TomsHardware: Core i7-3970X Extreme Review: Can It Stomp An Eight-Core Xeon?

[nitromullet]

Dude let's just forget it. You can't make an argument when they think we don't know math while we think they don't understand English...

LOL! Point taken.

 

[AtenRa]

Dude let's just forget it. You can't make an argument when they think we don't know math while we think they don't understand English...

Do we all understand that MORE time is SLOWER ???

Does 100 seconds are DOUBLE than 50 seconds ??? Does that makes the processor that needs 100 Seconds being 100% SLOWER than the processor that needs 50 seconds ???

Do we all agree to that ??? no matter if we know math or understand English ???

ps: IDC, you seriously have some problems, i will suggest you take the following, it works 100% (i have used it my self)

http://www.powerhealth.gr/en/product/mens-x-complex-2/?gclid=CKzC5eHu27MCFczHtAodejQAaw

 

[antihelten]

No. It means it does the job in 2x longer time. We measure time. Time. Need I write it one more time? Oh sorry,I already did
100% slower is actually 2x slower than the base you are using. Which is 50s in the (hypothetical) example original poster used.

Problem is that by using your definition, something that is 100% faster is in essence infinitely fast (i.e. CPU A is 50 sec and CPU B is 0 sec, (50sec-0sec)/50sec = 100%. Taking 0 sec is the same as infinitely fast), which to most people would seem counterintuitive, since it would imply that something can never be more than 100% faster (unless the CPU can travel backwards in time, by taking less than 0 seconds).

I would guess the reason is probably that the adjectives "faster" and "slower" is usually used to describe the relative speed of a subject, and speed is usually defined a some unit of work per some unit of time, and not as the inverse (i.e. some unit of time per unit of work), which is the metric used in the benches here.

 

[TemjinGold]

Do we all understand that MORE time is SLOWER ???

Does 100 seconds are DOUBLE than 50 seconds ??? Does that makes the processor that needs 100 Seconds being 100% SLOWER than the processor that needs 50 seconds ???

Do we all agree to that ??? no matter if we know math or understand English ???

ps: IDC, you seriously have some problems, i will suggest you take the following, it works 100% (i have used it my self)

http://www.powerhealth.gr/en/product/mens-x-complex-2/?gclid=CKzC5eHu27MCFczHtAodejQAaw

We agree on these things:

1) More time is slower.
2) 100 seconds is double 50 seconds.
3) The intel is twice as fast.
4) The amd is half the speed (or 50% of the speed).

However, you seem to not understand the meaning of "100% slower." This is why I said you think we don't know math and we think you don't know English.

Here's an example: If you make $50,000 per year and I make 100% less than you, I would make $0. I would not make $25,000.


Another example: If you took one of the CPUs and I asked you to "slow it down 100%," you would bring it to a stop.

 

[Haserath]

Do we all understand that MORE time is SLOWER ???

Does 100 seconds are DOUBLE than 50 seconds ??? Does that makes the processor that needs 100 Seconds being 100% SLOWER than the processor that needs 50 seconds ???

Do we all agree to that ??? no matter if we know math or understand English ???

ps: IDC, you seriously have some problems, i will suggest you take the following, it works 100% (i have used it my self)

http://www.powerhealth.gr/en/product/mens-x-complex-2/?gclid=CKzC5eHu27MCFczHtAodejQAaw

Instead of saying slower, just say it takes 100% more time.

50% slower is 100% more time.

100% slower is an infinite amount of time.
---
(Is just happy he restarted an off-topic war^_^)

 

[Abwx]

100% slower than something would mean it is doing no work at all.

You are confusing with 100% less performing...

100% slower means it will take double the time to do the task ,
that is , +100% of the base time wich in your exemple is 50 seconds.

I didnt read this thread theses days , i wasnt expecting
that this primary school maths was still at stakes and even
became a new thread within the thread.

Of course , it would be irrelevant to send it in the
"Highly Technical " part of this forum...

Edit : For still better understanding , given a 50 seconds base ,
being 200% slower means 150 sec , while 1000% slower means
550 seconds and so on....

 

[Abwx]

100% slower is an infinite amount of time.
---

Wrong , read above why.....

50% slower is 100% more time.

100% slower means 100% more time, period.

heh, Resistance is Futile

Their fishy maths argument just exploded in the wings....

 

[nitromullet]

100% slower means 100% more time, period.

No. both "slower" and "faster" do not refer to time alone. They refer to speed, which is a function of time.

