News

 

Bookmark and Share

(3) 

Neither ATI Technologies, nor Nvidia Corp. are expected to release full lineups of their DirectX 10-compatible graphics processing units (GPUs) this year, which is why both companies are very likely to introduce higher-performance chips that belong to current generation of products this autumn.

ATI, which is expected to release its code-named R600 GPU late this year or early next, is preparing a breed of new chips based on the X1000 micro-architecture for its fall product lineup refresh, according to sources close to the company. In addition, ATI is projected to quietly roll-out less expensive versions of the current products. The new chips will serve high-end, performance-mainstream, mainstream and entry-level markets, providing either new levels of performance, or much lower pricing compared to today.

For the high-end market ATI is projected to release its code-named R580+ design, which will feature the same Radeon X1900 XTX chip, but now equipped with GDDR4 memory at higher clock-speeds compared to today’s GDDR3. The GPU’s memory controller will be reprogrammed to handle higher clock-speeds and the whole R580+ design will allow ATI to test the GDDR4 memory ahead of its broad deployment in the next-generations of products. The product may be approximately 15% faster compared to the current flagship offering from the company.

For the performance-mainstream market the company will offer its code-named RV570 graphics chip with 36 pixel shader processors, 12 texture units and 256-bit memory bus. The product will be branded as the Radeon X1900 GTO and will be targeted at below $299 segment. The RV570 is planned to be released commercially for back-to-school season.

The mainstream market will also see a new high-performance chip code-named RV560, which will be branded as the Radeon X1700-series and will cost up to $199. This part is expected to feature 24 pixel shader processors, 8 texture units and 128-bit memory bus. In order to minimize the cost of the chip, ATI is likely to use TSMC’s 80nm process technology for its manufacturing. Mass launch of the Radeon X1700 is scheduled during the back-to-school season.

In summer ATI will also make another aggressive attempt to reduce the costs of its Radeon X1600-series product family. Sources with knowledge of the matter indicated that there are chips code-named RV535 and RV536 in the roadmap, which are expected to be made using TSMC’s 80nm process technology and UMC’s 90nm fabrication process, both are much less expensive compared to TSMC’s 90nm manufacturing process. The specs of the Radeon X1600-series product will remain the same as today, but new manufacturing processes will allow ATI to set more aggressive pricing on those components.

Additionally, the Radeon X1300-series family will be updated with the code-named RV516 chip, which is produced at UMC’s 90nm nodes, and which will have 32-bit memory bus option (current Radeon X1300 only support 64-bit and 128-bit memory bus).

Officials from ATI Technologies did not comment on the news-story.

Discussion

Comments currently: 3
Discussion started: 04/07/06 01:29:02 AM
Latest comment: 04/09/06 05:14:59 AM

[1-3]

1. 
That's a nice piece of information
[Posted by: thnks  | Date: 04/07/06 01:29:02 AM]

2. 
ATI will kick nVidia's arse...
[Posted by: BorisFromStockdale  | Date: 04/07/06 07:19:07 PM]

3. 
32bit memory bus? Huh, even slower than the 64bit, what's the point with ULTRA-LOWEND cards? :P
[Posted by: aguy  | Date: 04/09/06 05:14:59 AM]

[1-3]

You must log in to add comments.

Forgot password? Registration

remember me



Related news

Latest News

Thursday, September 2, 2010

4:57 pm | Nvidia Solidifies Positions in the Professional Graphics Market Segment - JPR. Nvidia Quadro Remains Graphics Accelerator of Choice by Professionals

12:01 pm | Globalfoundries Tapes Out Industry's First 28nm Chip. Globalfoundries Tapes Out 28nm ARM Cortex A9 Qualification Vehicle

Wednesday, September 1, 2010

11:13 pm | AMD Announces Inexpensive DisplayPort to Single-Link DVI Adapter. AMD Expands Monitor Range for ATI Eyefinity with New Adapter

10:42 pm | Nvidia Reportedly Working on a New Core-Logic Set. Nvidia Rumored to Launch GPU with Chipset I/O Functions

10:01 pm | Globalfoundries to Start Risk Production Using 20nm/22nm in Late 2012. Globalfoundries on Track for Volume Production at 20nm/22nm Nodes in 2013

6:12 pm | AMD Displays Die-Shot of Upcoming Eight-Core "Orochi" Processor for the First Time [UPDATED]. AMD Shows Off Die-Short of 32nm Bulldozer-Based "Orochi" Chip

5:06 pm | Graphics Chips to Gain Horsepower and Programmability - Chief Scientist of Nvidia. CPUs and GPUs Will Get Close to Each Other, But Will Not Converge

3:29 pm | Globalfoundries Adds Another 28nm Fabrication Process to Roadmap. Globalfoundries New 28nm HPP Process to Boost Performance

2:18 pm | PNY Offers Discounts to Owners of BFG Technology Graphics Cards. Owners of BFG May Get New Boards with 25% Discount