Jump to content
Sign in to follow this  
mattwalter85

AMD Phenom vs Intel Penryn

Recommended Posts

well evidentally AMD's new Phenom is going to be coming out a week after the Penryn and has some extras intel hasn't thought of....L3 cache, Hyper-Transport 3 and it will be a true quad-core not just two duo's slapped together. i'm interested to see what happens. what do you all think?

 

http://www.amd.com/us-en/Corporate/Virtual...~117412,00.html

Share this post


Link to post
Share on other sites

being an amd fan I really hope amd can really outperform the intel's out currently and the penryn. Sadly ive seen some early benches using the server version of the phenom and it doesnt look to good. But well have to see when phenom samples get out. I like amd more than intel dunno why just do so ive been hoping that while im saving for my next rig which if i had the money now would be a c2d or quad core. but hopefully by the time i save up amd comes out with a beast. :D

Share this post


Link to post
Share on other sites
well evidentally AMD's new Phenom is going to be coming out a week after the Penryn and has some extras intel hasn't thought of....L3 cache, Hyper-Transport 3 and it will be a true quad-core not just two duo's slapped together. i'm interested to see what happens. what do you all think?

 

http://www.amd.com/us-en/Corporate/Virtual...~117412,00.html

 

 

 

fyi... L3 cache = pointless without software written to use it... aka less than a handful of apps.

 

 

oh and while we're at it. penryn isn't a pair of dual cores unlike the current quad core chips. it also has a mamoth L2 cache by comparison.

 

 

and finally, hyper transport... is only used by AMD.

Share this post


Link to post
Share on other sites
fyi... L3 cache = pointless without software written to use it... aka less than a handful of apps.

The software being run has very little to do with the effectiveness of the caching. It has absolutely nothing to do with how it is used, as software doesn't control caching at all.

Share this post


Link to post
Share on other sites

ok... the sky is NOT blue

 

 

 

the L3 cache is NOT controlled by the software? no chit. it can sit there and bounce it's sack around in public all it wants... but if software ain't asking to use it... it's USELESS.

 

 

history lesson time. P4 3.2 EE vs the P4 3.2c. difference? 2mb of L3 cache. performance gain? only in adobe / autodesk apps, and ID games of the day. NOTHING in real world (screw benchmark apps, they're only used to sell more CPU's) out side of that handfull of apps saw any gain because the software wasn't using the USELESS cache.

 

you can go back and talk about the other L3 debate, the K6-2 vs K6-3. however there was one other MAJOR factor that most buttnugget review sites missed out on while they were getting stiffies over the L3 cache. an FPU performance jump of 50%. wonder why real world benchmarks on the K6-3 went up like that? I don't. <_< try disabling the L3 in the bios... ooops same results!

 

another pattern to this... AMD and Intel have done the L3 thing on thier "flagship desktop CPU's" before. what happened? L3 went right back on the shelf as another gimick and the world moved on. apparently from that 2 things can be said... 1: it doesn't work as advertised and 2: if the software don't ask for it, it ain't doing anything.

 

 

but back to your regular scheduled session of arguing the color of the sky

Share this post


Link to post
Share on other sites
ok... the sky is NOT blue

the L3 cache is NOT controlled by the software? no chit. it can sit there and bounce it's sack around in public all it wants... but if software ain't asking to use it... it's USELESS.

history lesson time. P4 3.2 EE vs the P4 3.2c. difference? 2mb of L3 cache. performance gain? only in adobe / autodesk apps, and ID games of the day. NOTHING in real world (screw benchmark apps, they're only used to sell more CPU's) out side of that handfull of apps saw any gain because the software wasn't using the USELESS cache.

 

you can go back and talk about the other L3 debate, the K6-2 vs K6-3. however there was one other MAJOR factor that most buttnugget review sites missed out on while they were getting stiffies over the L3 cache. an FPU performance jump of 50%. wonder why real world benchmarks on the K6-3 went up like that? I don't. <_< try disabling the L3 in the bios... ooops same results!

 

another pattern to this... AMD and Intel have done the L3 thing on thier "flagship desktop CPU's" before. what happened? L3 went right back on the shelf as another gimick and the world moved on. apparently from that 2 things can be said... 1: it doesn't work as advertised and 2: if the software don't ask for it, it ain't doing anything.

but back to your regular scheduled session of arguing the color of the sky

 

^ agreed. You can't claim that L3 is something Intel never tought of, cause they did, look at the Itanium. The Itaniums 7 years ago performed like utter crap and the L3 cache did nothing of any benifits.

