Rated CPU Speed Charts

The purpose of this thread is to help answer the question of how does CPU A match up against CPU B in terms of clock speeds. Additionally, I hope it helps to eliminate the common misconception that CPUs with higher clock speeds are faster than CPUs with slower clock speeds. If you want to see really detailed CPU charts, please check out Tom's Hardware Interactive CPU Charts. You can compare all kinds of CPU in real world performance. Very nifty little tool.

The idea to make this thread came from a suggestion from adinu in the thread What is your "rated" CPU speed?. Essentially, this thread asks the participants to go to the System Requirements Lab website, and run a simple test that will tell you what your rated CPU speed is. The System Requirements Lab website runs a little program that prints outthat "You have x performance rated at y", where x is the current speed of your CPU in GHz, and y is your rated speed. This thread will simply compile the data that is generated in other threads to help show what CPUs rated performances are according to the System Requirements Lab. Hopefully it will provide some fun results that will help give an insight into different processors, and the evolution of processors over time.

Lastly, it should be noted that just because a CPU may have a low rated performance compared to another, it does not mean it is a horrible CPU. This is just simply a way of trying to compare things on an "apples to apples" basis instead of "apples to oranges." In addition, keep in mind that this is just a synthetic benchmark, and I believe it only uses one Core for all of you have have two cores in your CPU, so obviously this benchmark does not tell the whole story.

As this tread develops, please feel free to post comments, concerns, recommendations, corrections in a constructive manner. Enjoy and thanks for contributions in advance!

I will post each CPU results in the following format. Also, as multiple people are submitting rated performances for the same CPUs, I will simply take the highest score if there are discrepancies in the contributions.

CPU Model-----Stock Clocks-----Rated Performance

---------------------------------------------------------------------------------------------------------------------------------------------------------------

Intel Laptop Processors

Intel Core 2 Duo
T7600-----2.33GHz-----6.28GHz
T7400-----2.16GHz-----5.51 GHz
T7200-----2.00GHz-----4.99GHz
T5600-----1.83GHz-----3.84GHz
T5500-----1.66 GHz----3.37 GHz


Intel Core Duo
T2050-----1.60GHz-----2.75GHz
T2300-----1.66GHz-----3.00GHz
T2250-----1.73GHz-----3.11GHz
T2400-----1.83GHz-----3.43GHz
T2500-----2.00GHz-----3.74GHz


Intel Pentium M
770-----2.13GHz-----3.00GHz
750-----1.86GHz-----2.79GHz

Intel Celeron M
350-----1.3GHz-----1.3GHz

Intel Pentium 4 Mobile
???-----1.60GHz-----2.39GHz

Intel Pentium 4
Prescott-----3GHz-----4.6GHz

---------------------------------------------------------------------------------------------------------------------------------------------------------------

AMD Laptop Processors

AMD Turion 64 X2
TL-60-----2.01GHz-----4.22GHz
TL-50-----1.60GHz-----3.11GHz

AMD Sempron
3300+-----1.99GHz-----3.30GHz

---------------------------------------------------------------------------------------------------------------------------------------------------------------

Intel Desktop Processors

Intel Quad Core
QX6700-----2.67GHz-----10.02 GHz

Intel Core 2 Duo
E6600-----3.0GHz-----8.58GHz (This is OCed, stock clocks are 2.4GHz)

---------------------------------------------------------------------------------------------------------------------------------------------------------------

AMD Desktop Processors

AMD Athlon 64
3200+-----1.99GHz-----3.20GHz

AMD Athlon XP
2600+-----2.09GHz-----2.60GHz
2400+-----2.00GHz-----2.40GHz

---------------------------------------------------------------------------------------------------------------------------------------------------------------

Please let me know what you think of this format. I am not sure the best/most useful way to put them. Let me know and I will be more than willing to change it.
I was also thinking of having the CPUs broken up into lists by type as shown above, but also to have a list of them in chronological order, from highest rated performance to lowest.

