• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

First 10 ghz equivalent chip? When?

D

Dario D.

Gawd
Joined
Dec 8, 2004
Messages
582
Once, at the launch of the P4's, there was this article that said, "at this rate, we'll have reached 10 ghz by the year 2010." (as opposed to the 2.0 ghz P4 that had just come out).

So, my question is, are CPU's still on track? Will there be a chip 4x more powerful than the first 2.0 ghz Pentium 4's in four years?
 
Conroes are darn fast. A P4 2.0ghz system is probably much slower than current system. Hard drives, memory, graphics have all gotten much faster than when the P4 was first released. Idk about 4x, but maybe.

Intel decided to not go the >4ghz route and decided to focus on dual core chips instead. That being said, many people around here routinely overclock their Intel dualcores over the 4ghz route. We won't be seeing a manufactured 10ghz chip anytime soon. Maybe overclocked....

 
I'd say the the fastest AMD and Conroe chips are close to 3x already... so yeah, 4x is definitely ahead. If you're talking overclocked there have been 7ghz+ P4's, and the 5ghz Conroe.
 
One thing with the netburst architecture was that it hit a plateau in efficiency and speed. One could overclock it to crazy levels but the benefits in calculation speed went down. The speed increase in calculations wasn't linear to the mhz increase. This was commented about in a lot of the extreme overclocking articles.
 
IBM's got a CPU that runs at 350 GHz @ room temp, and 500 GHz when chilled to -451.
 
tdg said:
IBM's got a CPU that runs at 350 GHz @ room temp, and 500 GHz when chilled to -451.
Click to expand...

You mean a single TRANSISTOR?? (or a few transistors) Not a FULL complex CPU.

I had seen something about the 500Ghz @ -471 somewhere, but I didn't see where it said it was a full chip.
 
I think its more about efficency and we will see 10 ghz but in cores which will probly add up to 20ghz in total.

You have amd working on theres and if one core is 2.5 GHZ then they add up to 10ghz and thats only year 2007 , in another three years who knows propably 25 ghz.

Theres loads of time and that time is loads especially for the computer industry where things double. :)
 
well, people are going to be claiming 10Ghz chips VERY soon, the engineer samples are already out there in the wild.

take a quad core processor clocked at 2.13Ghz, overclock it to 2.5Ghz x 4 = 10Ghz :eek: Ok not really, but sorta 10Ghz
 
I've been using a dual processor system for 5+ years... my first system was dual celerons... I will say this... I NEVER wait for a process to finish...

I still game on it.. far cry has a 1.2 GHz CPU requirement. My dual 575MHz OC celerons can play it fine... (on low settings of cource)

I do not however claim 1150MHz... but I would rate its AVERAGE performance around 1GHz... although, sometimes it just plain acts like 500MHz...
 
Like an Athlon64 10000+?
Click to expand...
If you're answering my original post, then exactly. I guess my question could also have been, "will there be an AMD 10000+ by 2010?" (assuming 10000+ means 10,000 ghz in Intel units)

I tried asking the same question as this thread on Ars Technica, and all I got was a load of snappy conjecture about how a 10 ghz chip technically won't exist anytime soon, and how gigahertz are complicated and can't be put on a scale very simply,... instead of anyone actually answering the general question.

Thanks for the good replies, guys :) All I get from Ars Technica is a bunch of technical arse.

So now, my question is, is it true that AMD's unit of measurement, like 2800+, means the equivalent speed of a P4 at 2.8 ghz? I know the speed of a Pentium ghz may have changed since AMD first started using this measurement, but I know that at one point or another it was their way of comparing their chips to Intel's. Is that still the case, or has the speed of the Pentium's ghz fluctuated too much?
 
AMD just makes up those numbers. Case in point: A64 2800+ is faster than AXP 2800+. I'm thinking they choose the numbers to compare to what they considor the equivalent P4 or Celeron, but as you can see it sometimes doesn't make sense. But yeah the speed to should be close to the competing cpu that came out at around the same time, give or take.
 
This isnt what you want to hear, but there is no true measure of processing power...

GHz stands for Giga Hertz. Which is merely how many times the power in the chip cycles. This is fine, as long as you are comparing apples to apples... but throw an orange in there... you should probubly be looking at how many instructions it can complete... now we are talking flops. G flops, T flops.... These numbers I dont have...
That too can be decieving. because some instruction sets are more efficient than others.

The numbers on the chips are irrelevant.. they dont cross over or compare... like ati and nvidia.

clock for clock the conroe is supposed to whoop the amd x64. (more flops)
my advice, is to check the bench marks for applications and games you use. then compare.
 
Conroes have been overclocked to 5Ghz.

The performance ratio between a Conroe and a A64 is roughly 1.3.

5 x 1.3 = 6.5Ghz Dual Core A64.

The performance ratio between a A64 and a P4 is roughly 1.6.

6.5 x 1.6 = 10.4Ghz Dual Core Pentium-D.

When will they come stock? Dunno. Shouldn't be too hard to overclock one of those 32nm Geshers to somewhere around there on light cooling by 2010.
 
xFlankerx said:
Conroes have been overclocked to 5Ghz.

