Value and Conclusion
|7.4||AMD Bulldozer is with us for months now, more than enough time has passed for the product to set into its market slot. Bulldozer and AMD did not live up to expectations, nor did they deliver a competitive desktop product which they very much need right now. For the past few years, AMD has always been two steps behind Intel, but always found a way to stay competitive in certain market segments that kept AMD operational. The Bulldozer architecture and Zamebezi processors should have narrowed that gap between AMD and Intel, but instead of competing on par with Sandy Bridge, a year-old architecture, Bulldozer ended up only as a mid range product. This means AMD will have to continue the two step lag behind Intel for another product cycle at least, and to be honest it's no wonder they recently announced that they won't be competing with Intel in higher-end desktop processors. They simply can't, not in the near future at least.|
There are many reasons that contributed to the "downfall" of the Bulldozer architecture. The most logical and the most simple one - Zambezi was not introduced on clock speeds that it should have. The whole architecture was designed for supreme multi-threading performance, and single threaded tasks were supposed to be dealt with very high clocks speeds enabled by Turbo Core feature. But the base clock for the whole series was set too low, hurting the FX-8150 flagship's performance the most, and Turbo Core was little help from there. I would guess Global Foundries is having a hard time with the new 32 nm fab process. First of all, a huge spike in power consumption (as well as higher than expected stock power usage) is noticed when playing with core voltage, somewhat disappointing overclocking results (clock wise), and then there's word on the street that almost all of the new FX processors get sold out as soon as shops restock. It's either that they offer best value for money, or Global Foundries can't achieve greater volume production of non faulty chips. We lean towards the latter.
Applying the patch to fix a core schedule bug in Windows 7 also didn't improve performance much, and it is doubtful Windows 8 will change things for better. Either way, AMD should have given us a competitive new architecture some time ago, and now that they finally did, they are playing the "architecture of the future" card and wants us to wait for full benefits of Bulldozer's architecture. That's not fair to AMD fans and users. A launched product should offer its end user everything it can, here and now, not in a year from now, when the whole world could end even before that. A future proof architecture should be of concern only to the company and it's management, and while they can be somewhat pleased with Bulldozer, desktop users cannot.
Why pleased you might ask? Well, Bulldozer on paper is future proof in some ways, and it can aspire some optimistic thinking within AMD. Its modular design and versatility gives Bulldozer a wide range of possible applications with minimum changes to the processor configuration. AMD is not trying to hide it, they are using the very same eight core die in server and desktop configurations, and with the coming revisions of Bulldozer processor, it will spread to every market segment, from mobile through home setups to high end server configurations. Possibly it will never be best in something, but with the right tweaks it might be just good enough for everything. Intel will be rough on AMD in the coming years, but if production problems are solved and AMD can tweak Bulldozer the way Barcelona was morphed to Phenom II, there might be a chance for AMD to deliver in the low end and midrange segment, and make a firm stand there. They said that they won't be competing with Intel anymore in the high end desktop CPU market, but let's be honest, when was the last time they did? Not since the last generation of Athlon64 FX processors.
Building a high end processor takes a lot more R&D investment, more man-hours, and reality finally caught up with AMD, they can't compete with a giant like Intel on that level. If they can play it smart and in the future deliver strong mid range processors, not much will change, really. It's been like that for years and there are many users out there quite happy with their previous generation of "mid-range" Phenom II processor, so there's no reason to go after AMD with guns blazing for not launching a new über high end desktop processor. Maybe we should support them in what they know and do best - affordable and fast budget and mid-range processors, because we the users would pay the biggest price if we'd be left with one processor manufacturer.
As for the reviewed FX-8150, simply put, it's a long shot from the reputation of what the first generation of Athlon64 FX processor had. It would have been a great CPU if it weren't for Intel's Sandy Bridge, and Ivy Bridge is just around the corner. The price tag on FX-8150 is not that bad actually and it could be the only thing that keeps FX-8150 afloat. If you're into gaming: Inte'ls Core i5 2500K is the way to go, while those spending many hours doing some serious work should look to FX-8150 as an valid option. Core i7 2600K can get you best of both worlds for a few bucks more so FX-8150 needs to stay well below that price segment and fight it off with i5 2500K.