Quote

It's good 3 months now since the notebook flavour of AMD 'Trinity' Fusion APUs arrived, including the high end K10 series. Their GPU performance is pretty darn well good for the die and process used, however the CPU core performance  was kind of just OK, still somewhat lower than the Sandy / Ivy Bridge parts, including those in the same price band. Why? Well, it seems the "Piledriver" core design, that Trinity pioneered in an actual product, isn't that much of an advantage over the faltering predecessor, Bulldozer, in per-core performance.

For the mobile market, that can be still forgiven if there's power efficiency to throw in, which Trinity does seem to have - the laptops do provide decent 6 - 10 hour battery life. If AMD was able to tweak in the process and the design further, and enable a 17 - 20 W 'Ultrathin' Trinity with all 4 CPU cores i.e. two dual-core blocks plus full GPU, integrated, it'd be a very decent competitor to Intel's Ultrabook designs.
However, the desktop PC market has more of performance focus, especially in the CPU department - since discrete GPUs are more common on desktops, anyway. Now, there's no 'Vishera', the Piledriver-based 4-core block (8-core in AMD speak or 4-core multithreaded in Microsoft Task Manager) replacement for the infamous  'Zambezi' Bulldozer desktop out yet, and the desktop Trinity flavours, some of which our guys have in samples, are also postponed without clear, if any, release date. Only the "Abu Dhabi' dual die lower clock 16-core total server chip should be out next month. Why?

Well, one reason could be that, maybe, there is not enough improvement in per core performance from "Bulldozer" to "Piledriver" to justify the new product release, as it wouldn't improve the competitive position against Intel in the desktop segment at all. Whether it is 'instruction per clock' IPC or pure clock frequency, both have to contribute to the end performance to justify the product.

The other reason, linked to the first one, can be - likely - that AMD now has to place its hopes to the upcoming 'Steamroller' core early next year, coupled with the 28 nm process. I guess we can accept that and give the AMD CPU team more time to do that migration, as long as it is real soon now. Else, another excuse this time next year, to maybe wait for 'Excavator' to fix things, would be really unacceptable...

Quote (NinjaSushi2 @ Aug 20 2012 08:42pm)

tl;dr thread.

What are we arguing about?

Quote (dolarsignzeroxeighty @ 20 Aug 2012 19:44)

nothing, i'm just posting about news i receive from google alerts about piledriver.

Quote (NinjaSushi2 @ Aug 20 2012 04:45pm)

Damn... PILE DRIVE THAT ASS HOLE! mmhmm

Quote

A day ago VR-Zone reported that AMD had cancelled their desktop Piledriver processors codenamed Vishera however latest reports suggest that Piledriver would launch but would be the last of performance CPU’s from AMD.

OBR’s the news breaker who acquired some internal documents of AMD which show that the company would discontinue its performance processors for desktop market after launch of Vishera. From them, AMD would focus more towards the development of its APU’s (Accelerated Processing Units) which are doing quite well against competitor Intel.

AMD’s performance offering such as Bulldozer didn’t do well against Intel’s Sandy Bridge and same is being said about the upcoming Piledriver architecture which doesn’t has sufficient IPC improvements to tackle Haswell or even the recently launched Ivy Bridge processors. The last of performance processors we would see from AMD would be the FX 8200/8300 series after which the series would be discontinued.

We still can’t confirm how legitimate the source is since there has been no word on this by AMD themselves but for now it seems as if Steamroller and Excavator are just names to be remembered in the performance market, their core architecture would still be used in APU development. On AMD’s APU’s forefront, we would be looking at Trinity and later on Kaveri which would be the 3rd generation APU based on 28nm Richland architecture featuring new HSA improvements and Sea Islands IGP chipsets.

Quote

A consumer platform guide of AMD has been leaked which confirms core frequencies of three upcoming Vishera FX processors including the flagship FX-8350 which turbos upto 4.2Ghz.

