Around this time last year, AMD's new bulldozer-based FX series was launched for light-eyed system manufacturers who expected the new architecture to question Intel's increasingly comfortable position in the high-end processor market. Unfortunately, bulldozer was not all that mattered. The performance lagged behind the then 9-month-old Sandy Bridge processors and in some cases could not even exceed the Phenom II series.
After the mediocre reception from Bulldozer, AMD insisted that the new architecture was still young and would serve as a "solid building block" for the FX series. Although hotfixes like one that addressed SMT inefficiency have slightly increased bulldozer performance, little has changed on the AMD-FX series over the past year – until now, at least, today, when the FX Offers of the second generation of the company were introduced.
AMD updates its desktop processors with Piledriver, an enhanced version of Bulldozer that focuses on improving instructions per clock and frequency. We saw this earlier this month when we tested the new Trinity APUs with Piledriver technology. In other words, instead of a major overhaul, Piledriver picks up where Bulldozer left off, which can disappoint those who wanted AMD to give up the architecture.
Perhaps the most controversial topic in last year's FX series was the new design of the integer core / module. Bulldozer's floating point unit has been completely redesigned and improved to support many new instructions while allowing resource sharing between cores. Two 128-bit FMACs were shared per module, allowing two 128-bit commands per core or one 256-bit command per dual-core module.
Although each module appears as two cores in the operating system, they do not have as many resources as conventional AMD cores. While AMD claimed its FX-8150 was the first true 8-core desktop processor, it wasn't in reality and its performance reflected that. Piledriver does not change this architecture significantly here either, which means that the module design is still present in today's processors of the Vishera FX series.
As we saw recently when testing the A10-5800K Trinity APU, Piledriver offered a significant improvement in power consumption, which allowed AMD to increase the chip's clock frequency by 31% over the older A8-3850. As an example of what this means for the new processors in the FX series, AMD has indicated that the chips will offer up to 20 to 30% more performance in digital media workloads – a significant boost, if accurate.
Parts released today include the FX-8350, FX-8320, FX-6300 and FX-4320 described above. Although AMD releases four new FX chips, we only test the FX-8350 and FX-6300. Compared to its predecessors, the FX-8350 is clocked slightly more than 10% higher than the FX-8150, although its maximum turbo core frequency is the same at 4.2 GHz, while the six-core FX-6300 is clocked 6% higher than the FX-6100.
Apart from that, the FX-4300 is interestingly clocked at the same speed of 3.8 GHz as the older FX-4150. While the older model had a full 8MB L3 cache, the FX-4300's cache was halved to 4MB. It is also noteworthy that all new Vishera FX models are supported by the existing chipsets of the Socket AM3 + and 9xx series, have Turbo Core 3.0 technology and use the same two-channel DDR3 memory interface.