AMD FX-8150 shows up in Passmark CPU benchmark

[Idontcare]

JFAMD was telling us not to believe everything we see on the internet, and tries to spin off most of these leaks as fakes (which incidentally had gotten to quite a number gullible OCN folks including the mods). Of course he never denies or point out at which are the fakes (or which ones are real).

At the risk of setting us off on a journey of semantics, this is a legitimate situation where being pedantic about what "real benchmarks" would even mean in terms of relevance to retail product performance on mature/validated platforms is relevant.

I submit that photoshopped/faked benches are just as relevant and useful as real/legitimate benches on non-retail stepping CPU's tested with beta BIOS revs and microcode updates that are still being hashed out in realtime.

That is not to say that it is not a worthwhile endeavor to seek out pre-release benches for the sake of fomenting discussion and pontification. But there is a difference between tracking the rate of progress in hitting necessary milestones on the journey towards polishing a product prior to its final retail debut versus attempting to divine and predict the performance of that yet-to-be-finalized retail product and platform.

For those who are aware of this and make use of this when engaging in discussions on pre-release benchmarks, there is no downside to the type of discussions we have here.

However, there is a group of folks who do not recognize the materially significant difference I highlight above, and unfortunately the reality for them and their efforts when intersecting threads such as these is that they, of their own volition, unwittingly sent themselves on a fool's errand when they committed to entering the thread.

And therein lies the source of the disconnect between expectations and reality, the source of frustration that comes to the surface in threads such as this throughout the CPU forum.

Recognition of this is key to understanding, and with understanding comes tolerance, and with tolerance comes community.

 

[Black96ws6]

Wasn't there a rumor that Intel was going to drop its prices when BD hits? They may have already started:

2500k: $149
http://www.microcenter.com/single_pr...uct_id=0354589

2600k: $279
http://www.microcenter.com/single_pr...uct_id=0354587

And if they drop a little lower (say a 2600k for ~$250 around Thanksgiving), how will BD compete?

 

[trollolo]

a 2500 for 149 would make my belly quite warm

 

[-Slacker-]

Wasn't there a rumor that Intel was going to drop its prices when BD hits? They may have already started:

2500k: $149
http://www.microcenter.com/single_pr...uct_id=0354589

2600k: $279
http://www.microcenter.com/single_pr...uct_id=0354587

And if they drop a little lower (say a 2600k for ~$250 around Thanksgiving), how will BD compete?

Could this be a hint that the bulldozers are serious competitors from a performance stand point?

 

[3DVagabond]

And if they drop a little lower (say a 2600k for ~$250 around Thanksgiving), how will BD compete?

Why do you believe that this will happen?

 

[LOL_Wut_Axel]

Could this be a hint that the bulldozers are serious competitors from a performance stand point?

No. It's from MicroCenter. They always have great deals and much lower prices than online retailers. The problem is that you only get to enjoy these deals if you have a store nearby.

 

[Gemmah]

That is not to say that it is not a worthwhile endeavor to seek out pre-release benches for the sake of fomenting discussion and pontification.

Fo' sure.

 

[inf64]

On a related note, all Interlagos submissions from sisoft website are gone . Gcache is our friend though.
Note that latest submission before takeoff was from 4 days ago and was highest one for 6282SE . 253Gops for integer and 632Mpix/s for Multimedia benchmark (32C 2.5Ghz/Turbo3.1Ghz-3.5Ghz 16C Opteron). This puts 8C Zambezi at 91Gops and 227Mpix/s(AVX ,brings 11% over legacy SSE in case of Zambezi;legacy SSE score should be around 204MPix/s).
Compare to 12C MC @2.5Ghz : 201Gops and 331Mpix/s and 1100T : 66Gops and 110Mpix/s. Massive improvement Vs previous generation with these non-retail samples.

