, you'd probably spend close to $1500 just for the CPU and cooler and PSU, just to MATCH the true performance of a E6700, maybe close to an X6800, which costs $500 and $800 each, respectively.

Again, only a die-hard IDIOT would fall for it.

So, the only real choice for AMD is to re-invent the wheel. Yes, it's a pain. Yes, it costs gobs of money. Yes, it takes time. And Barcelona looks interesting. But, even the experts (anadtech, xbitlabs, and other industry analysts), are concerned that from the info that AMD is releasing, it points to Barcelona being more evolutionary, then revolutionary. And simply tidying up a few areas isn't going to win back a lot of customers for AMD.

Also, AMD has NOT always been 'tight-lipped' this way. The A64's/Opterons were widely publisized and analyzed MONTHS(in some areas, up to a year) before they were ever released. Intel was the one who kept tight-lipped, and released the infamous 'NetBurst' we all love to hate. So, AMD is really shooting itself in the foot by witholding information and being (purposely)vague about the CPU and launch details. It's exactly this 'indecisivnous' that's causing their stock to tank, as stock-holders are insecure that AMD can continue to keep it's market-shares.
April 5, 2007 8:35:13 AM

INTEL is just scared that AMD will rise from the ashes and kick their proverbial butt again like they did with the Athlon64 processor.AMD caught INTEL with their pants down and emberassed the heck out of them.Now INTEL is being a poor sport and throwing sucker punches,hoping to keep AMD ducking.INTEL does not want AMD to get too big,otherwise AMD becomes a very serious competitor.Not that already aren't,but INTEL has more money than AMD and so they hit hard and fast to try and knock AMD out.NOT GONNA HAPPEN.

Dahak

AMD X2-4400+@2.6 TOLEDO
EVGA NF4 SLI MB
2X EVGA 7950GT KO IN SLI
4X 512MB CRUCIAL BALLISTIX DDR500
WD300GIG HD/SAMSUNG 250GIG HD
ACER 22IN WIDESCREEN LCD 1600X1200
THERMALTAKE TOUGHPOWER 850WATT PSU
COOLERMASTER MINI R120
3DMARK05 13,471
April 5, 2007 9:02:02 AM

Not quite the usual balance that I usually see... Just an observation.
April 5, 2007 9:17:01 AM

Quote:
You are comparing a mid-higher range core2 vs the fastest K8. not known for its efficiency. you are also quoting Intels TDP witch is not accurate.


Intel isn't accurate, but AMD is?

Even sites like Anandtech agree to this 65w TDP rating (for the E6300-E6700).

Quote:
INTEL is just scared that AMD will rise from the ashes and kick their proverbial butt again like they did with the Athlon64 processor.AMD caught INTEL with their pants down and emberassed the heck out of them.Now INTEL is being a poor sport and throwing sucker punches,hoping to keep AMD ducking.INTEL does not want AMD to get too big,otherwise AMD becomes a very serious competitor.Not that already aren't,but INTEL has more money than AMD and so they hit hard and fast to try and knock AMD out.NOT GONNA HAPPEN.


...except business isn't like a boxing match.

Intel have got all platforms and levels covered when it comes to processors:

Desktop (current):
- Allendale, Conroe, Kentsfield

Server (current):
- Woodcrest, Clovertown and Tulsa
- Itanium II

Mobile (current):
- Yonah
- Merom

Future:
- Penryn (Core 2 revision slated for this year for desktop, server and mobile)
- Nehalem (Major Overhaul in 2008)
- Gesher (TBA in 2010)

Even if Barcelona releases in Q2/Q3 of this year, Nehalem (which looks like a monster) will just rain on its parade all over again.
There's also the price cuts when Penryn/Nehalem come in, which will lower Conroe/Allendale prices to the gutter levels ($200 Q6600, anyone?).

What does AMD have to answer to Penryn, let alone Nehalem?
April 5, 2007 10:36:52 AM

Does anyone have benchmarks of Penryn and Nehalem?
April 5, 2007 12:03:53 PM

Quote:
Does anyone have benchmarks of Penryn and Nehalem?


Nope.

Nehalem hasn't even reached working silicon.
April 5, 2007 11:00:17 PM

simply redesigning the k8 would be enough to compete head to head with core2 arch. Obviously energy efficiency comes into play, but Netburst was much further away from the performance required to take the crown. Intel made a Big Leap ahead...(no punn intended). AMD needs a skip. AMD already has core2 beaten, its just a matter of how much more performace they can get. Lets not talk energy efficiency cause intel is new to that game.

As far as integrated graphics is concerned Intel wont be able to touch AMD on the high end because of ATI. High end GRFX is really hard to do and I dont see intel being able to compete for a couple of years, unless the NV wild card has something to say.
April 5, 2007 11:02:57 PM

I just new someone would run with that.
Im not sayin Intels power rating is incorrect, but not fair when comparing to an AMD proc
April 5, 2007 11:35:36 PM

Quote:
Obviously energy efficiency comes into play, but Netburst was much further away from the performance required to take the crown.
...Lets not talk energy efficiency cause intel is new to that game.