 

Also, Hyper Transport is AMD's tech of Bus speed, not Intel's, and to be honest, they don't need it either. If you can prove me wrong, please post benifits.

 

Alas, 2 cores on 1 die, without behing two physical cores...... Intel has thought of that, and probably already has done it (hmm, I believe Penryn is supposed to be 2 cores (1 physical) on 1 die) but the Kentsfield was different, because it is more of a marketing strategy. Why wait, when you can beat your opponent at it faster? Yes, it was more power hungry and produced more heat, but it still worked, did it not?

Share this post


Link to post
Share on other sites
well evidentally AMD's new Phenom is going to be coming out a week after the Penryn and has some extras intel hasn't thought of....Hyper-Transport 3

This is like saying Burger King is better than MacDonald's because MacDonald's hasn't thought of the whopper. :rolleyes:

 

Hyper Transport is AMD technology. Intel doesn't use it, and never has, never will. Also, I'd hardly say that Intel "hasn't thought" of L3 cache.

 

So far, AMD has just seemed to drop the ball over and over again with Phenom and all the Barcelona chips. Delays on top of delays, shady simulated benchmarks followed by weak real benchmarks, and so far, no one seems to be that excited about the server chips. Anything can happen with Phenom, and I truly hope that it comes out and stomps tail, but so far I just don't believe that's going to happen. We'll know in a little while though, I guess.

Share this post


Link to post
Share on other sites
This is like saying Burger King is better than MacDonald's because MacDonald's hasn't thought of the whopper. :rolleyes:

 

Hyper Transport is AMD technology. Intel doesn't use it, and never has, never will. Also, I'd hardly say that Intel "hasn't thought" of L3 cache.

 

So far, AMD has just seemed to drop the ball over and over again with Phenom and all the Barcelona chips. Delays on top of delays, shady simulated benchmarks followed by weak real benchmarks, and so far, no one seems to be that excited about the server chips. Anything can happen with Phenom, and I truly hope that it comes out and stomps tail, but so far I just don't believe that's going to happen. We'll know in a little while though, I guess.

 

I think the only thing that is certain is that the AMD chip will probably be cheaper then the Penryn but by how much nobody knows. AMD chips always seem to be cheaper then intel's, thats why i dont have one yet. If this Phenom comes out and falls on its face i'm joinin the intel team until AMD gets their stuff together.

Edited by mattwalter85

Share this post


Link to post
Share on other sites
ok... the sky is NOT blue

the L3 cache is NOT controlled by the software? no chit. it can sit there and bounce it's sack around in public all it wants... but if software ain't asking to use it... it's USELESS.

You do realize that software has nothing to do with what gets cached nor if it gets used, so why argue about the software side of things?

 

It gets asked to use it any time something misses on the L1 and L2 caches. It just so happens that the caching algorithms on the older CPUs didn't help performance much...that doesn't mean that it will do the same on the Phenom. If they lower the miss rate enough through intelligent caching it could very well increase performance on a large number of applications that are memory intensive. Also - newer applications benefit from L3 more than older ones. Look at recent benchmarks of the P4EE versus the regular P4. Despite the added latency of having the L3 cache it does increase performance.

Share this post


Link to post
Share on other sites

haha that guy beat me too it haha

 

that was kind what i was going to say. amd (MacDonalds) vs intel (Burger King) intel new whopper is probably going to be the best but amd's got the dollar menu... it might not be as good but cheep and fills you up...

 

 

haha

Share this post


Link to post
Share on other sites

you're not getting it. it can have all the level's of cache it wants. the cache is READY to go when asked to do it's job. but if the software doesn't ask. it CAN'T DO ANYTHING.

 

 

computer hardware is not self motivated. a computer is just like a car. and the L3 cache on a computer's CPU is no different than the air conditioner. it doesn't turn it self on to "do you a favor" or out of "good will". it does so because it was "told to". and without software telling the CPU what do, that L3 cache isn't going to do anything.

 

 

 

 

 

btw, itanium failed because of the IA64 architecture not the L3 cache. Itanium 2 moved to a 12mb L3 and more was planned. but x86-64 took over as the standard due to backwards compatibility.

Share this post


Link to post
Share on other sites
you're not getting it. it can have all the level's of cache it wants. the cache is READY to go when asked to do it's job. but if the software doesn't ask. it CAN'T DO ANYTHING.

All memory requests go through every level of cache. That's how they work, nothing is required in software other than a request for data.

 

This is why the memory latency is higher on the new AMDs, they have to go through an extra level of cache misses before getting to the main RAM.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

×
×
  • Create New...