 

[Idontcare]

On a related note, all Interlagos submissions from sisoft website are gone . Gcache is our friend though.
Note that latest submission before takeoff was from 4 days ago and was highest one for 6282SE . 253Gops for integer and 632Mpix/s for Multimedia benchmark (32C 2.5Ghz/Turbo3.1Ghz-3.5Ghz 16C Opteron). This puts 8C Zambezi at 91Gops and 227Mpix/s(AVX ,brings 11% over legacy SSE in case of Zambezi;legacy SSE score should be around 204MPix/s).
Compare to 12C MC @2.5Ghz : 201Gops and 331Mpix/s and 1100T : 66Gops and 110Mpix/s. Massive improvement Vs previous generation with these non-retail samples.

I don't understand why. Wasn't bulldozer launced in Q3? Its Q4 now, valid benches should be easy to find, not getting pulled such that we have to snoop google cache to find them.

Pulling benches is what AMD had to do when barcelona rolled out only to run into the TLB bug.

 

[inf64]

I don't understand why. Wasn't bulldozer launced in Q3? Its Q4 now, valid benches should be easy to find, not getting pulled such that we have to snoop google cache to find them.

Pulling benches is what AMD had to do when barcelona rolled out only to run into the TLB bug.

I think you got confused. Opteron 6200 haven't launched yet . It started shipping to partners 24 days ago,but launch is yet to happen. Supposedly it's very imminent which is probably why all those results got pulled off.
Good thing is that all these pre-release Opteron results are MagnyCours-crushing,especially in SIMD. How this translates to real world Zambezi vs Thuban clash is anyone's guess.

 

[BlueBlazer]

On a related note, all Interlagos submissions from sisoft website are gone . Gcache is our friend though.

Can't access other records and rankings? Looks like their database is down? :hmm:

I don't understand why. Wasn't bulldozer launced in Q3? Its Q4 now, valid benches should be easy to find, not getting pulled such that we have to snoop google cache to find them.

Pulling benches is what AMD had to do when barcelona rolled out only to run into the TLB bug.

Still waiting for those industry SPEC benchmarks also. Seems all quiet on the Bulldozer front? :hmm:

 

[Abwx]

I'm beginning to see the reason for the weirdness in Passmark results. One of the benchmarks rely on ancient x87 (floating point) which has been not been further developed/endorsed by Intel since Pentium4 (while it seems AMD is still strong with this legacy instructions).

Yet , Intel CPUs do better in Superpi , wich is an evidence
that they still have a robust X87 support....

This kind of post is just a desesperate move to downplay
whatever seems to make BD a good product...

 

[Black96ws6]

Once again I find myself going back and forth on this, first there was the Google cache hit review which said something similar to this, now there's this upcoming French magazine article, which, when translated, reads as such (slightly paraphrased by me):

"On the high-end, the FX processors based on the Bulldozer architecture were also a disappointment. While they are still generally more efficient than their predecessors and allow AMD to approach much of the performance of last generation Core i5 and i7 processors, unfortunately their performance remains below current expectations.

As we announced in a previous issue, they can sometimes compete with Sandy Bridge in the area of rough calculations, however the results in video games are very far behind. Only overclockers (and fanboys) will find them a great interest given their predisposition in this area.

The AMD FX-8120 model of the new Bulldozer series is the one that offers the best price / performance ratio. It is able to compete with the i5-2500K in most computing applications, even if it grossly lags behind in video games.

Nonetheless, it has substantial overclocking capabilities available at no additional cost. Compared with the old Phenom X4 980, this is a very good alternative.

Offered at a price slightly lower than the Core i7-2600K, the FX-8150 is currently the most powerful model of the new Bulldozer architecture from AMD. Unfortunately, at best, it matches the 2600k in some media processing applications while always falling behind in games. "

 

[cebalrai]

Once again I find myself going back and forth on this, first there was the Google cache hit review which said something similar to this, now there's this upcoming French magazine article, which, when translated, reads as such (slightly paraphrased by me):

"On the high-end, the FX processors based on the Bulldozer architecture were also a disappointment. While they are still generally more efficient than their predecessors and allow AMD to approach much of the performance of last generation Core i5 and i7 processors, unfortunately their performance remains below current expectations.

