AMD Apologize for Incorrect Bulldozer Transistor Count

The company initially told everyone in the press, that Bulldozer/Zambezi chips were made up of two billion transistors. It turns out that the transistor count number AMD supplied to the press was incorrect. Today AMD made following statements:

We always strive to be proactive and open. Last week, AMD confirmed the transistor count in the AMD FX CPU line-up at 1.2 billion, a correction from the earlier count of 2 billion. The earlier figure of 2 billion transistors was unfortunately shared in error. This correction is not the result of a new revision to the Bulldozer design. The correct count of 1.2 billion applies to all recently introduced 8-core AMD processors that are based on the new Bulldozer core – AMD FX family of desktop CPUs and AMD Opteron™ family of server and HPC processors. We apologize for the confusion.

Bulldozer Lack of Fine Tuning?

Why performance of the long-awaited Bulldozer was below expectations is not only because it was late, but because AMD had adopted design techniques that did not allow it to tweak performance. At least that is according to Cliff A. Maier, an AMD engineer who left the company several years ago.

The management decided to abandon practice of hand-crafting various performance-critical parts of its chips and rely completely on automatic tools. While usage of tools that automatically implement certain technologies into silicon speeds up the design process, they cannot ensure maximum performance and efficiency.

Apparently, automatically-generated designs are 20% bigger and 20% slower than hand-crafted designs, which results in increased transistor count, die space, cost and power efficiency.

"I had been in charge of our design flow in the years before I left, and I had tested these tools by asking the companies who sold them to design blocks (adders, multipliers, etc.) using their tools. I let them take as long as they wanted. They always came back to me with designs that were 20% bigger, and 20% slower than our hand-crafted designs, and which suffered from electro-migration and other problems," the former AMD engineer said.

AMD publicly said that each Bulldozer dual-core CPU module with 2MB unified L2 cache contains 213 million transistors and is 30.9mm2 large. By contrast, die size of one processing engine of Llano processor (11-layer 32nm SOI, K10.5+ micro-architecture) is 9.69mm2 (without L2 cache).

As a result, all four CPU modules with L2 cache within Zambezi/Orochi processor consist of 852 million of transistors and take 123.6mm2 of die space. Assuming that 8MB of L3 cache (6 bits per cell) consist of 405 million of transistors, it leaves around whopping 800 million of transistors to various input/output interfaces, dual-channel DDR3 memory controller as well as various logic and routing inside the chip.

800 million of transistors - which take up a lot of die space - is an incredibly high number for various I/O, memory, logic, etc. For example, Intel's Core i-series "Sandy Bridge" quad-core chip with integrated graphics consists of 995 million.

As a consequence of inefficient design and relatively low performance, AMD has to sell its eight-core FX series processors (315mm2 die size) for up to $245 in 1000-unit quantities. By contrast, Intel sells hand-crafted Core i-series "Sandy Bridge" quad-core chips (216mm2 die size) for up to $317 in 1000-unit quantities. Given the fact that both microprocessors are made using 32nm process technology [and thus have comparable per-transistor/per square mm die cost], the Intel one carries much better profit margin than AMD's microprocessor.

For more information, read here:
Insideris : AMD Spreads Propaganda, Ex-Employee Speaks Out
Mr. Maier posting on MacRumor Forum

Bulldozer for Desktop is Finally Out

Zambezi is the fourth-longest river in Africa (the one with the beautiful Victoria Falls), and also the codename for the new FX series of "Bulldozer" desktop enthusiast CPUs from AMD. After months of speculation and spin about the launch date and performance numbers, the new processors have finally hit the online shops. Yes, we saw the launch today, and the long awaited Bulldozer is finally out.


That new AMD processors based on Bulldozer microarchitecture are one of the most highly anticipated products not only of the year, but at least of the last five years. There are several reasons for that as well as for the fact that AMD products have so many fans. Remember the times when AMD processors were better than Intel ones in all aspects? Some users like AMD products for the balanced combination of price and performance that they have to offer. And some may have been carried away by the passion, with which they talked about the advantages of the new microarchitecture they have been working on. All this combined with the years of waiting for the new Bulldozer processor generation produced high interest and excitement.

Here we bring you some reviews of the top of the line FX-8150:
Tech Report
X-Bit Labs
Guru3D
Bit-Tech
AnandTech

Trinity & AMD Mobile Roadmap 2012

According to the latest AMD Mobile Roadmap, Trinity scheduled to arrive in 2012, initially will co-exist with Llano. Trinity will be based on AMD's new Pilderiver core which is derived from Bulldozer, the same core which we'll see in the soon to launch AMD FX series of CPUs.


The platform will be known as Comal and will continue to use the same chipsets A70M and A60M, and as far as we can tell, the same socket FM1. We'll also be seeing improved graphics, currently codenamed London. The TDP will remain the same as for the current APUs.