The guide details specifications of three upcoming Vishera FX processors which would be featuring the Piledriver x86 core architecture. Leading the pack is the Octa-Core FX-8350 running at a stock speed of 4GHz with Turbo Core frequency of 4.2GHz on all cores. The CPU has a rated TDP of 125W and features 8MB of L3 Cache.

The Hexa-Core “FX-6300″ operates at 3.5GHz stock and 4.1GHz Turbo Core speed with a rated TDP of 95W and featuring 6MB L3 Cache. There’s a entry level Quad Core model too known as the “FX-4320″ which runs at 4.0GHz stock and 4.2GHz Turbo on all cores, featuring 95W TDP and 4MB L3 Cache.

Price wise, the new Vishera processors would cost lower than their Intel Ivy Bridge counterparts. The guide also details a few Trinity APU’s which we already covered back here. You can also check out a brief performance measurement of the FX-8350 processor here. Launch of Piledriver is expected in Q4 2012.

In other news, there are chances we won’t be looking at a successor to Piledriver since rumors suggest it is the last of performance offering from AMD. AMD has yet to comment on that.

Quote (Battleship @ Jun 28 2012 10:03am)

Nope im not that excited at all tbh.

Quote

Advanced Micro Devices, the second largest x86-64 chip designer, is between a rock and a hard place. Although it recently went “fabless” by giving up its ownership stake in GlobalFoundries to shed some debt, the x86 underdog is surrounded by the booming (primarily ARM-based) mobile market on one side, and chip juggernaut Intel — which has vastly more resources to devote to pursuing new (smaller) process nodes, and CPU technology — on the other. Further, both ARM and Intel are moving towards a point of convergence as Intel’s x86-based CPUs get smaller and more efficient, and ARM’s SoC (system-on-a-chip) processors continue to get faster and add functionality. This raises the question of where exactly that leaves AMD.

At the Hot Chips conference today, AMD CTO Mark Papermaster provided an answer as he spoke about the future and continued relevance of AMD. In many respects, the keynote reiterated the company’s intention to bet its future on (the success of) APUs – an idea first revealed at the AMD Fusion Developer Summit (AFDS) earlier this year. The overarching theme at AFDS was that AMD was adapting to the current state of affairs, and would be — to a certain extent — conceding the high-end desktop market to Intel. Rather than compete directly against Intel for the absolute fastest performing chip crown, AMD has decided to focus on the areas where it has the biggest advantage over the competition. Primarily, the company’s fortés include graphics processing and heterogeneous computing architectures.

AMD talked heavily of the graphics processing portion and unified address space in its next generation Kaveri APU at AFDS. However, in today’s keynote, the chip designer focused on the CPU side of things by talking about the Steamroller x86-64 CPU architecture. Steamroller is the successor to Piledriver, and Piledriver is the name of the x86-64 CPU architecture used in the (upcoming) Vishera desktop processors and Trinity APUs. Steamroller will further improve upon Piledriver’s refinements to the original Bulldozer architecture, making it a much more efficient overall design.

Bulldozer vs. Steamroller architecture size

AMD has done away with hand-drawn processor schematics (which define how all the internals are laid out and interconnected) in favor of using a computer-assembled design (a more automated design approach). Using a high density library, the company was able to achieve a claimed 30% reduction in power draw and die area in the final chip without reducing the number of logic blocks. The Steamroller cores further boast reduced latency, increased bandwidth, instruction fetching and pipeline optimizations, inter-process communication tweaks, power efficiency improvements, and a dynamically-sized (shared) L2 cache. In simple terms, Steamroller is Piledriver 2.0 — a slightly tweaked Piledriver architecture with processing and power efficiency in mind.

On the instruction-fetching front Steamroller has been heavily refined, and should reap some decent performance gains due to its ability to keep the CPU cores (modules) fed with data. It features a 20% reduction to branch prediction errors, and 30% fewer cache misses, for example. Further, the floating point (FP) scheduler in Steamroller continues to be shared between two CPU modules (cores). It features two 128-bit FMAC (fuse-multiply-add capability) units, but it has only one MMX unit — versus Piledriver’s two MMX units. AMD has stated that this change is in response to changing computing situations, and by removing the MMX units they can reclaim die space without too large of a performance hit.

