AMD FX-8150 Review: From Bulldozer To Zambezi To FX

[tigr1982]

Indeed, much much better performance was expected from BD. I was an AMD focused PC buyer since 2005, at AMD "golden age", when I purchased AMD Turion-based laptop. That CPU was actually better than the corresponding Intel competitor at the moment - Pentium M Dothan, as probaly some people remember.

We know the rest of the story since then till now...

But the released BD-based product in its current state seems to be barely concurrent at all on the desktop market. Presumably, its popularity will be much lower, than in case of previous Phenom II lineup...

 

[DjEaZy]

... i gonna buy it... because i have the ecosystem in my box... it's a simple chip switch... and i found alternatives to single trended and/or 32bit applications, that will use all 8 cores and in 64Bit...

 

[tigr1982]

Saying "barely concurrent", I actually meant "barely competitive".

 

[Indy-Go]

I predict that AMD will be Steam-Rolled by Intel and become assimilated into its ranks in the next 5 years.

 

[Guest]

The thing here is lots of people are listening to the hype. Fact of the matter, it is an all-new architecture. Accusations being thrown around won't make the processor any better, or worse. Why is it when a new product,,,,yes it is NEW, fails initially to live up to other's expectations, it's a fail? AMD knew that it would take a linux/unix fix along with a Microsoft fix. Something to do with randomizing core addresses, or some such thing. Also, a new stepping is allready coming. As far as AMD losing in the market? Hm, LLano, Bobcat, anyone? And Trinity is coming. I like the performance of intel's processors, don't get me wrong, it's just that I like the idea of supporting the underdog, and loyalty to the first processor I ever used in 1972.

 

[dragonsqrrl]

Interesting interview with a former AMD engineer on Bulldozer's efficiency and performance problems:

http://www.guru3d.com/news/former-amd-engineer-talks-about-bulldozer-fiasco/

 

[Draven35]

[citation][nom]rikaroo[/nom]The thing here is lots of people are listening to the hype. Fact of the matter, it is an all-new architecture. *snip* Why is it when a new product,,,,yes it is NEW, fails initially to live up to other's expectations, it's a fail? AMD knew that it would take a linux/unix fix along with a Microsoft fix. Something to do with randomizing core addresses, or some such thing.[/citation]

So because its a new architecture, we should give them a pass and pretend there isn't problems? When the Core i7 shipped, was there 'performance problems' because it was 'a new architecture'?

No, it blew the doors off of everything else on the desktop.

If AMD 'knew it was going to take a fix' then they should have had the fix ready when they shipped the units to reviewers instead of doing what is likely to be irreparable harm to their reputation. "NEW AMD PROCESSOR SUX" is the equivalent of 'front page news' in our world, while "amd issues patch that fixes issues with new processors' is likely to end up on page 12.

 

[Why_Me]

That sounds pretty good - the part about IB running on all LGA 1155 (would be good if you could post a source/proof for such relieving news - not that I am planning to upgrade to IB, but it'd be nice to know that I can, should I ever need more CPU power). But I don't agree about SB-E overtaking the multi-core market... FX-8150 is $240 and SB-E will be $1000; there will be a market for SB-E but not everyone will be able to afford the extra $760 for their workstation.

All AMD needs is to drop FX-8150 to $200-210...

http://www.asrock.com/microsite/PCIe3/overview.html

 

[Why_Me]

So because its a new architecture, we should give them a pass and pretend there isn't problems? When the Core i7 shipped, was there 'performance problems' because it was 'a new architecture'?

No, it blew the doors off of everything else on the desktop.

If AMD 'knew it was going to take a fix' then they should have had the fix ready when they shipped the units to reviewers instead of doing what is likely to be irreparable harm to their reputation. "NEW AMD PROCESSOR SUX" is the equivalent of 'front page news' in our world, while "amd issues patch that fixes issues with new processors' is likely to end up on page 12.

C2D, socket 1336, and socket 1155 all exceeded expectations.

 

[Guest]

Ahhhaaaaa !!!
Bye bye Intel and a big Welcome to u AMD the best Processor Performance/Price it is unbeatable. Beware... He is heavy and full leaded called BULLDOZERRRRRRRRRRRRRRRR !!! .

 

[dragonsqrrl]

[citation][nom]D0zEr[/nom]Ahhhaaaaa !!! Bye bye Intel and a big Welcome to u AMD the best Processor Performance/Price it is unbeatable. Beware... He is heavy and full leaded called BULLDOZERRRRRRRRRRRRRRRR !!! .[/citation]
...troll.

 

[LORD_ORION]

Not surprising that it only does well in multi-threaded heavy apps.

Price is too high, and it is not a gaming CPU, much like their AMD X6.

Considering it will beat a 2500K by 25% in thread heavy apps, and matches the 2600Ks in the same apps... there is something to look forward to for AMD. This architecture will only get better as badly coded software dies off.