Quote:
Im not sayin Intels power rating is incorrect, but not fair when comparing to an AMD proc.


But isn't this EXACTLY what AMD hammered them on for the last 3 years? NOW it's not fair? *Oops, my 'Biased-Fanboy Detection System'(also known as BS) radar is going off.* BTW, Intels TDP is based on worst case, full-load conditions. Most 'real-life' tests show it using much less (-20% avg.).

And how is it not fair? We're comparing 65nm dual-cores to 65nm dual-cores.. am I missing something?

Quote:
Intel made a Big Leap ahead...(no punn intended). AMD needs a skip. AMD already has core2 beaten, its just a matter of how much more performace they can get.


Interesting.. somehow AMD leaped ahead again, and I haven't heard about it yet. 8O Please, I need links to where I can see the benchies and technichal reviews. Please, supply. Pretty please.. :wink:
April 5, 2007 11:57:30 PM

Quote:
Not only is AMD meeting Intel head-to-head;

BS! AMD have no CPU that can compete against E6700, X6800, Q6600, QX6700.

Hungry fish = easy troll.
April 6, 2007 12:40:46 PM

Ya know, you have not been answering my questions, and have been missing my points all along.

everytime i post something you reply with captain obvious type info. You have not told me anything I dont already know.

I never said AMD made a leap.
You still dont get the whole power consumtion thing
To prevent myself from hearing the same thing over and over (from you) we are done. Redundancy KILLS me

Next time we get into it please answer my questions or respond to them as they come, and not go into your own fanboi tangent everytime

If you backtrack you will see this, and i didnt say anything in the beginning.
Until next time. :) 
April 6, 2007 3:32:49 PM

Quote:
Ya know, you have not been answering my questions, and have been missing my points all along.

everytime i post something you reply with captain obvious type info. You have not told me anything I dont already know.

I never said AMD made a leap.
You still dont get the whole power consumtion thing
To prevent myself from hearing the same thing over and over (from you) we are done. Redundancy KILLS me

Next time we get into it please answer my questions or respond to them as they come, and not go into your own fanboi tangent everytime

If you backtrack you will see this, and i didnt say anything in the beginning.
Until next time. :) 


He's got a point. I would also love to see the benchmarks that show AMD beating a Core2. Anyone else?
April 6, 2007 5:26:51 PM

Quote:
Ya know, you have not been answering my questions, and have been missing my points all along.

everytime i post something you reply with captain obvious type info. You have not told me anything I dont already know.

I never said AMD made a leap.
You still dont get the whole power consumtion thing
To prevent myself from hearing the same thing over and over (from you) we are done. Redundancy KILLS me

Next time we get into it please answer my questions or respond to them as they come, and not go into your own fanboi tangent everytime

If you backtrack you will see this, and i didnt say anything in the beginning.
Until next time. :) 


In other words:
"I can't refute your arguments, and I can't support mine."
April 6, 2007 5:33:00 PM

:roll:

Yeah,like... Whatever. Make sure you don't miss that little yellow bus in the morning. Tell your pediatrician he needs to up the Rydelin dosage, and work harder on your basic comprehension skills.

I get the power consumption thing just fine. You, however, are the one who doesn't seem to get it. I have an AA in electronics. And I like to try to keep somewhat current in electronics, even if I don't work in that field anymore. Especially where it pertains to computers. So, I can seperate a lot of the 'market speak' from the facts, and deduct and infer from there.

I don't take either Intel or AMD strictly at their word. Most companies try to 'spin' things to try to show everthing in the most positive light. Because they want and need your dollars. So, you need to take a somewhat cynical view when ANY company is being vague and 'tight-lipped'. More information shows more progress, because they usually don't reveal a lot of details until they are finalized. Based on details, Intel definitely knows where it's going. AMD seems to be having more difficulty.
a b à CPUs
April 8, 2007 12:34:00 AM

Quote:
I just got the LAUGH OF MY LIFE with the K6-3 VS P3 compairison.Have you ever even owned a friggin K6-2 or K6-3 series which was released about the time of the K7.They K6-3 was NOTHING MORE than a k6-2 produced with 256K of L3 Cache running at the processor speed which was never higher than 570Mhz.All it was made for was a last ditch effort to use up the remaining SS7 parts and to provide an upgrade path for these boards.Now a K6-3 + , which is a mobile chip that can overclock like the wind , can run 650MHZ-700 EVEN but will never in a million years beat a good P3 , I mean they are different generations I mean come on , you defend AMD with an intel mindset(oh goodie a K6-3 must compete with a P3 because it has a 3 in it too , ohhh wonder wat the K7 runs with.Come on AMD has been and will be " around" but they will never have the Abilities that intel has but what good are they if intel is tooo stupid to use them.