Speed = distance/time.

On AT speed is slightly more abstract because we don't usually talk about the speed of cars or other physically moving vehicles. We usually refer to speed as some unit of work/time. Frames/second is probably the most common example.

The percentage refers to the work units done per second.

• If cpu A is doing 100% less work units per second than cpu B, cpu A is not doing any work.
• If cpu A is doing 50% less work units per second than cpu B, cpu A is 50% slower than cpu B, and cpu B is 100% faster than cpu A.

 

[AtenRa]

No. both "slower" and "faster" do not refer to time alone. They refer to speed, which is a function of time.

Speed = distance/time.

On AT speed is slightly more abstract because we don't usually talk about the speed of cars or other physically moving vehicles. We usually refer to speed as some unit of work/time. Frames/second is probably the most common example.

The percentage refers to the work units done per second.

• If cpu A is doing 100% less work units per second than cpu B, cpu A is not doing any work.
• If cpu A is doing 50% less work units per second than cpu B, cpu A is 50% slower than cpu B, and cpu B is 100% faster than cpu A.

The original post clearly speaks of TIME in MINUTES.

50% better performance? How about some facts from the article? Time to completion of whole THG test suite(ST and MT workloads) for 3970x is 91.5 minute. FX8350 took 116 minutes . From this we have: 91.5/116=0.788 or ~21% faster than 8350. 3770K is similarly 12% faster than 8350 when you take its total runtime in the equation.

 

[Abwx]

No. both "slower" and "faster" do not refer to time alone. They refer to speed, which is a function of time.

Yes , but you are confusing slowness and fastness , going as wrong
as using the same formulae for the two definitions while one is the
reciprocal of the other.

If cpu A is doing 100% less work units per second than cpu B, cpu A is not doing any work.

Doing 100% less work is not is not the same as being 100% slower ,
it means that it is 100% less faster....

Doing 100% less work means that it takes an infinite time to do the
work done , hence it is infinitly slower , so it will take an infinite amount
more % (of the reference time) to do the job.

Slowness and fastness are mathematicaly unbounded in their superior limit , that is , infinitly slow and infinitly fast have both a meaning.

As such , infinitly faster will be expressed by an infinite amount of % ,
while infinitly slower will be equally expressed by an infinite amount of %.

 

[TemjinGold]

Okay promise I will stop after this post because it's pointless.

Those of you arguing against nitromullet and me: The problem is I can tell based on your posts neither of you are native English speakers. Why? Because things like "less faster" are not proper English. This is why we won't ever understand you and you won't understand us. As I explained, you cannot resolve an argument when one side thinks math is the problem while the other side thinks English is the problem. I don't mean this to be mean but you guys simply DO NOT UNDERSTAND ENGLISH PROPERLY. That is the problem. Now let's please put this to rest.

 

[AtenRa]

Okay promise I will stop after this post because it's pointless.

Those of you arguing against nitromullet and me: The problem is I can tell based on your posts neither of you are native English speakers. Why? Because things like "less faster" are not proper English. This is why we won't ever understand you and you won't understand us. As I explained, you cannot resolve an argument when one side thinks math is the problem while the other side thinks English is the problem. I don't mean this to be mean but you guys simply DO NOT UNDERSTAND ENGLISH PROPERLY. That is the problem. Now let's please put this to rest.

It has nothing to do with understanding English or not, it has to do about acknowledging the fact that the OP was talking about TIME and Less time is FASTER.

Taking LESS time to do the same thing is FASTER, now when people argue that faster is not about time but work done in the same time i beginning to believe that someone here is trolling or, im sorry to say, he or she is stupid.

Also, people are too proud to admit they are wrong in this forum. Nobody knows everything(except me).

 

[Idontcare]

Okay promise I will stop after this post because it's pointless.

Those of you arguing against nitromullet and me: The problem is I can tell based on your posts neither of you are native English speakers. Why? Because things like "less faster" are not proper English. This is why we won't ever understand you and you won't understand us. As I explained, you cannot resolve an argument when one side thinks math is the problem while the other side thinks English is the problem. I don't mean this to be mean but you guys simply DO NOT UNDERSTAND ENGLISH PROPERLY. That is the problem. Now let's please put this to rest.

It has nothing to do with understanding English or not, it has to do about acknowledging the fact that the OP was talking about TIME and Less time is FASTER.

Taking LESS time to do the same thing is FASTER, now when people argue that faster is not about time but work done in the same time i beginning to believe that someone here is trolling or, im sorry to say, he or she is stupid.