 

Intel Core 2 Duo (Merom) T7200 2.0GHz CPU=

Rated speed 4.99GHz

 

Intel Core 2 Duo T5600.

1.83 GHz Performance Rated at 3.84 GHz

(wow...there seems to be a BIG difference between the T7200 and T5600)

 

Intel Core 2 Duo (Merom) T7600 2.33GHz
Performance rated @ 6.28 GHz

 

Intel Core 2 Duo
T7200-----2.00GHz-----4.99GHz
T5600-----1.83GHz-----3.84GHz


Shows how much L2 does in benchmarking

 

This is all meaningless though, for a number of reasons. First, dual core processors do not actually function as these ratings dictate for both software related and hardware related reasons. They will probably in time function better, but will never achieve the equivalent speeds posted.

Second, the comparison is to the equivalent of a P4 GHz rating. The newer processors are equivalents to a higher GHz rated P4 for many tasks, but it is not equal amongt these tasks, and this comparison is only valid for those tasks that it is best at.

I just think these figures are very misleading, and a processor cannot be viewed so single-mindedly as they used to be. Despite that most people are aware that a processor cannot be viewed strictly by it's frequency rating, we all seem keen to try and compare them by the same, and now much less relevant measure.

 

Oh nice, systemrequirementslab have changed their rating system: my C2D T7200 2.0GHz was rated at 2,99 GHz, now it's rated at 4,94 GHz

 

You are correct to point out that these benchmarks, like any other benchmarks, need to be interpreted carefully.

However, apart from a couple of extreme results for Intel desktop CPUs, the results so far give a pecking order much as I would expect. It is useful to get another comparison between AMD and Intel CPUs and it is also useful to see that the faster Pentium M CPUs can keep up with the slower of the Core series. Unless I am mistaken, the benchmark is single-threaded so it gives a fair indication of the performance of a CPU with a single-threaded application (which most still are).

What the results don't say, but the experienced user knows, is that having an extra core on the CPU avoids the performance hit when one core is loaded with work so, in reality, the T2050 is likely to feel faster than the Pentium M 770.

John

 

I'm sorta surprised actually...

The turions compared to the core duos (not c2d) seem to be better at each speed. The 1.6 Turion is equivalent to a 1.73 cd, and the 2.0GHz Turion is quite a bit faster than the equivalent 2.0 cd.

 

^ Me too.
It will be interesting to see how the new dual-core drivers will affect the benchmarks.

 

core duo 1.66 ghz = 2.99 ghz

 

First off, that you for the constructive criticism. I really appreciate it that you didn't just say "This is all meaningless" and then stop at that. Thanks for exemplifiying. Second, I wouldn't go as far to say this is meaningless. There is something to be learned by everything. I did however, put a link to the Interactive CPU Charts on Tom's Hardware website. These charts will let you compare processors in many real world applications for more realistic results. I couldn't have summed it up better than what Jon Ratsey said "You are correct to point out that these benchmarks, like any other benchmarks, need to be interpreted carefully." That goes for this thread as well. I certainly don't want people picking processors solely based off of the performance ratings on this page. Hopefully, they will post in the "what notebook should I buy thread", and we can see what their needs really are and what laptop, or desktop processor would best suit them. I hope this alleviates some of your concern. If you want me to add a big disclaimer to the post, I will be more than happy to do it. Again, thank you for the constructive criticism, I appreciate it.

In addition, this isn't different at all than the thread "Measure your Notebook CPU Speed." In this thread, Super-Pi is used as the benchmark, which is also a single threaded application. Just a different way of benchmarking, but again, not necessarily a prediction of "real world" performance. For "real world" benchmarks, I would look at Tom's Hardware Interactive CPU Charts.

 

I came across a little strong, I meant no offense. I should have not stated that it was "...all meaningless..." even if I did offer some points that followed - I must admit, I did a poor job of reading throught the description of the thread and posted a knee-jerk reaction - apologies. Rather, I should have more gracefully stated that these ratings need to be taken with a grain of salt. I don't pretend to be particularly knowledgable about the operations of processors, but I just thought it prudent to point out to those who would look at this chart and assume something that may not be realistic given current software and hardware limitations.