P2 350 vs K6-2 500, the P2 aces it too easily

i try overclocking the 500 to 600, it kills it in munites and never boots again! (2.2v -> 2.4v)

i overclock my dual celeron 333a's to 550 at 2.4v and they work for months! (still own em, in my cupboard)

the only time AMD ever owned Intel was with the K8, and only because Intel kept there acthitecture (netburst) for too long and got lazy.

Intel is back on track

Long Live P6, the best core ever made.
April 8, 2007 2:06:49 AM

Hello Jesus's friend :D 

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx...

Quote:
Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".
April 8, 2007 3:31:40 AM

Quote:
I'd like to clear out some misinformation that is going on in this forum:

1) Penryn's 45% advantage over Cloverton is just a blatant lie from intel.
Penryn's 3.2 GHz advantage over a 3.0GHz cloverton would be only 20% (according to different media) and even Ed (from overclockers) has also agreed with this. My best guesstimate is that Penryn would be better than cloverton 8-10% clock for clock. This is NOT enough to counter barcelona. Intel's only answer is clock speed which is something that they are accustomed to deal with.

2) Penryn is intel's answer to catch up to Barcelona's level. Still, we have to wait on the benchmarks to get a final conclusion, but to say that Penryn will roll over barcelona (on servers) and Agena (on desktops) is just FUD and a nasty wet dream.

Well, that's all my take for now on this topic.
I believe it would be prudent not to talk about "1 vs 2" performance before we get some benchmarks.



Can you provide a factual basis for making these unsubstantiated claims? While every single statistic provided by a manufacturer should be questioned until independently verified I am at a loss for why you seem to accept AMD's unsubstantiated data as fact yet proclaim Intel's a lie?

You do realize that your forum namesake would likely not approve of such Hypocrisy? And your "best guesstimate" is something we should accept over what limited info has been released from AMD and Intel?
April 8, 2007 3:33:35 AM

Quote:
1) Penryn's 45% advantage over Cloverton is just a blatant lie from intel.
Penryn's 3.2 GHz advantage over a 3.0GHz cloverton would be only 20% (according to different media) and even Ed (from overclockers) has also agreed with this. My best guesstimate is that Penryn would be better than cloverton 8-10% clock for clock. This is NOT enough to counter barcelona. Intel's only answer is clock speed which is something that they are accustomed to deal with.


Gee, that's amazing. They supossedly just came out with Penryn on 'first silicon', and you've already got benchmarks! And of course, your 'best guesstimate' is a recognized standard in both the media and engineering world.. :roll:

Penryn is more than just a die shrink. It will also incorporate Intels' new "Metal Gate and High K Die-electric" (MG+HK) transistor technology, SSE4, and bigger caches. Basically, it's testing alot of the technologies that will be going into Nehalem next year. Anyway, there's a lot more involved, and it could add up to a bigger increase than you expect. I sincerely believe it will be a lot better than Brisbane..

Quote:
2) Penryn is intel's answer to catch up to Barcelona's level. Still, we have to wait on the benchmarks to get a final conclusion, but to say that Penryn will roll over barcelona (on servers) and Agena (on desktops) is just FUD and a nasty wet dream.


No, Penryn is not Intel's answer to Barcelona. It is a set-up to Nehalem (next year), and eventually Gesher (2010). THEY are Intel's response to Barcelona, Agena,and Kuma.

Quote:
It is really lame from intel that a 3.2GHz Penryn with an 1600FSB can only beat a Cloverton core clocked at 3.0GHz with an 1333FSB with just a mere 10% on average and 20% in specific benchmarks.

Just as I said before, let's wait for REAL benchmarks before making any judgements and not eat intel's marketing fud.


Maybe you should follow your own advice, and stop spreading your own FUD, because I doubt you can back it up either. So, unless you can post some benchmarks, with a Penryn, your just spouting the same FUD and marketing talk you claim to hate.
April 8, 2007 3:36:47 AM

Quote:
It is really lame from intel that a 3.2GHz Penryn with an 1600FSB can only beat a Cloverton core clocked at 3.0GHz with an 1333FSB with just a mere 10% on average and 20% in specific benchmarks. :wink:

Just as I said before, let's wait for REAL benchmarks before making any judgements and not eat intel's marketing fud.


Where has Intel claimed 10% anything? Did you expect any gains from the FSB to 1600 in gaming when it seems to be negligible 800 to 1066 (E4300 benches)?
April 8, 2007 5:36:19 AM

Quote:
It is really lame from intel that a 3.2GHz Penryn with an 1600FSB can only beat a Cloverton core clocked at 3.0GHz with an 1333FSB with just a mere 10% on average and 20% in specific benchmarks. :wink:

Just as I said before, let's wait for REAL benchmarks before making any judgements and not eat intel's marketing fud.


Where has Intel claimed 10% anything? Did you expect any gains from the FSB to 1600 in gaming when it seems to be negligible 800 to 1066 (E4300 benches)?

It will be the switching speed of the transistors, added cache, FSB increase, new instructions, and increased clock speed that will make it faster. I was just told how much in a meeting but can not disclose it.
!