News
 

Bookmark and Share

(6) 

Fujitsu, a maker of various electronics and also a developer of microprocessors, this week disclosed some additional details regarding its Venus processor. Unfortunately, the company is still unsure about availability timeframe for the chip as well as systems on its base.

According to Fujitsu, 128GFLOPs peak performance of eight-core SPARC64 VIIIfx (Venus) is 2.5 times higher compared to Intel’s top-of-the-range dual-core Itanium 2, however, even with built-in memory controller Venus consumes only 33% of what Itanium 2 uses, hence, about 35W, according to PC Watch web-site.


Fujitsu SPARC64 VIIIfx processor. Image by PC Watch

It should be noted that Intel Itanium 2 chips only have two processing cores and are made using 90nm process technology, meanwhile, Fujitsu’s SPARC64 VIIIfx (Venus) is made using 45nm low-power process, sports eight processing cores and is not yet available on the market. Intel plans to unveil a new breed of Itanium chips with four cores and produced using 45nm technology. It is highly likely that Fujitsu’s Venus will have to compete against those chips.


Fujitsu SPARC64 VIIIfx processor, die shot. Image by PC Watch

At present Fujitsu is only testing the new processors and so far it has made no comments regarding actual availability. The only official clue is that the chips “will find themselves in applications in the course of the following years” can hardly be considered as an update.

Fujitsu’s SPARC64VIIIfx is aimed at supercomputers and enterprise-class servers from Fujitsu. Theoretically, the chip could be found inside Sun’s servers, however, given that Sun Microsystems was recently acquired by Oracle, the destiny of SPARC-based servers from Sun is not completely clear.

Tags: Fujitsu, Venus, 45nm, Sparc

Discussion

Comments currently: 6
Discussion started: 05/15/09 12:11:56 PM
Latest comment: 01/18/11 04:40:38 PM
Expand all threads | Collapse all threads

[1-4]

1. 
"...Sun Microsystems was recently acquired by Novell..."

Oracle rather than Novell...
0 0 [Posted by: bluvg  | Date: 05/15/09 12:11:56 PM]
Reply

2. 
Hopefully, this chip will help Fujitsu smack Intel upside their iTanic head and get a good piece of the server market and drive the prices of the moronic iTanic chips downwards.

As for Sun Microsystems ... I'm very afraid they'll end up like 3DLabs in the Creative belly or 3dfx in the nVidia stomach ...
0 0 [Posted by: East17  | Date: 05/15/09 01:54:06 PM]
Reply

3. 
Two things to recognise here: 1) Current ATI/AMD graphics cards hit 240 GFlop/sec at double precision, and 1 TFlop at singe precision. 2) Intel are yet to release Larrabee (32 cores) - notional performance is 2 TFlops as per the Wikipedia article. I'm glad that the author of the post did his / her homework...
0 0 [Posted by: funkygibbin  | Date: 05/16/09 05:44:54 PM]
Reply
- collapse thread

 
Please stop comparing CPUs to GPUs. Yes, they can achieve those peak computations, but they are limited to what they can do.
For example, there are some type of Folding@Home workloads that cannot be done on GPUs.
Besides, the processor in the article is a "stand-alone" processor, whereas AMD/nVidia (and Larabee) are all co-processors, meaning they still need a CPU to work.
GPUs have another momentary advantage: the memory bus is much faster. Disadvantage: memory cannot be share between 2 boards and if you need to increase it, you have to change the board.
0 0 [Posted by: mathew7  | Date: 05/18/09 12:40:23 AM]
Reply
 
Guess what ?!!? Larabee is DEAD
0 0 [Posted by: East17  | Date: 01/18/11 04:40:38 PM]
Reply

4. 
"Current ATI/AMD graphics cards hit 240 GFlop/sec at double precision, and 1 TFlop at singe precision. "

Yes but what you are failing to mention is that the support for MIMD is lower on GPUs due to the reduced branching logic. This makes some algorithms quite hard to port to the GPU paradigm.

Furthermore, the per-core cache is about 9 KB (or double that or so). For many algorithms, 9KB is not acceptable.

0 0 [Posted by: mickey1  | Date: 02/09/10 11:07:55 AM]
Reply

[1-4]

Add your Comment




Related news

Latest News

Wednesday, August 20, 2014

10:53 am | AMD to Cut Prices on FX-9000, Other FX Processors: New Prices Revealed. AMD to Make FX Chips More Affordable, Discontinue Low-End Models

10:32 am | LG to Introduce World’s First Curved 21:9 Ultra-Wide Display. LG Brings Curved Displays to Gamers, Professionals

9:59 am | AMD Readies FX-8370, FX-8370E Microprocessors. AMD Preps Two New “Mainstream” FX Chips

Monday, August 4, 2014

4:04 pm | HGST Shows-Off World’s Fastest SSD Based on PCM Memory. Phase-Change Memory Power’s World’s Fastest Solid-State Drive

Monday, July 28, 2014

6:02 pm | Microsoft’s Mobile Strategy Seem to Fail: Sales of Lumia and Surface Remain Low. Microsoft Still Cannot Make Windows a Popular Mobile Platform

12:11 pm | Intel Core i7-5960X “Haswell-E” De-Lidded: Twelve Cores and Alloy-Based Thermal Interface. Intel Core i7-5960X Uses “Haswell-EP” Die, Promises Good Overclocking Potential