As we announced in a previous issue, they can sometimes compete with Sandy Bridge in the area of rough calculations, however the results in video games are very far behind. Only overclockers (and fanboys) will find them a great interest given their predisposition in this area.

The AMD FX-8120 model of the new Bulldozer series is the one that offers the best price / performance ratio. It is able to compete with the i5-2500K in most computing applications, even if it grossly lags behind in video games.

Nonetheless, it has substantial overclocking capabilities available at no additional cost. Compared with the old Phenom X4 980, this is a very good alternative.

Offered at a price slightly lower than the Core i7-2600K, the FX-8150 is currently the most powerful model of the new Bulldozer architecture from AMD. Unfortunately, at best, it matches the 2600k in some media processing applications while always falling behind in games. "

I think if Bulldozer is landing around Nehalem-level i5 and i7 performance while competing with SB in one or two areas then it will meet a lot of expectations. Not sure many folks expected it to meet or exceed SB.

As far as lesser gaming performance goes, as long as it's not going to keep games under 60 FPS for a few years then it meets my expectations. ()

 

[LOL_Wut_Axel]

Once again I find myself going back and forth on this, first there was the Google cache hit review which said something similar to this, now there's this upcoming French magazine article, which, when translated, reads as such (slightly paraphrased by me):

"On the high-end, the FX processors based on the Bulldozer architecture were also a disappointment. While they are still generally more efficient than their predecessors and allow AMD to approach much of the performance of last generation Core i5 and i7 processors, unfortunately their performance remains below current expectations.

As we announced in a previous issue, they can sometimes compete with Sandy Bridge in the area of rough calculations, however the results in video games are very far behind. Only overclockers (and fanboys) will find them a great interest given their predisposition in this area.

The AMD FX-8120 model of the new Bulldozer series is the one that offers the best price / performance ratio. It is able to compete with the i5-2500K in most computing applications, even if it grossly lags behind in video games.

Nonetheless, it has substantial overclocking capabilities available at no additional cost. Compared with the old Phenom X4 980, this is a very good alternative.

Offered at a price slightly lower than the Core i7-2600K, the FX-8150 is currently the most powerful model of the new Bulldozer architecture from AMD. Unfortunately, at best, it matches the 2600k in some media processing applications while always falling behind in games. "

This would go pretty much with what I said, but I wouldn't call it a complete failure. It'll more than probably be decent in multi-threaded, with the FX-8120 beating the 2500K slightly in multi-threaded and the FX-8150 only matching the 2600K there.

Given its projected price, I wouldn't call the FX-8120 a bad deal. It'll be a good chip for people with multi-threaded workloads, but for everything else Sandy Bridge would be much better. Most enthusiasts would go for Sandy Bridge given its balanced performance in different workloads.

 

[Idontcare]

"Offered at a price slightly lower than the Core i7-2600K, the FX-8150 is currently the most powerful model of the new Bulldozer architecture from AMD. Unfortunately, at best, it matches the 2600k in some media processing applications while always falling behind in games."

Considering the resource situation, morale situation, etc within AMD...if this is true then that's a real testament to AMD's design team :thumbsup:

Being able to hit 200mph in your $500k lambo is not quite the same accomplishment as managing to hit 180mph in your $30k Civic.

One is kinda expected, if not downright pedestrian given the budget, the other is rather unexpected and a bit inspiring.

 

[LOL_Wut_Axel]

Considering the resource situation, morale situation, etc within AMD...if this is true then that's a real testament to AMD's design team :thumbsup:

Being able to hit 200mph in your $500k lambo is not quite the same accomplishment as managing to hit 180mph in your $30k Civic.

One is kinda expected, if not downright pedestrian given the budget, the other is rather unexpected and a bit inspiring.

I don't get it... they basically said it matched it in multi-threaded and was much slower in everything else. Not much different than comparing the Phenom II X6 1100T and the Core i5-2500K. Overall, it's not a failure, but it's not a success either.

This isn't really exciting for anything other than people do this type of workload only.

 

[Idontcare]

This isn't really exciting for anything other than people do this type of workload only.