The roadmap also contains some of what AMD is planning for 2013. Kaveri which is Trinity's replacement will feature Steamroller CPU cores and a yet unknown graphics core. Below that we have Kabini which will feature Jaguar cores, a feature it shares with the tablet friendly Samara APU. Kaveri will be teamed up with the Bolton FCH, while Kabini gets the integrated Yangtze FCH and finally Samara will feature the integrated Salton FCH.

AMD FX Sets a Guinness World Record

On August 31, an AMD FX processor achieved a frequency of 8.429GHz, a stunning result for a modern, multi-core processor. The record was achieved with several days of preparation and an amazing and inspired run in front of world renowned technology press in Austin, Texas. This frequency bests the prior record of 8.309GHz, and completely blows away any modern desktop processor.

The result was achieved in a technology grand prix that must be seen to be believed. The refinement of techniques employed by the world’s finest overclockers creates a stunning scene that looks as if we are about to bring the machine to life. It was recorded by a fantastic video team and validated with a most useful tool called CPU-Z, the de facto standard for overclockers looking to capture a moment of stability under extreme conditions, or more practically under conservative tuning. The video has now launched and you will enjoy the spectacle of extreme overclocking and the record-breaking capabilities of upcoming AMD FX processors.

AMD Confirms Delay of Bulldozer

During the company's press conference dedicated to the launch of AMD 9-series core-logic sets, AMD announced that its long-awaited FX-series microprocessors will be available within the next 60 to 90 days from now, which points to August or September. Mainboards based on the new AMD 9-series chipsets with AM3+ socket will support both existing high-performance multi-core Phenom II microprocessors as well as AMD FX-series chips with up to eight cores.


Below is official rendered image of AMD 8-Core FX Processor Black Edition box (PIB):

Bulldozer Maybe Delayed and Need More Tuning

Advanced Micro Devices may further delay release of its highly-anticipated FX-series microprocessors code-named Zambezi that are based on the Bulldozer micro-architecture, according to leaked slides published by a Chinese web-site.


The slides also point to new model numbers for the chips: FX-8150, FX-8100, FX-6100 and FX-4100, which may indicate that the company intends to alter previously set specifications of the chips.

Earlier it was believed that AMD would launch its FX-series of chips in June, 2011. It was also reported that the highest-performing eight-core offering would cost $320, which is in line with the top-of-the-range Core i-series "Sandy Bridge" CPU.

The long-awaited CPUs are now fully-functional and work without flaws, according to a person who wished to remain anonymous. The problem with the delay of the AMD FX family of chips is that they currently cannot operate at truly high-speeds and thus cannot achieve performance levels that AMD wanted them to. As result, AMD will need to design a new stepping of the processor and therefore delay the commercial launch to September.

The currently available B0 and B1 stepping Zambezi/Bulldozer processors can function at around 2.50GHz/3.50GHz (nominal/turbo) clock-speeds and at such frequency they cannot deliver performance AMD considers competitive, a person with knowledge of the situation said on Monday. As a consequence, AMD needs to tune the design of the processor and create B2 stepping of the chip with better clock-speed potential amid similar thermal design power (TDP), which will take several months to complete.

This will slowdown revenue growth of the chip-maker and will also hit its reputation, as this is by far not the first or second delay of Bulldozer in general and Zambezi in particular.

AMD to Fight i7 Sandy Bridge with Bulldozer

Advanced Micro Devices at present is not able to compete in terms of pure performance against Intel Corp.’s Core i7-series “Sandy Bridge” microprocessors, according to benchmark results, but with its next-generation Bulldozer-based chips the company fully aims to fight.


A leaked document from AMD seen by us here claims that AMD FX-series central processing unit (CPUs) powered by the Bulldozer micro-architecture will be fully able to rival Core i7 2600-series chips. Below is the art-work for FX processor box:


AMD plans to introduce AMD 900-series chipsets compatible with Zambezi processors in Q2 2011. The Bulldozer processors, Radeon HD 6000 "Northern Islands" discrete graphics cards and AMD 900-series core-logic sets will power AMD's next-generation enthusiast-class platform code-named Scorpius.

The initial family of AMD's FX-series microprocessors based on Bulldozer micro-architecture will include four models with eight, six or four cores and 95W or 125W thermal design power (TDP).

The first breed of AMD FX8000, FX6000 and FX4000 currently known under Zambezi code-name will completely support all the advantages that the Bulldozer micro-architecture is supposed to bring, including new Flex FP floating point processing unit. The new chips in maximum eight-core configurations are projected - by AMD's internal documents - to offer roughly 50% performance improvement over Phenom II-series microprocessors in multimedia applications.


