Next-generation 28nm GPUs Could Be 45 Percent Faster
November 4, 2011 10:37 AM
comment(s) - last by
Compute Unit (click to enlarge)
Graphics Core Next (click to enlarge)
Sources claim TSMC's manufacturing is that good
Taiwan Semiconductor Manufacturing Company (TSMC) is the world's largest semiconductor foundry, and as such is constantly under pressure from its customers and competitors. The company recently announced that its
newest 28nm process
has entered mass production. Smaller process geometries typically mean that a faster, smaller, and cheaper processor will be available.
Two of TSMC's biggest customers are AMD and NVIDIA. Both have taped out GPU designs using TSMC's 28nm High Performance (28HP) process. AMD's next-generation GPU series is codenamed
GPU is supposed to launch early in December, while NVIDIA's
GPU will launch in February of 2012.
28HP is the first process from TSMC to use High-k Metal Gate (HKMG) technology, as opposed to the typical silicon oxynitride (SiON) found in 40nm GPUs. HKMG uses a material with a high dielectric constant instead of the traditional silicon dioxide gate dielectric. This allows for a substantial reduction in gate leakage, thus lowering overall power consumption and allowing for higher clock speeds.
According to sources within TSMC, the 28HP HKMG process is doing very well. So well, in fact, that it supports up to a 45 percent clock speed improvement over the firm's own 40G process used to make the last two generations of video cards. This speed improvement is based on the same leakage per gate, but the GPU firms may choose to favor lower power consumption over a pure speed boost.
Our AMD contacts declined to respond to these assertions and directed our attention to a presentation made in June at the AMD Fusion Developer Summit by Eric Demers, the Chief Technology Officer of AMD Graphics.
The flip side of the manufacturing process is the architecture and design, and AMD has already previewed the basis for its future graphics architectures, known as Graphics Core Next (GCN). The basic design of GCN will form the foundation for the next few generations of AMD graphics processors.
The fundamental unit of
AMD’s previous designs
has been the Streaming Processor, utilizing a Very Long Instruction Word (VLIW) architecture designed to take advantage of Instruction Level Parallelism (ILP). That will be supplanted by the Compute Unit, comprised of multiple 16-wide vector Single Instruction, Multiple Data (SIMD) units designed for Thread Level Parallelism (TLP).
L1 and L2 caches will support the CUs, while the GPU will have access to the main system memory. Support for C++ programming has been added, making it easier to program for the GPU and CPU within the same application.
While all of these changes will benefit compute applications, it is not yet clear what impact this will have on current games and those already in the pipeline. The architectural changes may end up helping or hindering performance. Nevertheless, the entire 28nm graphics lineup will support
resolutions of up to 16000 x 16000 pixels
Of course, yields have always been a big problem with introducing a new node. TSMC had significant challenges with its
, leading to shortages of the
Radeon HD 5800 series
. HKMG processes typically use atomic layer deposition for high-k materials, which has been a challenge even for Intel.
Ultimately, we know that the next-generation of GPUs will be "significantly" faster, but the effects of the new architecture are still unknown and could change things positively or negatively. The better the performance, the more likely it is to sell out. The final clock speeds are still being determined, so final performance numbers will have to wait until the first official launch. The latest word on the street is December 6.
This article is over a month old, voting and posting comments is disabled
RE: Process isn't everything
11/4/2011 12:57:23 PM
Bulldozer performance per Watt being low is related to the design AMD has designed. They used a long pipeline while hopping that speed and better branching prediction would reduce the impact. It didn't. The architecture could be good, but it still need significant tweaking to outperform previous K10.5 design. A K10.5 with a die-shink, like the A8350 but without the integrated GPU, would absolutely blast Bulldozer. But after investing so much in Bulldozer, AMD isn't going to scrap it. I do think they'll do what Intel did with the first Core design. It combined the best of the Pentium Mobile and of Netburst. Ultimately they will end up with a better CPU design. And Intel isn't going to let AMD die, as other said, the Damocles sword of anti-trust being already upon them.
RE: Process isn't everything
11/4/2011 1:29:30 PM
The only benefit I saw in someone getting Bulldozer is if they plan on massively overclocking it. It sure overclocks a lot higher than the previous design but sadly clock for clock the bulldozer is a dud. Still I would probably go for a Intel 2600K and overclock that from all the benches Ive seen.
"Intel is investing heavily (think gazillions of dollars and bazillions of engineering man hours) in resources to create an Intel host controllers spec in order to speed time to market of the USB 3.0 technology." -- Intel blogger Nick Knupffer
AMD Video Cards Now Support 16000x16000 Resolution
November 1, 2011, 5:43 AM
TSMC Confirms 28nm Chips in Mass Production
October 24, 2011, 3:43 PM
AMD Launches 2GB Radeon HD 6900 Series Video Cards
December 15, 2010, 8:47 AM
TSMC Says 40nm Problems Resolved, Preparing 28nm Fab Production
January 21, 2010, 8:15 AM
ATI Launches Radeon HD 5800 Series, First DirectX 11 GPUs
September 23, 2009, 12:01 PM
Tim Cook Touts Apple’s Product Portfolio, Performance in Letter to Employees
October 21, 2014, 8:05 AM
Lenovo Once Again The Top Global PC Maker, Apple Takes 50% of PC Profits
October 9, 2014, 7:46 AM
It’s Official: HP to Split Into Two Business Units
October 6, 2014, 8:14 AM
Quick Note: HP Plans to Split PC, Enterprise Units
October 5, 2014, 4:31 PM
Samsung Prepping 14-nm for Apple A9 Production, Handling 30 Percent of A8 Orders
October 2, 2014, 1:17 PM
Lenovo to Complete $2.1B Acquisition of IBM's x86 Server Operations on October 1
September 29, 2014, 3:19 PM
Most Popular Articles
Chinese Government Declares Digital War Against America's Top Tech Firms
October 20, 2014, 12:07 PM
Samsung Announces Galaxy S5 Plus with Snapdragon 805 Processor, LTE-A
October 22, 2014, 3:40 PM
Windows 8.1 + Android "Sell Mini PC" w/ Bay Trail Creates New PC Form Factor
October 20, 2014, 5:07 PM
Apple's iPad Air 2 Features Triple-core A8X Processor, 2GB of RAM
October 21, 2014, 8:32 PM
Apple Releases iOS 8.1; Adds Apple Pay Support, SMS Relay, Instant Hotspot
October 20, 2014, 1:00 PM
Latest Blog Posts
The Surface Mini That Was Never Released Gets "Hands On" Treatment
Sep 26, 2014, 8:22 AM
ISIS Imposes Ban on Teaching Evolution in Iraq
Sep 17, 2014, 5:22 PM
Space Terrorism is a Looming Threat For the United States
Apr 23, 2014, 7:47 PM
Facebook Aims to Provide Internet to "Every Person in the World" with Drones, Satellites
Apr 1, 2014, 10:20 AM
Retail Mobile Sites Experience Outages in Light of Simplexity's Bankruptcy
Mar 14, 2014, 8:48 AM
More Blog Posts
Copyright 2014 DailyTech LLC. -
Terms, Conditions & Privacy Information