Which is what JFAMD has said is AMD's intention, to go after the markets in which Intel's mainstream products are neglecting or leaving room for improvement on the table. Not exactly niche, but not exactly broad-spectrum performance appeal either.

 

[notty22]

Here is something interesting.
AMD Bulldozer FX CPUs dated: October 12th. Shhh.

AMD has not been too well received in the upper end of CPUs for quite some time now. Once Intel started pushing for performance with their Conroe core, AMD was forced to stay competitive in the mainstream market to survive and that is pretty much where we have been for the last 5 years. Also returning after a nearly 5-year hiatus is the FX moniker for AMD’s flagship products. According to leak(s) from Microcenter that floated past our desks we should see a resurgence of at least one of those two on October 12th, 2011:
"AMD is launching their new AM3+ FX series processors on 10/12/11. We currently have a number of AM3+ compatible motherboards in stock. These motherboards will support the new AM3+ FX processors as well as legacy AM3 processors."

 

[BD231]

nm

 

[VirtualLarry]

Still sounds decent to me. I might get one. Looking at this motherboard. Looks pretty sick to me. Imagine that with an 8-core BullDozer, and four GTX460s, all crunching Distributed Computing.

Edit: I think that AMD is still ahead of Intel, not in pure computing horsepower, but as a platform company.

AMD's newest chipsets sport ALL SATA 6G ports, not this pathetic mish-mash of SATA 6G and SATA 3G like Intel. I think (could be wrong), that AMD's newest chipsets also support USB3 natively. Also, AMD's higher-end chipsets like the 890FX and 990FX, have mucho PCI-E lanes. That's one thing that you cannot get with Sandy Bridge, and that's a motherboard with four PCI-E x16 slots (that are at least PCI-E 2.0 x8 electrically). SB's on-chip PCI-E lanes simply aren't numerous enough, and even if they were, they cannot be split four ways, AFAIK.

 

[Idontcare]

Here is something interesting.
AMD Bulldozer FX CPUs dated: October 12th. Shhh.

AMD has not been too well received in the upper end of CPUs for quite some time now. Once Intel started pushing for performance with their Conroe core, AMD was forced to stay competitive in the mainstream market to survive and that is pretty much where we have been for the last 5 years. Also returning after a nearly 5-year hiatus is the FX moniker for AMD’s flagship products. According to leak(s) from Microcenter that floated past our desks we should see a resurgence of at least one of those two on October 12th, 2011:
"AMD is launching their new AM3+ FX series processors on 10/12/11. We currently have a number of AM3+ compatible motherboards in stock. These motherboards will support the new AM3+ FX processors as well as legacy AM3 processors."

This has got to be the longest "60 to 90 days" I've ever witnessed...

 

[LOL_Wut_Axel]

Which is what JFAMD has said is AMD's intention, to go after the markets in which Intel's mainstream products are neglecting or leaving room for improvement on the table. Not exactly niche, but not exactly broad-spectrum performance appeal either.

But it's not really being neglected... the 2500K and 2600K are both good at everything they do, including single and multi-threaded.

 

[zebrax2]

But it's not really being neglected... the 2500K and 2600K are both good at everything they do, including single and multi-threaded.

True but if all i do with my computer is rendering or encoding then i would rather go with a processor that is slower in general but faster for my workload rather than a processor that is faster in general but is slower in my workload. That is probably what AMD is trying to accomplish. Excel in at least one area rather than trying to catch up at all fronts at once

 

[LOL_Wut_Axel]

True but if all i do with my computer is rendering or encoding then i would rather go with a processor that is slower in general but faster for my workload rather than a processor that is faster in general but is slower in my workload. That is probably what AMD is trying to accomplish. Excel in at least one area rather than trying to catch up at all fronts at once

That's what I said earlier, but it doesn't mean Intel has neglected multi-threaded performance. Sandy Bridge will only be a bit slower there.

People that do this for a living would be better served picking up a Core i7-3930K anyway. Home users that have many multi-threaded workloads, then it could make sense if you neglect performance in other workloads.