A First Look at AMD's Triple Core Phenom

[muk]

Although it will not be available before the second quarter of 2008, we decided to take an early look at the performance of AMD's upcoming triple core Phenom 8000 series. Read on for a comparison of Phenom performance at one, two, three and four cores.

http://www.tomshardware.com/2008/02/28/a_first_look_at_amd_triple_core_phenom/

 

[MrLinux]

Error404

Sorry! The Requested Page was not Found!

Dear Visitor. The page you requested couldn't be found 🙁 - If you followed a link from another Website please inform their Webmaster. If you happen to get this message while browsing Tom's Hardware Guide please inform the Webmaster of Tom's Hardware Guide.


?

 

[gpippas]

The tri-core gets more L3 cache per core to play with.

 

[PeterHighlander]

I still visit Tom's Hardware regularly however the reviews over the past say 4-6 months really seem to have missed the boat. Is it just me or have others found the comments / questions answered by reviews to be poor?

I don't think there is much surprise in this article. Wow the tri core performs between the dual and quad in thread optimized code! Shocker! Wow, in semi optimized code the Tri and Quad perform very similar wow! I mean come on.

Here are some bigger questions IMO:
1) Cache. Will the Tri core have access to the same cache as the Quad? Thus in applications which require a large cache, will the increased cache hit rate make up for the loss of a core?

2) Overclocking! Let's face it if you don't overclock or are on a tight budget, AMD is likely your choice. Yeah it gives up some 200-600MHz to the C2D but for a given amount of cash, AMD has a solid position. Once you begin overclocking this all changes. The C2D simply blows AMD away (Or if you have several hundred to spend on the CPU). For example, I bought X2 5600 w/ 1Meg cache. $100 bucks delivered back before the abundance of C2D options. OC'd it to 2.95G. 690 chipset, AM2 socket. Current quads are a bad upgrade choice. Well this tri may be perfect! One less core, less heat, better OCing, yes yes?
*note* I did read the article, heat production was still up indicating that the core was likely running to some degree. However you didn't even attempt OCing!

Anyway, Tom's is no longer my first stop for "solid" reviews because these reviews simply don't answer the important questions.

Simply Disappointed.

 

[Reynod]

An interesting read,

The point about power consumption is a good one.

I wonder what the impact will be for a disabled core and how well the Agina X2's will fare.

Thanks.

 

[jimmysmitty]

I still visit Tom's Hardware regularly however the reviews over the past say 4-6 months really seem to have missed the boat. Is it just me or have others found the comments / questions answered by reviews to be poor?

I don't think there is much surprise in this article. Wow the tri core performs between the dual and quad in thread optimized code! Shocker! Wow, in semi optimized code the Tri and Quad perform very similar wow! I mean come on.

Here are some bigger questions IMO:
1) Cache. Will the Tri core have access to the same cache as the Quad? Thus in applications which require a large cache, will the increased cache hit rate make up for the loss of a core?

2) Overclocking! Let's face it if you don't overclock or are on a tight budget, AMD is likely your choice. Yeah it gives up some 200-600MHz to the C2D but for a given amount of cash, AMD has a solid position. Once you begin overclocking this all changes. The C2D simply blows AMD away (Or if you have several hundred to spend on the CPU). For example, I bought X2 5600 w/ 1Meg cache. $100 bucks delivered back before the abundance of C2D options. OC'd it to 2.95G. 690 chipset, AM2 socket. Current quads are a bad upgrade choice. Well this tri may be perfect! One less core, less heat, better OCing, yes yes?
*note* I did read the article, heat production was still up indicating that the core was likely running to some degree. However you didn't even attempt OCing!

Anyway, Tom's is no longer my first stop for "solid" reviews because these reviews simply don't answer the important questions.

Simply Disappointed.

Well considering they have yet to recieve a real tri core from AMD, OC'ing will not fare any better. It is the same 9900 they have been using so even with 1, 2 or 3 cores disabled it will more than likely OC the same.

Once THG gets a real tri core they will try OC'ing it but there is no guarantee it will do any better just like B3 has no guarantees.

 

[zenmaster]

I still visit Tom's Hardware regularly however the reviews over the past say 4-6 months really seem to have missed the boat. Is it just me or have others found the comments / questions answered by reviews to be poor?

I don't think there is much surprise in this article. Wow the tri core performs between the dual and quad in thread optimized code! Shocker! Wow, in semi optimized code the Tri and Quad perform very similar wow! I mean come on.

