What is the performance of AMD bulldozer?

ByGiovanna Carosella 2020-01-06 583

The AMD bulldozer is at the heart of AMD's complete redesign and will be AMD's next-generation high-performance processor technology for client and server applications, with a 33% increase in core and about 50% performance compared to the Opteron 6100 series.

As a new generation of processor architecture, the AMD bulldozer will use 32nm SOI technology, which allows the bulldozer to increase the number of cores and throughput by 33% and 50% without increasing power consumption compared to the "Magny-Cours" Hao long processor.

Unlike all processors before AMD, the bulldozer uses a "modular" design, with each "module" containing two processor cores, which is a bit like a single-core processor with SMT enabled. Each core has its own integer scheduler and four proprietary pipelines, and the two cores share a floating-point scheduler and two 128bit FMAC multiplication accumulators.


Turbo Core full core acceleration technology

Turbo Core technology mainly refers to some workloads that are not completely consumed to the maximum extent, to speed up the clock speed. Using Turbo Core on a variety of different workloads can increase performance by up to 500 MHz. The most important point of, Turbo Core acceleration refers to the acceleration of all nuclei. Unlike some nuclear acceleration technologies, previous nuclear acceleration technologies may need to shut down some cores and accelerate only some of them. The use of Turbo Core technology can increase the growth rate of all nuclei by up to 500 MHz, and if some more nuclear operations are shut down, the acceleration will exceed 500 MHz. At the same time, we further optimize the memory controller to improve the memory throughput.

In addition to 4 integer calculation pipelines for each core, the bulldozer uses "FlexFP" technology for floating-point operations. the two cores share a floating-point scheduler and two 128bit FMAC multiplication accumulators, which can be combined to perform two 64-bit double-precision calculations or four 32-bit single-precision calculations per clock cycle. If one core does not perform floating-point operations, the other core can occupy the two 128bit FMAC, to complete four double-precision operations or eight single-precision operations in one clock cycle, which AMD names as AVX mode. This technology ensures the floating-point computing power of the "bulldozer" and does not sacrifice performance because of "sharing" in high-performance computing.

New interface and new technology

The bulldozer processor will adopt a Socket AM3+ interface with 942 pins, which is different from the current 938-pin Socket AM3 interface. Its advantage is that it can support DDR3-1600 memory and advanced energy-saving technologies, and AM3+ will be the last generation of AMD's pins grid array (PGA) package, and then it will switch to the contact grid array (LGA),. When the Fusion fusion processor arrives, it will use the new LGA AF1 interface, with as many as 1591 contacts. Support DisplayPort 1.2standard, PCI-E 3.0specification (32 channels), four-channel memory.

Enhanced memory controller

8 years ago, AMD first introduced integrated memory controllers. According to AMD's experience in this field and very good technology, it has comprehensively improved the performance of memory controllers in this generation of products. First of all, the memory controller has been specifically redesigned and improved in terms of efficiency, so as to achieve a 30% improvement in memory performance. On the basis of a 30% performance improvement, an additional 20% performance can be achieved by making the memory support 1600MHz frequency. Taken together, the throughput of the memory controller can be improved by 50%.

Support both AVX instruction and SSE instruction

FLEX FP is by far the most innovative floating-point computing technology in AMD. Each module has a FLEX FP for floating-point operations. If you use traditional 128-bit coding, it means that each core will have a separate floating-point unit. Compared with competitors, AMD performs twice as much under the premise of 128bit coding. If it is a 256bit AVX code, Bulldozer can execute two floating-point units together. So in 256-bit coding execution mode, the number of execution is the same as that of competitors. But Bulldozer has a great advantage that it can execute both 256bit AVX instructions and SSE instructions at the same time. Competitors can't do this, they can only choose between AVX or SSE, this advantage will enable Bulldozer to have higher performance in high-performance computing, media codec and some technical computing.

More advanced power management technology

The circuit required by the second integer core in each module accounts for only 12% of the total core area, which will only add 5% of the circuit to the entire core at the chip level. More core, less space, which obviously helps to improve the performance of unit power consumption and unit cost.

The amount of energy consumption is determined by the number of clocks being powered on, and it depends on how many transistors need to be powered on to execute a normal instruction (operation). Under the maximum percentage of clock power, Bulldozer has a very good energy consumption performance in both normal application state and idle state. At the same time, it is optimized on each energy consumption unit, and the power supply can be turned off under various units. The reason for the high energy consumption of high-performance computing is mainly due to floating-point computing, while the general application operation is mainly consumed in the execution unit. At the same time, when idle, AMD's technology can turn off power completely for cores that are not needed at all. There was a big shift in AMD products last year, when AMD introduced a new slot, and the 2011 bulldozer could use the 2010 slot. Competitors launch new platforms and new slots at the same time, which also gives AMD an advantage. 


 You may also want to read:
 Does the HUAWEI FreeBuds 3 Work Well With iPhone?
 Honor Magic Watch 2 VS Amazfit Stratos 3: Which Smart Running Watch is Better?
 EMUI 10 Always-On Display (AOD) feature: supported devices and everything you need to know about it
 HUAWEI Band 4 Pro VS Band 4 VS Band 3 Pro: What are the Differences Between these there Smart Band?

Gearbest Daily deals up to 69 percent off

Extensive Product Selection

● Over 300,000 products

● 20 different categories

● 15 local warehosues

● Multiple top brands

Convenient Payment

● Global payment options: Visa, MasterCard, American Express

● PayPal, Western Union and bank transfer are accepted

● Boleto Bancario via Ebanx (for Brazil)

Prompt Shipping

● Unregistered air mail

● Registered air mail

● Priority line

● Expedited shipping

Dedicated After-sales Service

● 45 day money back guarantee

● 365 day free repair warranty

● 7 day Dead on Arrival guarantee (DOA)

You might also like

HOT TOPICS

Related Products