AMD Steamroller Instruction Fetching Improvements

When Steamroller CPU cores are used in an AMD APU specifically, they are able to realize further power-saving features. Namely, the chip is able to dynamically adjust clockspeeds (and as a result power usage) depending on the current workload. If the CPU is sitting mostly idle and only the GPU is under load (watching a hardware-accelerated H.264 encoded movie for example), the majority of available power can be allocated to the GPU along with increasing the clockspeed (if necessary) up to the rated TDP thanks to AMD’s Turbo Core technology (no word yet on default or boost clockspeeds).

AMD has recognized the importance of, and need for, power efficient processors as millions of mobile devices are sold each year and people are increasingly turning to the internet to store and process their data — where servers eating up electricity and requiring extensive cooling can be major operating expenses. It remains to be seen whether or not AMD made the smart bet in basing the company’s future on APU technology, but the Steamroller cores bring some promising improvements to the table that may just help the company realize its goal of bringing HSA to the here-and-now (and keeping the company competitive as a result).

Quote

This fall AMD will release large number of new desktop processors, built on the latest Piledriver microarchitecture. With the exception of a couple of chips, details of upcoming lineup were known well ahead of the launch. Yesterday Expreview published a slide, titled "2012 Consumer Platform Quick Reference Guide for Desktops", that confirmed clock frequencies of upcoming "Piledriver" products, and shed some light on supported technologies.

The slide lists three high-performance FX processors with "Vishera" core: quad-core FX-4320, six-core FX-6300, and eight-core FX 8350. Details of FX-4320 and FX-8350 were known for quite some time. The CPUs run at 4 GHz, and Turbo Core technology can boost clock speed as high as 4.2 GHz when half of the cores are idle. The FX-6300 will operate at 3.5 GHz, and its maximum Turbo Core speed is 4.1 GHz. Quad- and six-core FX microprocessors will be rated at 95 Watt TDP, 30 Watt lower than the FX-8350.

Four Accelerated Processing Units, included on the slide, are quad-core A8-5500 and A10-5700, and dual-core A4-5300 and A6-5400. All of these are 65 Watt chips. We reported on A10, A8 and A6 APUs in the past, and the slide only confirms previously published specs. Less known from these 4 SKUs is A4-5300. This part has 2 cores, clocked at 3.4 GHz. The processor supports Turbo Core, which can raise the speed by another 300 MHz. The A4-5300 comes with integrated Radeon HD7480 GPU, which has 128 shader cores. The A4-5300 is the only "Piledriver" based desktop processor, that will not support a number of AMD features, such as accelerated video conversion and playback, dual-graphics, and Eyefinity technology.

The slide states that all above mentioned Piledriver processors will be available in late Q3 2012. It's important to note that the guide does not show several "K" APUs, that will be released at the same time frame. The summary of details of forthcoming FX CPUs and A-Series APUs, referenced by the document, is provided below:

Model Cores Frequency Turbo Frequency L2 cache L3 cache Socket Type TDP
A10-5700 4 3.4 GHz 4 GHz 4 MB N/A Socket FM2 65 Watt
A4-5300 2 3.4 GHz 3.7 GHz 1 MB N/A Socket FM2 65 Watt
A6-5400K 2 3.6 GHz 3.8 GHz 1 MB N/A Socket FM2 65 Watt
A8-5500 4 3.2 GHz 3.7 GHz 4 MB N/A Socket FM2 65 Watt
FX-4320 4 4 GHz 4.2 GHz 4 MB 4 MB Socket AM3+ 95 Watt
FX-6300 6 3.5 GHz 4.1 GHz 6 MB 8 MB Socket AM3+ 95 Watt
FX-8350 8 4 GHz 4.2 GHz 8 MB 8 MB Socket AM3+ 125 Watt