Also, people are too proud to admit they are wrong in this forum. Nobody knows everything(except me).

1) Let me be the first to admit I am wrong, I asked for confirmation and can now confirm that I was wrong when I posted "I am 50% shorter than I claim", apparently there is a temperature effect in play that I was not accounting for, let's call it "shrinkage" for lack of a better term at the moment, and with this being the start of the winter season the cold factor apparently makes my situation more like "60% shorter" :|

2) Regardless the reasoning for the disconnect between faster/slower, longer/shorter, it is in ALL of our best interests to come to an agreement on the matter of accepted convention that is to be employed within these hallowed halls of the anandtech forums

We are an international community, we will forever be butting heads over terminology and vernacular unless we come to some middle ground and try to see eye-to-eye.

Personally what I'd like to see are some proposals on the mathematical definition for what we shall adopt as a community on what we mean when we use the words faster and slower, then we can iterate until we all agree.

This is not one of those things where "agreeing to disagree" is a viable outcome because it will just continue to resurface ad nauseum down the road.

If we can leave out the jabs and snarky quips that would be better too. This isn't about schooling the mathematically ignorant americans or slamming down the ESL (english as a second language) members.

So lets hash this out.

We have two different kinds of benchmarks, one type is the type that results in a number where the higher number is "better" i.e. "faster"
Data Type 1:

Computer A finishes a task in 10s resulting in benchmark score of 10 A.U. (arbitrary units)

Computer B finishes a task in 5s resulting in benchmark score of 20 A.U. (arbitrary units)

The second type is the type that results in a number where the lower number is "better" i.e. "faster"
Data Type 1:

Computer A finishes a task in 10s resulting in benchmark score of 10 A.U. (arbitrary units)

Computer B finishes a task in 5s resulting in benchmark score of 5 A.U. (arbitrary units)

Goal - develop an objective method for ascribing adjectives and numerical percentages which capture and correctly convey the relative performance differences between Computer A and Computer B regardless whether we are assessing the results of Type 1 benchmarks (higher is better) or Type 2 benchmarks (lower is better).

With that, I yield the floor to those of you who wish to weigh in with your proposals

 

[guskline]

If it aids in this "dispute" I humbly bow with my lowly FX8350 at 4.6Ghz in homage to the Kingly Core I7-3970X Extreme!

Look at this link for benchmarks comparing the 8350 to the 3960x which is slightly slower if you don't believe me. http://www.anandtech.com/bench/Product/697?vs=443

The 3930/3960 and now 3970 are really the UBER chips.

 

[kevinsbane]

We have two different kinds of benchmarks, one type is the type that results in a number where the higher number is "better" i.e. "faster"
Data Type 1:


The second type is the type that results in a number where the lower number is "better" i.e. "faster"
Data Type 1:


Goal - develop an objective method for ascribing adjectives and numerical percentages which capture and correctly convey the relative performance differences between Computer A and Computer B regardless whether we are assessing the results of Type 1 benchmarks (higher is better) or Type 2 benchmarks (lower is better).

With that, I yield the floor to those of you who wish to weigh in with your proposals

Speed; higher = better
Car 1: 100 km/h.
Car 2: 200 km/h.
Car 2 is twice as fast (ie, it is 100% faster) than car 1.

Time: lower = better
Car 1 travels 100 km in 1 hour.
Car 2 travels 100 km in 1/2 hour.
Car 2 is still twice as fast as car 1. Car 2 completes a task in 50% of the time.

Referencing to something being faster should be a a reference to speed alone. A CPU is faster; it is an inherent property of what it is.
Referencing to something doing something in a shorter time period should be a reference to time alone. A CPU can complete a task in 5 seconds (instead of 10); it is a property of what it can do.
A car that is 100% faster, can complete the same task in 50% of the time.

EDIT: I realised that a CPU can complete a task 100% faster and mean the same thing as a CPU is 100% faster. A car that drives 100 m twice as fast as another car means it is going twice the speed as that other car. Clarified...

 

[TemjinGold]

IDC: In the interest of diplomacy, here is a suggestion: We can't seem to agree on faster/slower mostly because of the nuances of the language. What if our official convention is simply "X is y% the speed of Z" and leave it at that? That is something everyone seems to readily agree on.

