A list of the specs for the new family of processors based on Piledrive has just been released! The family’s codename is Vishera, named after an island that forms a part of the Southern Islands to the south of Singapore. The new family has the product names as expected, but a lot has changed in the meantime while Bulldozer has been fighting it out for market relevance.
Hit the jump for more info.
I’ll admit that I’ve been waiting a while for something concrete to pop up for AMD’s next batch of chips. Since the release of Bulldozer the company has been focusing its efforts on their APU lineup and working on the mobile Piledriver cores, dubbed Trinity, which NAG recently reviewed in their September edition in a HP Pavilion DM6 Ultrabook alternative. While the updates Trinity brings to the notebook segment are noteworthy, its the desktop that needs a significant performance overhaul as Intel’s Ivy Bridge CPUs begin to chew into more market share. And it seems like the company has been doing a lot of work judging by the changes in the specs sheet.
First up though, is the APU lineup. The moniker has changed to the Ax-5000 series with dual and quad cores available. The K-versions of the APUs won’t be released until much later but for now you’ll have to make do with the top-of-the range A10-5700, boasting a Radeon HD7660D integrated GPU, a 3.4GHz clock speed with 4GHz available for one or two cores in boost mode and, surprise surprise, a 65W TDP. If you’re confused, that’s because its the same TDP as Intel’s low-end quad-core chips and even some of their dual-core offerings. It looks like Piledriver brought the expected power savings to the APU lineup as promised. Is it possible these are 22nm chips?
The rest of the APU lineup differentiates by dropping clock speeds and putting in less shaders for the GPU component, dropping down to the low-power E2-1800, designed for use in low-power desktops and SFF nettops that use ITX motherboards. Because it seems that AMD will keep the Athlon II brand for now, this won’t be the most affordable chip on the market but it will hopefully be the first one able to dupe it out with Intel’s Celeron G540 with its integrated HD2000 graphics chip.
It also looks like AMD is employing the same TDP tactics as Intel in order to not confuse customers. All of the APU chips feature a 65W TDP but its unlikely that they all go up to that limit. I expect the E2-1800 to max out with a 35W TDP, the higher-clocked A4 and A6 chips with a 45W TDP and the A8 and A10 quad-cores pushing the 65W envelope. All you really have to do is find a board that supports AM3+ 65W chips and you’re good to go. If this is true, its the kind of simplicity AMD needs for their product desperately.
The FX family offers no graphical component but as you can see in the specs sheet, there are some things GPU-related that the chip will offer, but more on that later. Of note is the lowered 95W TDP for the quad and six-core chips, better matching their Intel Core i5 counterparts at 77W. The FX-8350 tops out at 125W but its not as bad considering it is their highest-SKU chip. At 125W its unlikely that AMD has made the transition to 22nm for their FX lineup, but the APU is entirely possible.
Its also evident that some headroom has been created thanks to the lower power requirements of the Vishera family. The FX-4320 in particular operates at stock speeds of 4GHz – with decent air or, it could be possible that people will break the 5GHz barrier with ease. Got a water cooler on hand? 6GHz is my estimation as to where most chips will top out. For the overclocker, the new chips will be a great challenge and finding their limits using LN2 will surely be fun.
It also looks like AMD’s technology engineers have been hard at work catching up to Intel. The company is really pushing the dual-GPU benefit of running an APU with a discrete card up to a certain size but I wonder if anyone’s really going to use that. One new thing is AMD’s Wireless Display technology. I haven’t seen anything about that yet but its inclusion is telling that AMD is playing catch-up with Intel in areas where consumers are likely to appreciate the better value from the red team’s chips. I see they’ve also listed App Acceleration as a feature, likely due to the company’s push to have OpenCL accepted as an acceleration standard, rather than relying on their ATI Stream or Nvidia’s CUDA offering. For the most part though, the technology that Piledriver brings to the desktop doesn’t change much.
But where things do change, its for the better. The IPC improvements inside the mobile chips are enough to secure Trinity’s place as a viable alternative to the Core i5 and i7 mobile processors with the HD4000 GPU, especially where gaming is concerned. With a lower TDP and faster clock speeds. this is going to be an interesting time for upgraders with desktops in need of a refresh. Lets hope AMD has the chops to compete once again.