Here are some bigger questions IMO:
1) Cache. Will the Tri core have access to the same cache as the Quad? Thus in applications which require a large cache, will the increased cache hit rate make up for the loss of a core?

2) Overclocking! Let's face it if you don't overclock or are on a tight budget, AMD is likely your choice. Yeah it gives up some 200-600MHz to the C2D but for a given amount of cash, AMD has a solid position. Once you begin overclocking this all changes. The C2D simply blows AMD away (Or if you have several hundred to spend on the CPU). For example, I bought X2 5600 w/ 1Meg cache. $100 bucks delivered back before the abundance of C2D options. OC'd it to 2.95G. 690 chipset, AM2 socket. Current quads are a bad upgrade choice. Well this tri may be perfect! One less core, less heat, better OCing, yes yes?
*note* I did read the article, heat production was still up indicating that the core was likely running to some degree. However you didn't even attempt OCing!

Anyway, Tom's is no longer my first stop for "solid" reviews because these reviews simply don't answer the important questions.

Simply Disappointed.

I have to agree.
The reviews have been really lacking.
With the Tri-Core Phenom I can understand the limits because they did not have one to test.
So the results are somewhat limited.
They simply show the painfully obvious.

I would still love tot see THG publish a REAL Athlon vs Phenom article.
The previous one was a joke. (OC'd Phenom vs UnderClocked Athlon)
Let's See a Stock Phenom vs a Stock X2-6400.
Let's OC both to reasonable levels.
What are the results.

Core vs Core Test - Yet another joke.
This fails to recognize that when more cores are enabled the L3 cache becomes less efficient.
By running the X2 slow, it fails to account for the fact Phenom performance scales less per Mhz than the X2.
So while it may be faster @ 2.4ghz with a single core, that will drop as speeds reach 3.0Ghz and Beyond.
It will drop when you have more cores enabled.

I must sadly admin that I like the THG forums, I love the monthly Video Card Roundup, the news links are useful, but the actual articles are really slipping.

 

[mi1ez]

I think the only thing this review shows is how much a second core helps single threaded applications by offloading background tasks (networking etc.) from where the app is running.

 

[coret]

I think the only thing this review shows is how much a second core helps single threaded applications by offloading background tasks (networking etc.) from where the app is running.

And along the same lines, it shows the advantage for tri has over dual being equal to the advantage quad possesses over dual for dual core optimised software.

Actually seeing this properly, I'm actually quite tempted by a tri-core ... depending on how they do once released anyway.

 

[caamsa]

Hmmm this could be the recession chip. With the economy slowing down this could be exacty what AMD needs.......kina ironic if you think about it. Cheap CPU, good performance. McDonalds could add it to their dollar menu.

 

[BSMonitor]

Well considering they have yet to recieve a real tri core from AMD, OC'ing will not fare any better. It is the same 9900 they have been using so even with 1, 2 or 3 cores disabled it will more than likely OC the same.

Once THG gets a real tri core they will try OC'ing it but there is no guarantee it will do any better just like B3 has no guarantees.

Exactly.

AMD is simply floundering. All this points to is very poor yields of the Agena core. A 2.6GHz chip was supposed to be available in November. Now reviewers of the Tri-Core chip still do not have samples weeks from release? 3 months after the 2.6GHz chips were supposed to be available?

Hell, we are only 6 months away from another new Intel architecture. By the time anyone can get a 2.6 GHz AMD quad-core, Intel will have Penryn 3.2GHz quads in the bargain bin.

Of course knowing Intel, we won't see Nehalem in desktops until there is a competitive reason to do so. So maybe Penryn won't be really cheap for a while.

 

[Scarchunk]

I know the Vista index score is BS, but does anyone else think that its weird that a single core Phenom @ 2.6 scores a perfect 5.9 for the cpu score?

http://www.tomshardware.com/2008/02/28/a_first_look_at_amd_triple_core_phenom/page7.html

 

[yomamafor1]

I think its bogus. But again, to use WEI as a benchmark is bogus to begin with.

 

[Scarchunk]

Yep, I call shenanigans on the cpu index scores.

 

[FrancoisPiednoel]

Why not comparing them to the new 3.13Ghz Dual core Penryn? It could be interesting to add this to the mix.

Francois

 

[sunangel]

poor amd just can't catch a break. it still cost the same to make a tri- just as a quad-core. the best they can do is sell at cost or maybe 1-3% margin. i thank amd for their athlons, but competition between amd and intel is coming to an end... and their (amd) biggest downfall is releasing testing samples to any joe & jane that wanted to do a review.