Would also like to add--I didn't mean to insult anyone when bringing up the English issue. Merely pointing out that the problem arose because they are using "terminology" that isn't proper English (and therefore does not have proper meaning.) When you ascribe meaning to such and then base an argument on it, there is no basis for sound argument. This was in no way meant to be a "we're better than them" thing.

 

[Haserath]

We have two definitions of slower:

1) A function of time ie 100%-2x the length of time, 200%-3x, etc.

2) A function of 'speed'(positive velocity/time) ie 50%-half the speed(twice as much time), 75%-a quarter of the speed(4x the time), etc.

To me, if you're moving 100% slower, you're not moving at all(def #2).

You can move infinitely faster, but you cannot move infinitely slower since you can't move slower than 'no movement'.

 

[HutchinsonJC]

Time: lower = better
Car 1 travels 100 km in 1 hour.
Car 2 travels 200 km in 1/2 hour.
Car 2 is still twice as fast as car 1. Car 2 completes a task in 50% of the time.

I think on the Car 2 there is some inaccuracy.

If Car 2 were to be *twice* as fast as Car 1, then wouldn't you only double the km or halve the time as opposed to both?

Car 2 should travel 100 km in 1/2 hour to be twice as fast as Car 1

or

Car 2 should travel 200km in 1 hour to be twice as fast as Car 1.


Edit:

If I am driving 70MPH, and I slow down by 100%, then 100% of 70MPH is the removal of all 70MPH; I have stopped moving.

If I am driving 70MPH, and I speed up by 100%, then 100% of 70MPH is the addition of another 70MPH; I have doubled my speed.

100% slower = stopped

100% faster = doubled

 

[Abwx]

To me, if you're moving 100% slower, you're not moving at all(def #2).

You can move infinitely faster, but you cannot move infinitely slower since you can't move slower than 'no movement'.

The confusion is because you re doing an amalgam between "X% slower" and "Y% less" (time...)

You can move as much slower as you want it will still be superior
to 0 wich is the boundary of "infinitly slow".

Hence , being 1000% slower means a 10 ratio in "slowness".

It s the same as saying that it takes 90% less time , see ??..

 

[kevinsbane]

I think on the Car 2 there is some inaccuracy.

If Car 2 were to be *twice* as fast as Car 1, then wouldn't you only double the km or halve the time as opposed to both?

Car 2 should travel 100 km in 1/2 hour to be twice as fast as Car 1

or

Car 2 should travel 200km in 1 hour to be twice as fast as Car 1.

If it makes more sense,

Car 2 is still twice as fast as car 1. Car 2 completes a task in 50% of the time (compared to car 1).

You can travel twice as fast and complete the same task in half the time. In your example...
Car 2 travels 100 km in half an hour - it goes at 100/0.5 = 200 km/h.
Car 2 travels 200 km in an hour - it goes at 200/1 = 200 km/h.

Both statements are true at the same time. Car 2 is still twice as fast as car 1. Basically, doing something twice as fast means doing it in half the time. They are two ways of saying the same thing.

 

[kevinsbane]

The confusion is because you re doing an amalgam between "X% slower" and "Y% less" (time...)

You can move as much slower as you want it will still be superior
to 0 wich is the boundary of "infinitly slow".

Hence , being 1000% slower means a 10 ratio in "slowness".

It s the same as saying that it takes 90% less time , see ??..

So how would you like to define "100%" faster?

 

[Abwx]

100% slower = stopped

100% slower is twice as slow , mind you...

So being twice as slow would be stopping ?...

Actually it is going at half the speed.

So how would you like to define "100%" faster?

As twice as fast or half as slow.

 

[HutchinsonJC]

Both statements are true at the same time.

I'm not so sure.

Time: lower = better
Car 1 travels 100 km in 1 hour.
Car 2 travels 200 km in 1/2 hour.
Car 2 is still twice as fast as car 1. Car 2 completes a task in 50% of the time.

Translation:
Car 1 travels 100km in 1 hour = 100km per hour.
Car 2 travels 200km in 30min = 400km per hour.

 

[kevinsbane]

I'm not so sure.



Translation:
Car 1 travels 100km in 1 hour = 100km per hour.
Car 2 travels 200km in 30min = 400km per hour.

I stand corrected.

 

[HutchinsonJC]

100% slower is twice as slow , mind you...

So being twice as slow would be stopping ?...

Actually it is going at half the speed.

I'm sorry, but I do not think I would ever use English in that way.

If I wanted to say something was going at half speed relative to something else, I would not say that it was going twice as slow. I would say it's going half as fast; or 50% slower.