Next Generation Changes Hardware/Architecture
on
New ATi 3D Chip
·
· Score: 2
Due to the complexity of the problems surrounding graphics there are a number of really innovative solutions coming to market. This particular chip and the nVidia product represent evolutionary changes, however there will be a number of revolutionary changes coming to market around mid-next year.
A number of companies have specialized in developing really unique and interesting graphics engines which they are marketing to the big makers in the form of logic cores. They range from faster engines to render more polygons per microsecond, to more unique designs that render blocks of the screen at a time, to really uniqe designs that take polygons completely out of the picture. These engines will find their way into the next generation graphics chips very shortly.
Additionally there have been a number of advances in the architecture of the systems themselves that will make significant advances in graphics. The advent of really high performance memory (Rambus, DDR, etc.) will certainly improve performance. And then there are some really uniqe designs coming that use copper-wire interconnect technology to give roughly the same performance as "system on chip" designs. And with all that, they die sizes of the chips themselves will keep shrinking (I think the new ATI chips are done on a.18 micron process). So suffice it to say, things are only going to get more interesting.
This is a GREAT idea. Now us poor hackers in the bay area can afford to live here without having to shell out ungodly rent rates for little closets full of mice. Now the tables are turned and the mice ARE the closets! Seriously, this gives the terms "mobile home" a whole new set of meanings. One question though? What if ferrets turned out to be more viable large family dwellings? Do you think the state of California would lift the restriction on them if they were to be used as dwelling space only?
As I understood it, the interview section of Slashdot was there for us to hear from acknowledged experts and/or controversial figures involved within different technology fields. I think its pretty clear based on the postings that most people don't find JP to be an expert in any field. Then the question remains, is he a good candidate for this area? Does being controversial alone merit him the chance to sound off? In some cases, I might say "yes" however in this case I don't. JP has shown himself, time and time again, to be a young man with no professional ethic or interest in forwarding anything other than his own agenda. As an avid reader of Slashdot, I personally would hate to see it stoop so low. There are a number of far more knowledgeable/interesting people in technology today. I have no real desire to hear anything he has to say (and frankly anything he says is suspect). I don't think the expected flame-fest is really worth it. Let's find someone worth talking to. Perhaps Mike Shiffman has some time?
I will be the first to admit that I'm not clued in about either technology so I'm hoping the more clueful here can offer some insight. How does the performance benefit derived from such a device compare with what could be achieved from the research being done into optical devices (ie: those using light rather than electrons). It seems to me that if one of these two technologies gained favor, the research being done into the other would be lost. Is this the case? Is it possible to use the processes mentioned in this article to replicate an optical device? Besides eliminating heat emissions problems, I'm sure there are plenty of other benefits. Does anyone know?
You're right your assessment. Also, I would point out that AMD went on a tour 3 years ago saying that they were going to take 25% market share from Intel capturing much of the low-end market. Now its true that they didn't make alot of money doing it, but remember this is Intel, the company the fights tooth and nail for 5% market share changes. Now AMD has a really good chance to gain further strength with the Athlon. With the credibility that Intel's lost due to the whole Rambus debacle, there are a number of OEMs and Resellers that wouldn't mind bypassing Intel completely. Expect good things from AMD in the near future.
It's interesting that ZDnet folks are now coming out of the woodwork to throw punches. Whether roblimo's comments and actions were inproper is beside the point (and if you're asking my opinion they were). The fact of the matter here is that there would be no news at all had the ZDnet editor in question bothered to check his sources (and himself) before writing the article.
The article is clearly taken right from the website with no effort made to confirm anything. Stating as fact, and hinting that the information came as part of an interview. Since when do journalists start writing everything down as fact that they find on websites? The fact that the article itself appeared as "news" isn't testemony to the skill of the hoaxer, but rather to the fact that the editor was too lazy or incompetent to follow up. This goes back to Journalism 101 people.
Sun could merely be joining a trend that is now becoming more popular in the industry. There are a number of companies that produce configurable processor cores, allowing licensees to make alterations to suit their individual needs. The level of alteration possible differs from processor to processor, however the basic idea is the same. Instead of licensing a processor that started life 15 years ago for use in a UNIX workstation, they can tailor-make a processor quickly and easily. In many cases this actually turns out to shorten design time immensely, because the designers don't have to learn all the ins and outs of the processor and can tailor it to fit an pre-existing product. ARC Cores (http://www.arccores.com) claims to have shortened some design times from 6 months to 2 days. It makes sense to me as a logical evolutionary step.
I did not mean to imply that the MIPS processor was the only proprietary processor out there. It clearly isn't. My complaint was more that SGI tried to do far too many things at once. It succeeded for quite awhile doing all these things. But keeping a processor architecture on the cutting edge is an inordinately difficult task. And SGI wasn't exactly forthcoming about the architecture (I don't remember the licensing fee for a MIPS core, but it was ungodly even then) so there were very few ports. And I'm sorry, but IRIX is a poor unix.
Now, I don't know what the ultimately best choice for SGI is or will be. They had an extremely strong server following for a long while and did quite well against Sun, IBM and HP. Additionally, their workstation business was also extremely successful, however the main niche there was its extreme graphics performance. That industry has now become over-populated and would have overtaken SGI's dominance in another generation or two. The decision to offload to nVidia was a wise one. So like I said, I don't know what they should do. I just know what they did, and that it failed.