 

[thefumigator]

"we were disappointed to see that Phenom support was virtually nonexistent."
What a laugh, I did the research myself. Its low, but no "Virtually non existant". 59 out of 162 mobos are Phenom ready
http://www.tomshardware.com/forum/248676-28-mobo-phenom-compatibility-research-here

The Tom's article that at first was called "Phenom fails compatibility test" then changed to "Most AM2 Motherboards Not Phenom Ready" should have been called "We fail to make a proper selection of AM2 mobos".

And the really interesting read that could have been "Phenom on old AM2" hasn't seen the light yet.

They choose 10 mobos, give me a break...

 

[tamalero]

poor amd just can't catch a break. it still cost the same to make a tri- just as a quad-core. the best they can do is sell at cost or maybe 1-3% margin. i thank amd for their athlons, but competition between amd and intel is coming to an end... and their (amd) biggest downfall is releasing testing samples to any joe & jane that wanted to do a review.

actually, you're wrong
if AMD will sell the toliman with a defective core.. then its LOSE AN AGENA (due of low yields ), BUT TO SELL IT AS TOLIMAN
thats better than sell nothing.

 

[jimmysmitty]

"we were disappointed to see that Phenom support was virtually nonexistent."
What a laugh, I did the research myself. Its low, but no "Virtually non existant". 59 out of 162 mobos are Phenom ready
http://www.tomshardware.com/forum/248676-28-mobo-phenom-compatibility-research-here

The Tom's article that at first was called "Phenom fails compatibility test" then changed to "Most AM2 Motherboards Not Phenom Ready" should have been called "We fail to make a proper selection of AM2 mobos".

And the really interesting read that could have been "Phenom on old AM2" hasn't seen the light yet.

They choose 10 mobos, give me a break...

I will agree that 10 mobos is short and your review was fantastic but the mobos they chose were the ones that are used most. Meaning that these mobos are the one that most people have. Kinda like the Asus P5K-E is one of the most widely used P35 mobo.

They did state that they will go back and check it again. Maybe there will be a better amount this time.

 

[jimmysmitty]

actually, you're wrong
if AMD will sell the toliman with a defective core.. then its LOSE AN AGENA (due of low yields ), BUT TO SELL IT AS TOLIMAN
thats better than sell nothing.

He wasn't saying to not sell them. He was saying that a tri core cost the same as a quad to make. And they cannot expect people to buy a tri core if the quad core is the same price.

Let me ask you this. If you had the option of a tri core @ 2.3GHz and a quad core @ 2.3GHz and both were the same price which would you go for?

Its like asking if a Ford GT and a Chevy Cavelier were the same price. If you select the latter you are one crazy....

mmmm..... Ford GT....

 

[computertech82]

I still visit Tom's Hardware regularly however the reviews over the past say 4-6 months really seem to have missed the boat. Is it just me or have others found the comments / questions answered by reviews to be poor?

I don't think there is much surprise in this article. Wow the tri core performs between the dual and quad in thread optimized code! Shocker! Wow, in semi optimized code the Tri and Quad perform very similar wow! I mean come on.

Here are some bigger questions IMO:
1) Cache. Will the Tri core have access to the same cache as the Quad? Thus in applications which require a large cache, will the increased cache hit rate make up for the loss of a core?

2) Overclocking! Let's face it if you don't overclock or are on a tight budget, AMD is likely your choice. Yeah it gives up some 200-600MHz to the C2D but for a given amount of cash, AMD has a solid position. Once you begin overclocking this all changes. The C2D simply blows AMD away (Or if you have several hundred to spend on the CPU). For example, I bought X2 5600 w/ 1Meg cache. $100 bucks delivered back before the abundance of C2D options. OC'd it to 2.95G. 690 chipset, AM2 socket. Current quads are a bad upgrade choice. Well this tri may be perfect! One less core, less heat, better OCing, yes yes?
*note* I did read the article, heat production was still up indicating that the core was likely running to some degree. However you didn't even attempt OCing!

Anyway, Tom's is no longer my first stop for "solid" reviews because these reviews simply don't answer the important questions.

Simply Disappointed.

I feel the same. Well said. Sad the REAL question/Answer wasn't done....how fast will the tri run at? The crappy slower then the 6400+ at 2.6ghz? Or will it be clocked much higher since core2 of the quads that is bad will be disabled? Or it's random cores being disabled?

 

[jimmysmitty]

I feel the same. Well said. Sad the REAL question/Answer wasn't done....how fast will the tri run at? The crappy slower then the 6400+ at 2.6ghz? Or will it be clocked much higher since core2 of the quads that is bad will be disabled? Or it's random cores being disabled?