Perhaps it would be more accurate to state that the speed ratings represent potential so as not to be confused with an absolute rating?

 

Also read this new article at Tom's Hardware on benchmarking mobile CPUs and then move onto the Interactive Mobile CPU Charts.

John

 

Good reads, thanks.

 

I like the chart. Nice to simplify things a bit. I realize that synthetic benchmarks don't always equate to reality, but they generally come close. I'm guessing they do decent job at rank ordering. They are also useful to compare across brands and generations. My only criticism is, can you flip the order of the core duo's to match the rest? You went slow to fast instead of fast to slow like the other charts. I do think its neat that they seem to base their score of AMD's desktop ratings. (i.e. 2600+ = 2.6GHz) I'll add my scores when I get home today

It may also be useful to rank order them, like below.

QX6700-----2.67GHz-----10.02 GHz
E6600-----3.0GHz-----8.58GHz (This is OCed, stock clocks are 2.4GHz)
T7600-----2.33GHz-----6.28GHz
T7200-----2.00GHz-----4.96GHz
P4-Prescott-----3GHz-----4.6GHz
TL-60-----2.01GHz-----4.22GHz
T5600-----1.83GHz-----3.84GHz
T2500-----2.00GHz-----3.74GHz
T2400-----1.83GHz-----3.43GHz
3300+-----1.99GHz-----3.30GHz
3200+-----1.99GHz-----3.20GHz
TL-50-----1.60GHz-----3.11GHz
T2250-----1.73GHz-----3.11GHz
770-----2.13GHz-----3.00GHz
T2300-----1.66GHz-----2.99GHz
750-----1.86GHz-----2.79GHz
T2050-----1.60GHz-----2.75GHz
2600+-----2.09GHz-----2.60GHz
2400+-----2.00GHz-----2.40GHz
P4-M-----1.60GHz-----2.39GHz
Celeron-----1.3GHz-----1.3GHz

*edit - NMV, I don't like the whole rank orderd look after all.

 

Bah! I can't believe my old 3.4GHz P4 @ 5.1GHz beats out the T7200 C2D @ 4.96GHz. Furthermore, what's up with the huge rating differential in the C2D T7xxx lineup in comparison with every other architecture? I suspect that extra core and 64 bitness is factoring in somehow. Take everything you get from THG with a grain of salt methinks.

 

^just to expand on what starcub said....whats the architectural differences between the T5600 and the T7200 C2D's? Just 2MB L2 cache?

 

^ adding to above, this problem is pretty common with synthetic benchmarks. They only scale so well. Once you get past the top end of what it was designed for, the measurement becomes pretty exagerated.

I just ran this on my 2 PC's and noticed that it isn't actually doing any testing (task manager never said my CPU utilization increased). It is just reading various system settings and comparing results to a pre-defined table. This means that is taking into account single, dual and multi-core processors. But it's all just based on someones pre-defined ideas.

Here's another problem I found. I would have expected the T7600 vs. T7200 to scale the same way, but they don't. The T7600 has a 16.5% clock rate increaseover the T7200 (2.33 vs. 2.00, no architectual diferences at all), but the rated CPU speed is more than 22% higher.

I think the rankings are pretty decent from everyones results so far. But you cannot say that something with a rated speed of 6.0 is twice as fast as something rated at 3.0. Still nice to see, but limited in usefullness. Here's my results.

Laptop
AMD ML-30 (1.60 GHz) rated at 2.0 GHz

Desktop
Intel E6300 (1.86 GHz) rated at 4.90 GHz

 

Don't worry I didn't take any offense at all and I couldn't agree with you more. Many people in this thread have the same concern. You stated it well "These rating should be taken with a grain of salt." While they are interesting, they obviously don't take many things into consideration coming up with the rating.

 

Core 2 Duo---1.66 GHz---3.37 GHz

Observing the results, the Cache makes the difference among the Core Duos. Impressive.