The performance ratio between a Conroe and a A64 is roughly 1.3.

5 x 1.3 = 6.5Ghz Dual Core A64.

The performance ratio between a A64 and a P4 is roughly 1.6.

6.5 x 1.6 = 10.4Ghz Dual Core Pentium-D.

When will they come stock? Dunno. Shouldn't be too hard to overclock one of those 32nm Geshers to somewhere around there on light cooling by 2010.
Click to expand...

Excellent
 
Demon_of_The_Fall said:
HAHAHA, don't you just love it when people double / quadruple the speed of dual / quad core systems?

:rolleyes:
Click to expand...

I have to agree. But. If Intel Mitosis works as advertized . I am sure that by 2010 10ghz. Is more than likely a reality . Cores only need to be running 5 ghz because the other cores are also using ghz.on the same app. so the total ghz used on that app.Would be the total ghz. With a four core cpu 10 ghz is very possiable.
 
Alas, the increase in performance is not 100%, more like 2/3s more, or 66%.

Also, Quad Cores cannot Multiplex to a single core, as far as I know. Only to a supercharged Dual Core. I'm not knowledgeable about how they do it, but thats how some of the diagrams on XS showed it.
 
xFlankerx said:
Alas, the increase in performance is not 100%, more like 2/3s more, or 66%.. Sorry something came up . Olden shows a lot better than a 66% increase with 4 cores.


Figure 3. In this graph, the blue bars show the performance improvement going from an in-order to an out-of-order
core with about twice the amount of resources. The red bars indicate the performance of a processor with perfect
memory, illustrating the potential performance improvement for any technique that targets simply reducing memory latency. The yellow bars show the performance gains that result when using Mitosis with a four-core processor.

The results obtained by the Mitosis compiler/architecture for this subset of the Olden benchmarks are very encouraging, outperforming single-threaded execution by 2.2x. When compared with a big out-of-order core, the speed increase is close to 2x. One can also see that the benefits of Mitosis do not come only from reducing memory latency—using Mitosis enables the system to outperform an ideal system with perfect memory by about 60 percent. Overall, this work demonstrates that significant amounts of thread-level parallelism can be exploited in irregular codes, with a rather low overhead in terms of extra (wasted) activity.
Click to expand...
 
Dario D. said:
Will there be a chip 4x more powerful than the first 2.0 ghz Pentium 4's in four years?
Click to expand...

I wasn't going to say anything but I couldn't resist and I haven't seen anyone mention it yet but you do know that 2.0 ghz * 4 is 8 and not 10 right? I'm just giving you hard time; I have no opinion on the actual question.
 
meatycheesyboy said:
I wasn't going to say anything but I couldn't resist and I haven't seen anyone mention it yet but you do know that 2.0 ghz * 4 is 8 and not 10 right? I'm just giving you hard time; I have no opinion on the actual question.
Click to expand...

Lol, missed that. :p
 
Speculative parallelization can provide significant sources of additional thread-level parallelism, especially for irregular applications that are hard to parallelize by conventional approaches. In this paper, we present the Mitosis compiler, which partitions applications into speculative threads, with special emphasis on applications for which conventional parallelizing approaches fail.The management of inter-thread data dependences is crucial for the performance of the system. The Mitosis framework uses a pure software approach to predict/compute the thread's input values. This software approach is based on the use of pre-computation slices (p-slices), which are built by the Mitosis compiler and added at the beginning of the speculative thread. P-slices must compute thread input values accurately but they do not need to guarantee correctness, since the underlying architecture can detect and recover from misspeculations. This allows the compiler to use aggressive/unsafe optimizations to significantly reduce their overhead. The most important optimizations included in the Mitosis compiler and presented in this paper are branch pruning, memory and register dependence speculation, and early thread squashing.Performance evaluation of Mitosis compiler/architecture shows an average speedup of 2.2.
 
I'm thinking that the complexity of processors will increase very rapidly, but I doubt that the clock speed will. In other words, I have no doubt that by 2010, processors will be available that are four or five times as capable as an early Pentium 4, but I seriously doubt that raw clock speed is how it will be accomplished.

If you look at AMD's offerings, the clock speed they run at has stayed in the 2 - 2.5 or so GHz range for two or three years, but undeniably, an Athlon64 X2 is more capable than an AthlonXP, even if they're "clocked the same."
 
What you say about AMD clocks over last 2 years is true. Intel came out with a new tech that out of the box has shown 3.85ghz is easily attainable. when this new cpu matures i suspect 3.85ghz from intel stock and common o/c of 4.5 ghz. IF intel used the new tera herts transistor on 45nm I supect Penryn will Start life on the high end @ 4.5 ghz.

Than when they go to32nm 3d gates with high k(nehalem c) I would believe 5ghz. for the low end chips. But who knows? This will be With Penry on 45nm (4.5ghz) . Nehalem may be something all together differant. I think that we may see 5ghz a lot sooner than we think . Time will tell.
 
You must log in or register to reply here.
Back
Top