Quote:In the end, AMD's FX-8320E is an affordable quad-core processor that overclocks decently, but even if you pushed it to 5GHz it would struggle to match the slightly pricier Core i5-4430 and even the Core i3-4360 at times. Then after you take the power consumption figures into account, arguments for the FX-8320E begin to seem rather indefensible.

Free performance isn't exactly free if it comes at the expense of drawing more power and needing a more efficient cooler, so the humble Core i3-4360 makes more sense than the FX-8320E for budget users. There may be some great reasons to buy the FX-8320E, but we don't think it's the chip you want if you're after the best overall performance for the price.

Separately, we were surprised by how well the dual-core i3-4360 performed against the quad-core i5-4430. Although it runs 500-700MHz faster and has Hyper-Threading, we thought there would be a larger gap in our application tests as the i5-4430 actually has four cores. Given the price difference, we would also suggest the i3-4360 over the i5-4430.

Quote:First things first: at Hardware.Info we are no fans at all of significant product changes that are not reflected in a change in product codes and preferably product names. Even when products improve, as happens frequently with PSU manufacturers, this simply should not happen.

If Gigabyte had provided a different product and EAN code for the Rev 2.0 and Rev 3.0 boards, as well as updated product photos to the shops, there would not be much of a problem as far as we're concerned. In that case it would be clear to everyone that these are new, different products where the specifications and test results of earlier revisions do not necessarily apply to the newer revisions.

Now however a shop or end user can order a board based on reviews of the original and receive a product that is clearly different and, as we illustrated in this test, potentially inferior. And that is very questionable indeed.

The analysis of the B85M-D2V is shocking as Gigabyte has removed a feature from the original list of specifications, i.e. Dual BIOS. It's like ordering a car from a catalogue and discovering a missing right hand side mirror on delivery: not acceptable.

With the B85M-HD3 at least on paper there seems to be less of an issue: all the original Rev 1.0 specs are present in Rev 2.0. To play devil's advocate: Gigabyte doesn't promise anywhere that the board comes with a 4-phase power supply. Still, even if it passes muster legally, this is hardly the way to do business. Our test results moreover show that the Rev 2.0 has a potentially lower lifespan thanks to its lower number of mosfets and without any doubt will deliver lower performance than the first version in normal consumer applications (such as video encoding).

Quote:When I first saw the report of the massive loss of $364 million, and the revenue guide for Q1 2015, I was tasting that chili layer all over again. Part of me was shocked, part of me was dumbfounded, part of me was laughing at just how incompetent AMD was, never mind I had the stock. But things got better, but never to mango. Maybe more like carrot, or cabbage.

Quote:AMD has spoken repeatedly about preserving its research and development budget, and cutting elsewhere when it announced cuts. The facts do not bear this out. In Q3, AMD's R&D budget was $278m, in Q4 $238m. When looking at the cost reduction AMD enjoyed in Q4, this is a very large proportion of that. By contrast, administration costs only went from $150m to $144m. The way AMD portrayed this, you would have thought the proportions were reversed. This is of great concern not only because of the implications for future development, but also because it makes one question the veracity of AMD's management team.