 

I noticed that as well, what exactly are the differences between the C2D's besides the clock speeds and the L2 cache?

 

Price

 

Is the Celeron-M a 300, 400 or 500 series and why is it a clock speed that is hardly ever found in mainstream notebooks?

 

The latest celeron is a 400 series I believe, based on the yonah core.

 

You Have: 2 processors running - AMD Turion(tm) 64 X2 Mobile Technology TL-56

You Have: 1.81 GHz Performance Rated at 3.66 GHz

 

Intel Core 2 Duo (Merom) T7400 2.16GHz
Performance rated @ 5.51 GHz

 

Thanks, added.

 

Intel Pentium D 925 (3.0GHZ)
performance rated at 5.53 GHz

 

The Mobile CPU Chart from Tom's hardware is still rather incomplete, and has been for some time. Hopefully they will continue adding to it, as I always thought their desktop chart was pretty informative from year to year.

Notebookcheck has a chart as well, although less detailed and also incomplete http://www.notebookcheck.net/Mobile-Processors-Benchmarklist.2436.0.html

 

I have a T5300 core 2 duo 1.73Ghz processor in my dell E1505

It is rated at 3.37Ghz

My Voodoo ENVY U709
Processor AMD Opteron 185 @ 2.6Ghz dual core
rated at 6.68 GHz

K-TRON

 

laptop:
You Have: 2 processors running - Intel(R) Core(TM)2 Duo CPU T7800 @ 2.60GHz
You Have: 2.59 GHz Performance Rated at 6.42 GHz

desktop:
You Have: 2 processors running - AMD Athlon(tm) 64 X2 Dual Core Processor 4000+
You Have: 2.11 GHz Performance Rated at 6.00 GHz

 

Intel C2D T5250 @1.5GHz rated at 2.92GHz

 

Pentium M 1.50 GHz Performance Rated at 2.24 GHz.

 

Those charts are 100% flawed. The speed ratings are absolutely incorrect when comparing different CPUs.

Unfortunatly this thread is a 100% complete misrepresentation if comparative CPU performance. OP, Please remove this misinformation as it will only confuse the general public. I will contact moderators to persue removal of this misinformation.

It is 100% clear this laughable benchmark is dependant on ghz as its also clear it only does simple mathematics to claculate speed which is the WORST way to benchmark CPUs.

Proof is in the laughably high score the Pentium 4 3.0ghz received compared to the other CPUs. The pentium 4 3.0ghz got such a high score becuase again, this flawed speed rating benchmark is dependant on ghz as it only does simple mathematics.

Again, OP, please remove these flawed rating. You should know better than to post such misinformation.

 

Also note that the AMD Turion x2 cpus are rated nearly the same clock for clock as core 2 duos. This is laughably incorrect and Im shocked this misinformation it being displayed.

 

I have a question, which would make Onion right. How come a 3.0Ghz pentium 4 has a rated speed of 3.4Ghz. What is the processor being compared with. I thought all of the processors were being compared to pentium 4 architecture. If so, as I thought how can a 3.0 have arated speed of 3.6. And how comea E6700 intel, which is 2.86% faster in mflops as my opteron scores 10+ghz, and mine scores 6.68Ghz.

I also agree that the program is BS.

It is really stupid to compare this way. If you want proven speeds, why dont we have a post where we all give our whetstonre and dhrystone mips and mflops scores from SIS Sandra. That would be more beneficial.

K-TRON

 

You forget the Hyper Threading ????

 

Which actually slowed down the processor for some applications. HT on the Pentium was flawed implementation by Intel and didn't work half as well in practice as in theory.

For once I have to agree with onion, SRL hasn't been known to put out good benchmarks while WPrime would be a much better way to test dual core CPU speed.

Edit: Just looked at the dates. Holy thread necromancy Batman.

 

Actually, wPrime is pretty crappy as well. Best is actual benchmarks in different applications and games, like Xvid encoding, H.264 encoding using different programs and encoders, vmw-encoding, 3DS max rendering, winrar compression, low resolution and quality gaming...