Also, considering this is how much the Phenom IIs were 1st priced, we can expect to see this at $150 next year?

 

[pcito]

It seems like the only thing this CPU is really good at is heavily threaded applications, such as most media encoding, etc. It would have been useful for toms to focus on that area a bit more, so that anyone with that type of workload could determine if this CPU makes any real sense.

I would like to see it benchmarked with VMWare workstation, for example with a few virtual machines running vs a sandy bridge I5.

Regardless, this CPU is way overpriced for its performance level. The i5 seems to outperform it in almost all other usage scenarios except a few heavily multithreaded ones. Even in those cases it holds its own.

AMD really needs to tweak this one fast, and drop its price as well. From what the benchmarks show, this should be in the $165 range, down there with the Phenom II x6 cpus.

 

[someonewhoknowsalittle]

If AMD wants to sell any of these chips to any of the major desktop computer manufacturers like Dell, HP, Acer, etc., they will have to drop the price of this chip to them to not more than $149. Any more than that and they will not sell more than a handful of these chips to those companies. We know not more than a handful of gamers will buy it also. No companies will buy more than a few of them for their servers because they run too hot and use too much expensive electricity.

Let's face it, this is not an 8 core chip. It's a 4 core chip with 2 integer modules in each core. That's it. That's all. The L2 and L3 latencies are too high. Branch prediction and penalties for wrong prediction are too high. The time to switch the streams from integer to fpu modules is too high and encountered too frequently.

This cpu is a dud. The performance at rated speed is too low and the performance per watt and IPC are way too low. This is bloatware at its worst.

The projected 10%-15% increases in performance per year over the next 3 years are in reality a 0% increase, because they just indicate what we already know, that anyone can overclock this 3.6GHz cpu to 4.6GHz on air, so AMD promises to crank out higher rated speed cpus in the next 3 years that are in reality not any better or faster than this cpu is now.

This is the end for AMD as a cpu designer. They've already sold off their fabs. They can't design cpu architecture for beans. All that's left is their gpu unit for design and production of video cards and gpus for super computers. AMD? It was nice to have known you. And thanks for the Athlon XP's I had. They were hot!

 

[nosnitrousx]

How about Tom´s doing a review of the FX-4000 and -6000 series??

Could it be possible this being like the Phenom X4 and X6 difference? Where few cores actually perform "better"....gamewise though.

And pleaaase....with an ATI/AMD card.
Don´t flame/tag me as a fanboy...but to get a chance (at least in that scenario) to see one brightside on this so-expected-but-far-from-convincent release.

 

[ChromeTusk]

I predict that AMD will be Steam-Rolled by Intel and become assimilated into its ranks in the next 5 years.

Although this could happen, that action could potentially put Intel into a monopoly situation. Intel may just let AMD try to keep itself afloat as "competition" and avoid anti-trust/anti-monopoly laws. AMD still has revenue coming from low-end chips and ATI.

Interesting interview with a former AMD engineer on Bulldozer's efficiency and performance problems:

http://www.guru3d.com/news/former-amd-engineer-talks-about-bulldozer-fiasco/

I read the same article (and posted a comment in a different thread). If the information is true, automated designs increased the cost by 20% versus hand-crafted designs ($245 divided by 120% = 204.17) and were also 20% slower. Imagine how the reviews would look with a 10-20% increase in performance.

 

[littleleo]

Complete FAIL on AMDs part, they let the entire industry and all the consumers down. I hope pile driver isn't just a pile.

 

[technoholic]

Has anyone heard a "registry fix" for BD in win7? there are some rumors on the net about this.

 

[Guest]

Good to see AMD is taking the Duke Nukem Forever approach with their products..oh wait, that's bad......

 

[amk-aka-Phantom]

Has anyone heard a "registry fix" for BD in win7? there are some rumors on the net about this.

Right, blame the registry... :lol:

 

[nikorr]

It doesn't have that punch I thought it would have : )

 

[Guest]

what ?
hahaha
where AMD FX-FX8812 benchmarking ?
no FX-8150
no mistake PLZ
no ad PLZ

 

[redraider89]

Tom said they tested a "wide range of processors, graphics cards, and in-game settings" for World of Warcraft and then acts like they did the same thing here and then said AMD still struggled. Still? Like you were doing the same thing and got the same results? All I see is that you used a single nvidia card, HARDLY A "WIDE RANGE of...graphics cards"!

 

[redraider89]

[citation][nom]nikorr[/nom]It doesn't have that punch I thought it would have : )[/citation]

If you don't know, maybe you out to keep quiet.

 

[techpops]

Some time has passed, the dust has settled and now I'm ready to say what I really think, But you don't care about that. If you're reading this you already ploughed through a tonne of comments, probably read multiple reviews too. You know how it is, there are no more answers to find, nothings going to change in the short term, it's time to move on.