I'm sad to say it, but frankly this was in the cards long ago. SGI's reliance on proprietary processors hurt it greatly. In addition, the fact that SGI plays into so many different markets (servers, workstations, high-end graphics, processor cores, supercomputers, OS) against so many large established players didn't bode well. The high-end graphics market, while being very sexy, wasn't ever going to yield the kind of revenue they needed. Now they're slowly trying to re-position themselves and cutting loose alot of employees. Frankly it was a necessary move. I'm sorry to see so many of the "oh greedy corporation, cutting loose people to turn a profit" arguements. Perhaps one day when you get out into the real world, start your own corporation and are answering to a board, and your stockholders, you might see why something like this is necessary. Now that SGI is finally trying to get a handle on the market and refocus their efforts, perhaps they'll be able to turn it around. I certainly hope so.
Slashdot Mirror
A number of companies have specialized in developing really unique and interesting graphics engines which they are marketing to the big makers in the form of logic cores. They range from faster engines to render more polygons per microsecond, to more unique designs that render blocks of the screen at a time, to really uniqe designs that take polygons completely out of the picture. These engines will find their way into the next generation graphics chips very shortly.
Additionally there have been a number of advances in the architecture of the systems themselves that will make significant advances in graphics. The advent of really high performance memory (Rambus, DDR, etc.) will certainly improve performance. And then there are some really uniqe designs coming that use copper-wire interconnect technology to give roughly the same performance as "system on chip" designs. And with all that, they die sizes of the chips themselves will keep shrinking (I think the new ATI chips are done on a .18 micron process). So suffice it to say, things are only going to get more interesting.
I suppose that before long GraphOn (or perhaps AntiOnline) will announce that statd will be the official exploit of the Chinese cyberwar effort?
This is a GREAT idea. Now us poor hackers in the bay area can afford to live here without having to shell out ungodly rent rates for little closets full of mice. Now the tables are turned and the mice ARE the closets! Seriously, this gives the terms "mobile home" a whole new set of meanings. One question though? What if ferrets turned out to be more viable large family dwellings? Do you think the state of California would lift the restriction on them if they were to be used as dwelling space only?
As I understood it, the interview section of Slashdot was there for us to hear from acknowledged experts and/or controversial figures involved within different technology fields. I think its pretty clear based on the postings that most people don't find JP to be an expert in any field. Then the question remains, is he a good candidate for this area? Does being controversial alone merit him the chance to sound off? In some cases, I might say "yes" however in this case I don't. JP has shown himself, time and time again, to be a young man with no professional ethic or interest in forwarding anything other than his own agenda. As an avid reader of Slashdot, I personally would hate to see it stoop so low. There are a number of far more knowledgeable/interesting people in technology today. I have no real desire to hear anything he has to say (and frankly anything he says is suspect). I don't think the expected flame-fest is really worth it. Let's find someone worth talking to. Perhaps Mike Shiffman has some time?
I will be the first to admit that I'm not clued in about either technology so I'm hoping the more clueful here can offer some insight. How does the performance benefit derived from such a device compare with what could be achieved from the research being done into optical devices (ie: those using light rather than electrons). It seems to me that if one of these two technologies gained favor, the research being done into the other would be lost. Is this the case? Is it possible to use the processes mentioned in this article to replicate an optical device? Besides eliminating heat emissions problems, I'm sure there are plenty of other benefits. Does anyone know?
You're right your assessment. Also, I would point out that AMD went on a tour 3 years ago saying that they were going to take 25% market share from Intel capturing much of the low-end market. Now its true that they didn't make alot of money doing it, but remember this is Intel, the company the fights tooth and nail for 5% market share changes. Now AMD has a really good chance to gain further strength with the Athlon. With the credibility that Intel's lost due to the whole Rambus debacle, there are a number of OEMs and Resellers that wouldn't mind bypassing Intel completely. Expect good things from AMD in the near future.
The article is clearly taken right from the website with no effort made to confirm anything. Stating as fact, and hinting that the information came as part of an interview. Since when do journalists start writing everything down as fact that they find on websites? The fact that the article itself appeared as "news" isn't testemony to the skill of the hoaxer, but rather to the fact that the editor was too lazy or incompetent to follow up. This goes back to Journalism 101 people.
Sun could merely be joining a trend that is now becoming more popular in the industry. There are a number of companies that produce configurable processor cores, allowing licensees to make alterations to suit their individual needs. The level of alteration possible differs from processor to processor, however the basic idea is the same. Instead of licensing a processor that started life 15 years ago for use in a UNIX workstation, they can tailor-make a processor quickly and easily. In many cases this actually turns out to shorten design time immensely, because the designers don't have to learn all the ins and outs of the processor and can tailor it to fit an pre-existing product. ARC Cores (http://www.arccores.com) claims to have shortened some design times from 6 months to 2 days. It makes sense to me as a logical evolutionary step.
Now, I don't know what the ultimately best choice for SGI is or will be. They had an extremely strong server following for a long while and did quite well against Sun, IBM and HP. Additionally, their workstation business was also extremely successful, however the main niche there was its extreme graphics performance. That industry has now become over-populated and would have overtaken SGI's dominance in another generation or two. The decision to offload to nVidia was a wise one. So like I said, I don't know what they should do. I just know what they did, and that it failed.
I'm sad to say it, but frankly this was in the cards long ago. SGI's reliance on proprietary processors hurt it greatly. In addition, the fact that SGI plays into so many different markets (servers, workstations, high-end graphics, processor cores, supercomputers, OS) against so many large established players didn't bode well. The high-end graphics market, while being very sexy, wasn't ever going to yield the kind of revenue they needed. Now they're slowly trying to re-position themselves and cutting loose alot of employees. Frankly it was a necessary move. I'm sorry to see so many of the "oh greedy corporation, cutting loose people to turn a profit" arguements. Perhaps one day when you get out into the real world, start your own corporation and are answering to a board, and your stockholders, you might see why something like this is necessary. Now that SGI is finally trying to get a handle on the market and refocus their efforts, perhaps they'll be able to turn it around. I certainly hope so.