We don't know which core will be disabled but once we get one can see. Also we cannot tell if it will clock higher as the higher clock speed deficiency seems to be due to the B2 stepping. Or it could just be the architecture/process aka SOI.

We will have to wait and see.

 

[thefumigator]

I will agree that 10 mobos is short and your review was fantastic but the mobos they chose were the ones that are used most. Meaning that these mobos are the one that most people have. Kinda like the Asus P5K-E is one of the most widely used P35 mobo.

They did state that they will go back and check it again. Maybe there will be a better amount this time.

Good point...!
However, the testing was to show how does Phenom performs on old AM2 mobos, which is today an unanswered question, despite good reviews you can see at newegg but nothing pro actually.
I still feel that 10 mobos is still too low for the "mobos used the most" category.

Just realized something:
I think asus and any manufacturer got a point on not releasing bios update to support phenom on certain motherboards, specially the most populars. Those choosing a phenom will have to buy another motherboard which is what a manufacturer wants, to sell more. Specially now that people seems to be moving to Intel. And, towards the most potential buyers, those owning the popular boards, which are actually those who don't mind changing a platform for a better one. While releasing a bios for a not-so-popular motherboard may be ok just for getting rid of the last units around.

Still, I think the opposite at the same time. Why would a popular manufacturer give its back to a potential buyer by privating the release of a new bios-for-phenom? That potential buyer would choose another manufacturer to protect his investment. So, I think in this case the lack of new bios release is just hardware limitation (lack of 4MBit bios as example).

 

[jimmysmitty]

Good point...!
However, the testing was to show how does Phenom performs on old AM2 mobos, which is today an unanswered question, despite good reviews you can see at newegg but nothing pro actually.
I still feel that 10 mobos is still too low for the "mobos used the most" category.

Just realized something:
I think asus and any manufacturer got a point on not releasing bios update to support phenom on certain motherboards, specially the most populars. Those choosing a phenom will have to buy another motherboard which is what a manufacturer wants, to sell more. Specially now that people seems to be moving to Intel. And, towards the most potential buyers, those owning the popular boards, which are actually those who don't mind changing a platform for a better one. While releasing a bios for a not-so-popular motherboard may be ok just for getting rid of the last units around.

Still, I think the opposite at the same time. Why would a popular manufacturer give its back to a potential buyer by privating the release of a new bios-for-phenom? That potential buyer would choose another manufacturer to protect his investment. So, I think in this case the lack of new bios release is just hardware limitation (lack of 4MBit bios as example).

Some of this is true. But Asus usually will update if they said it will support a CPU. My mobo is a P5K-E with the P35 chipset and it stated on the box it will support Intels 45nm CPUs. And so far the last BIOS added support for every Wolfdale dual core and every Yorkfeild quad core.

They could have not added it and forced us to buy a X38/X48 mobo but Asus is one of the better companies out there. Hell my old P4P800 supports every PGA478 Pentium 4 CPU out there. It was the LGA775's when BIOS updates stopped.

Now to compare AM2 vs AM2+ isn't that hard even with thought. The HT 3.0 vs HT 2.0 will make the major difference in memory sensitive applications as the HT link is mainly a memory bandwidth booster hence why HT gets better memory bandwidth than FSB does.

I doubt the differemce will be big enough in gaming and other big user apps to really upgrade to a AM2+. But once apps and games start to take advantage of the extra memory bandwidth it might prove to be worthwhile. But I don't expect that to happen really until the time of Nehalem and AM3.

But yea I would love to see a comparison of the two sockets just to see what the difference is.

 

[thomasxstewart]

ITS' One CPU DIE, NEVER DICE. One die with 4 cores. blah. OK, Next I'm glad to see you used NT6 Vista, yet VISTA ULTIMATE is O/S of choice, in 64 bit, with about 8 gb on main, if possible.

Why? because thing is so complex that it needs all O/S & memory it can get, to be fairly tested.Some synthetic shows that one to two nearly doubles, three not much more, probably out of harmony & four another fair step up. About what early mainframes & on have experienced with cores. REMEMBER. TO TEST BEST, USE ALL TOP STUFF THAT WILL FIT. I think you havn't given 4 core enough quality in O/s to truely know its faults, Especially futurely intresting is VISTA ONLY 3d mark VANTAGE , WHEN available.

Signed😛HYSICIAN THOMAS STEWART VON DRASHEK M.D.