I'm not dissappointed at all because current CPUs are more than enough. This allows things to settle down and gives us many cheap upgrades that will let you run most anything. Technology wise it IS a bit boring to see so many incremental steps... here on in it's all gonna be multi-core computing.
RussianSensation, I know how you feel. My 3 year old chip is still faster than the latest Prescott :disgust:
No they didn't. ATI has no plans to release high end AGP, I got that straight from the horse's mouth. One of their partners had a typo on their website about it, but that was just a typo.
I never heard of this and I really don't think the added latency of an AGP bridge chip would hurt a teeny 6600GT.
Yes, it could hurt a current gen card; but I doubt it would need cooling.
I have to agree with the OP that CPU's are slowing down in progress, but (A) its not the engineers being lazy and (B) CPU performance can come a long way with proper software.
I think the biggest limitation today with CPU progress is physical. Why did Intel abandon a 4GHz P4? They were approaching the physical limits of the mfr process they were using; heat and size were increasing TOO much to be feasible as a long term plan. There is a LOT of research going on today as to how to improve transistor manufacturing, but IMO they are all just trying to find a way to extend the life of good ol' silicon. Realistically we're approaching the limits of silicon in its current incarnation, now matter if you use low-K, SOI, 90nm, 65nm, etc. Problems like leakage current due to shrinking gate/junction sizes, low yields due to complex processes, etc are all pointing to the fact that we are pushing silicon's limits. I think before we see a revolutionary change in CPU's, we'll need to see a revolutionary change in device fabrication.
On the other hand, we could see huge CPU advances with a change in logic. Kind of like Intel abandoning the MHz game, what if a completely new architecture made computing a LOT more efficient using current tech? Of course this also brings into consideration the software side of things, as you have to have an OS & programs to take advantage of this highly efficient architecture. I find the possibility of a logic revolution less likely than that of a fabrication revolution, mainly because (A) people are used to thinking in a certain way (i.e. x86) and (B) it would involve a massive change that would be damn near impossible to do unless it was led by someone like Intel. Sure, Cell has some promise and in future incarnations...who knows. I haven't read up enough on it to intelligently comment, but I think it is a step in the right direction.
I think the biggest limitation today with CPU progress is physical. Why did Intel abandon a 4GHz P4? They were approaching the physical limits of the mfr process they were using; heat and size were increasing TOO much to be feasible as a long term plan. There is a LOT of research going on today as to how to improve transistor manufacturing, but IMO they are all just trying to find a way to extend the life of good ol' silicon. Realistically we're approaching the limits of silicon in its current incarnation, now matter if you use low-K, SOI, 90nm, 65nm, etc. Problems like leakage current due to shrinking gate/junction sizes, low yields due to complex processes, etc are all pointing to the fact that we are pushing silicon's limits. I think before we see a revolutionary change in CPU's, we'll need to see a revolutionary change in device fabrication.
On the other hand, we could see huge CPU advances with a change in logic. Kind of like Intel abandoning the MHz game, what if a completely new architecture made computing a LOT more efficient using current tech? Of course this also brings into consideration the software side of things, as you have to have an OS & programs to take advantage of this highly efficient architecture. I find the possibility of a logic revolution less likely than that of a fabrication revolution, mainly because (A) people are used to thinking in a certain way (i.e. x86) and (B) it would involve a massive change that would be damn near impossible to do unless it was led by someone like Intel. Sure, Cell has some promise and in future incarnations...who knows. I haven't read up enough on it to intelligently comment, but I think it is a step in the right direction.
Zebo
Elite Member
- Jul 29, 2001
- 39,398
- 19
- 81
that aint sayin much.. prescott was always a POS.. No Intel guys I know went to them.. Duvie..thugsrook.. etc (well they tried and went back) Northwood C probably best CPU in last 5 years from intel.
It's AMD who has really improved. A-XP was getting a little long in tooth.. Bogus ratings attached to a Pentium A really which made them look rediculous against C. But AMD 64 changed all that. Not only surpasing Northwood but continues to scale.. some guys hit 3.0Ghz on default Vcores!!! FX-57 is 2.8ghz stock. Add in dual core which performs identical to single cores in single threads and many times double performance in multithreads it's not too hard to see improvement. All depends how you look at things and what your comp is used for if it's useful improvement or wasted. I've suggested to at least 10 people to stay w/ northwood when asking about upgrading due to thier circumstance.
Zebo
Elite Member
- Jul 29, 2001
- 39,398
- 19
- 81
RS- Here is another review with 3.2C, AXP 3200 and FX-55 for you to look at
http://www.aceshardware.com/read.jsp?id=65000327
Like win rar & statistical analysis..almost double the performance of 3.2C and totally demolishes 3200.
http://www.aceshardware.com/read.jsp?id=65000327
Like win rar & statistical analysis..almost double the performance of 3.2C and totally demolishes 3200.
- Sep 5, 2003
- 19,458
- 765
- 126
Yeah I remember this site. They used to have interesting reviews. Don't know why they stopped updating it....Their upgrade guides highlighted the importance of a graphics card all the way back in 2003; they were really on the ball.
I totally agree with everything that was said in this thread. Just my main concern is that I am somewhat disappointed that if I choose to upgrade soon to play games, I have to unnecessarily replace my cpu and motherboard. Yet I find both to be already fast enough for my needs. So essentially I am not happy with spending $250-300 to enjoy the benefits of PCIe, just to spend $300+ on a new generation videocard later. Imagine spending $600 towards a 512mb GTX AGP board instead?