The primary family of AMD FX-series central processing units (CPUs) will feature two eight-core models, one six-core flavour as well as a quad-core version. The chips will support dual-channel DDR3 1866MHz memory, will support Turbo Core dynamic acceleration technology and will come in AM3+ form-factor and will have 125W and 95W TDP. The second "wave" of Bulldozer chips will also include four microprocessors and will improve performance of the initial breed of CPUs.

AMD's Financial Analyst Day 2010

At its annual Financial Analyst Day, AMD demonstrated how its unique combination of CPU and GPU computing technologies on a single die will enable breakthrough capabilities in an innovative processor design with planned OEM system availability in early 2011. AMD executives detailed how this new class of processor, AMD Fusion Accelerated Processing Units (APUs), is poised to create a sustained position of advantage by powering demanding PC workloads in sleek form factors with long battery life, pacing future AMD growth. AMD Fusion APUs are built from DirectX® 11-capable GPU technology and either low-power or high-performance multi-core x86 CPU technology. These APUs are designed to vastly improve today's Internet, video processing and playback, and gaming (client and online) experiences. For the first time, AMD also demonstrated its new high-performance x86 multi-core CPU architecture codenamed "Bulldozer" and provided additional information around the "Bulldozer" launch schedule.

AMD also announced several notable updates to its 2011-2012 roadmaps including:

Server - All “Bulldozer”, all the time



By the end of 2012, every new server CPU AMD bring to market will be based on the new “Bulldozer” architecture. “Bulldozer” will power multiple platforms designed to give server buyers what they keep asking for: more threads, more memory and the best performance-per-watt, per-dollar available. The “Terramar” CPU, planned for release in 2012, will indeed scale to 20 cores in our mainstream, 2- and 4-processor server platforms. The cloud’s new best friend will be “Sepang,” planned for release in 2012. “Sepang” is planned to scale up to 10 cores and is targeted at low-cost, low-power 1- and 2-way platforms.

Client Roadmap - Velocity goal: the best APUs every year

Velocity is the term Rick Bergman introduced last year to signal the new AMD Fusion APU design methodology and product introduction cadence. AMD Velocity builds on AMD’s already established annual GPU design cycle to achieve a faster pace of innovation than AMD previously achieved with a CPU-only development focus.




"Krishna" and "Wichita" are two and four-core 28nm APUs based on the next-generation sub-one watt "Bobcat" CPU cores and a DirectX 11-capable GPU, designed for the tablets, notebooks, HD netbooks, and desktop form-factors. "Trinity" is a 32nm APU based on AMD's next-generation "Bulldozer" CPU cores and a DirectX 11-capable GPU, designed for mainstream and high-performance desktops and notebooks. "Komodo" is a 32nm CPU featuring up to 10 AMD "Bulldozer" CPU cores designed for high-performance and enthusiast desktops.

First Details about Bulldozer

Advanced Micro Devices revealed numerous details about the Bulldozer micro-architecture, which will power the company's next-generation central processing units (CPUs) for desktops, servers and workstations. Apparently, the main goal of AMD's designers when it came to Bulldozer was to ensure maximum sharing of resources within multi-core microprocessors to get high performance amid moderate low power and die sizes.

Traditional approach to creation of multi-core microprocessor is very straightforward: each core acts independently and shares only the most obvious resources with other: L3 cache, memory controller, processor bus, etc. In Bulldozer designs, cores will be able to dynamically share fetch and decode blocks, caches and other units. At least in initial designs, multi-core chips will consist of several major blocks, each of which will have two independent integer cores (that will share fetch, decode and L2 functionality) with dedicated schedulers and two 128-bit FMAC pipes with one FP scheduler. This means that each major block (Bulldozer Module) is, according to AMD, essentially a tightly-linked dual-core microprocessor with shared fetch, decode and floating point units.


Such "dual-core" major block will not be as efficient as two traditional cores, but will consume less power and will use less die space, which in effect means that more of major building blocks can be installed without increasing thermal design power or die size to unacceptable levels. Moreover, AMD reasonably claims that the approach is more efficient than simultaneous multi-threading or chip-level multi-threading. In fact, according to AMD, each major block can provide 80% performance of a dual-core microprocessor.

AMD also implied that Bulldozer will feature a new predication-directed instruction preset mechanism to overlap instruction miss requests to cache or memory and thus improve efficiency of execution. In particular, this will help AMD to maximize utilization of the "dual-core" major blocks of Bulldozer microprocessors.

Given the fact that AMD's Bulldozer architecture seems to be very modular, we can expect AMD to tailor designs in accordance with performance and/or power requirements.


The first Bulldozer processors will be made using 32nm SOI fabrication process in 2011 and that with 33% increase of the number of cores, up to 50% of additional performance may be received in server applications, at least, based on AMD's internal simulations.

More info : TechReport - AMD's Bulldozer architecture revealed