Had cpu progress or cpu speed increases for gaming and other tasks like archiving, opening files, loading game levels improved significantly, I wouldnt have been upset with cpu progress.
Every time we change a generation in games, games will use CPUs less and less for anything but physics and AI.
Option 5 in the poll makes you sound like a moron. "They should sell it at a lower cost." First, cost is their total dollars to sell (including overhead), you mean price. To sound like even more of an economist, they're selling at the price that the market will bear. See Opteron 165's going on backorder quickly, then the price going up? Capitalism at work. Supply meets demand. Don't be a whiny kid, the technology will be priced less when it becomes mature.
AMD has supply issues. People want to buy even $320 Opteron 165's or $335 3800 X2's. Demand might be "low" but supply is sufficiently low to justify that price. PC components are almost at commodity pricing, they can't cut it just for you.
AMD has supply issues. People want to buy even $320 Opteron 165's or $335 3800 X2's. Demand might be "low" but supply is sufficiently low to justify that price. PC components are almost at commodity pricing, they can't cut it just for you.
To poster In 36 short days the new dual core yonahs arrive. ECS M/B with the intel 975x chipset with a yonah socket for the desktop . Also there WILL BE a Yonah EE model rated above 50 watts. No 64 bit but who really cares as its hardly used on the desktop yet.
So is this going to really be an improvement or not .36 days left we shall see . There should be reviews before this however . 2006 is going to be really exciting. I think we actually see true cpu progress and features . OR hype who knows but we can all hope.
So is this going to really be an improvement or not .36 days left we shall see . There should be reviews before this however . 2006 is going to be really exciting. I think we actually see true cpu progress and features . OR hype who knows but we can all hope.
Option 4 and 5 for me.
More needs to be done to get dual core into the mainstream and particularly the low end of the market.
I would love to see an $80 dual core sempron within the next couple of years. I think the main thing that is stopping Intel and AMD offering such a product is that the 90nm process is just too costly to offer due an enormous die to the low end.
More needs to be done to enhance consumer level applications for dual core and eventually multi core setups. At a workstation and server level there is already a great deal of support for multi threaded applications.
More needs to be done to get dual core into the mainstream and particularly the low end of the market.
I would love to see an $80 dual core sempron within the next couple of years. I think the main thing that is stopping Intel and AMD offering such a product is that the 90nm process is just too costly to offer due an enormous die to the low end.
More needs to be done to enhance consumer level applications for dual core and eventually multi core setups. At a workstation and server level there is already a great deal of support for multi threaded applications.
- Sep 5, 2003
- 19,458
- 765
- 126
Yes I know what the difference between cost and price is. However, cost can also refer to "the amount of money that must be paid to acquire something; purchase price or expense" - Dictionary of Business Terms, Jack P. Friedman, Barron's Business Guides
In this case most people I thought would understand that I was phrasing the question towards the buyer's point of view; in which case the 2 synonyms could be used interchangeably.
Either way, when A64 x2 3800+ sells for higher price than 2x A64 3200+, your theory of supply and demand holds true, but tells us nothing about market efficiency. Most consumers buy on emotions and impulse, rather than rationality and comparison of facts. But if you take the latter into consideration you'll realize that dual-core processors are simply overpriced. Just because the consumers are willing to pay those prices, does not necessarily mean the prices are fair (it just means they have very little choices and decide to bear the burden of early technological adoption).
I am not asking for companies to necessarily reduce prices significantly. But except for Intel 820 processor which is clearly a "bargain" if you will when you compare it to its price to a single core 2.8ghz processor, there has been no initiative from Intel and especially AMD to make dual-core processors mainstream. The problem with that is lack of software developers committment for making programs and games that benefit from such technologies since the user-base is simply too small. Given that IBM 3.2 triple core, dual-threaded processor costs $106 to produce for microsoft, I simply cannot accept that cost of production scales linearly when you add a 2nd core to say 3200+ to make it an X3800+. What is happening is that AMD and Intel are capitalizing on people who believe this price is fair, yet logically it is clear that you pay a premium (in which case as a firm you are not really promoting the switch towards a multi-processor environment, at least on the desktop market).
cpu's arent the only things lacking. what about hard drives. unless you make a raid array, hard drives only average roughly 50-70mbps w/o raid, w/ raid id imagine considerably faster. but the thing is with raid if you get a drive that fails, you lose considerable data. so its not only cpu's that are lacking but also hard drives. that also falls on the software as well, bloatware. as in, windows xp vs windows 2000, i loved windows 2000, low ram useage, high uptimes, great compatibility and simple UI, wish windows would go back to that. only thing similar to that nowadays is linux.
Only for certain things. I had my 3.06ghz northwood @3.45ghz, and compared to my prescott @ 3.82ghz, the northwood still beat the prescott in a lot of things. The prescott did better with video encoding, and typicaly did better with programs that tookfull advantage of hyperthreading, but the northwood was still ahead in almost everything else, with a 370mhz clock speed differance..it was ahead in 3d rendering, MP3 encoding, gaming..with the prescott at 3.45..the northwood was ahead in everything but some synthetic benchmarks.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 384
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
Facebook
Twitter