What's new
Frozen in Carbonite

Welcome to FiC! Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

General CPU thread

Threadripper 2 has been released, here's a review:


Memory bandwidth hurts it a lot
 
Last edited:
We're hitting a plateau as far as technological innovation goes, it seems.
 
We're hitting a plateau as far as technological innovation goes, it seems.
People said the same thing when we hit the clockspeed hysteresis barrier back in the Pentium 4 days.
 
People said the same thing when we hit the clockspeed hysteresis barrier back in the Pentium 4 days.
Do you have anything to actually contribute to the thread or are you just trying to start a dogpile and make a mountain out of a mole hill? Notice I said "it seems" meaning that is my fucking opinion godammit. As you can see Vyor already made your post before you did, ergo, what you are doing is essentially spam.
 
Do you have anything to actually contribute to the thread or are you just trying to start a dogpile and make a mountain out of a mole hill? Notice I said "it seems" meaning that is my fucking opinion godammit. As you can see Vyor already made your post before you did, ergo, what you are doing is essentially spam.

No, no it isn't. You're just butthurt about it.

Also gaming performance was fixed in a nvidia driver update.
 
He basically repeated what you had already said a month ago. Yes, that is spam. And what you are doing is now spam as well.
 
Do you have anything to actually contribute to the thread or are you just trying to start a dogpile and make a mountain out of a mole hill? Notice I said "it seems" meaning that is my fucking opinion godammit. As you can see Vyor already made your post before you did, ergo, what you are doing is essentially spam.
He basically repeated what you had already said a month ago. Yes, that is spam. And what you are doing is now spam as well.

Seriously? Cool it off.
 
Seriously? Cool it off.
This thread isn't a debate about how fucking wrong IndyFront its the 'General CPU thread'. Period.

I will cool it off when they get back on topic and leave me alone.
 
Do you have anything to actually contribute to the thread or are you just trying to start a dogpile and make a mountain out of a mole hill? Notice I said "it seems" meaning that is my fucking opinion godammit. As you can see Vyor already made your post before you did, ergo, what you are doing is essentially spam.
Calm down Indy.
 
Do you have anything to actually contribute to the thread

Simple statement of fact: many thought that we couldn't possibly advance processors any faster than they were when 3.7GHz became the upper limit for Pentium 4's as the Prescotts simply generated too much heat and the processors would invariably short out from hysteresis unless more esoteric active cooling systems were used. And then we went multicore with more complex instructions running up the IPC combined with more cache and smaller construction processes. Any time someone says we've reached the performance limit, we invariably figure ways around it.

I suspect moving from silicon to, say, gallium arsenide, will yield another such bypass.
 
No, no it isn't. You're just butthurt about it.

Also gaming performance was fixed in a nvidia driver update.
If civility can't be maintained I will close this thread, and that goes for all of you.
 
Simple statement of fact: many thought that we couldn't possibly advance processors any faster than they were when 3.7GHz became the upper limit for Pentium 4's as the Prescotts simply generated too much heat and the processors would invariably short out from hysteresis unless more esoteric active cooling systems were used. And then we went multicore with more complex instructions running up the IPC combined with more cache and smaller construction processes. Any time someone says we've reached the performance limit, we invariably figure ways around it.

I suspect moving from silicon to, say, gallium arsenide, will yield another such bypass.

Also clockspeed kept going up anyway.
 
We're hitting a plateau as far as technological innovation goes, it seems.
I will agree with you as far as integration goes. There may be some advantages to using different materials, but certainly getting from 100 to 10 nm was easy compared to getting from 10 nm to anywhere else. OTOH there's plenty of room for innovation in architecture.
 
I will agree with you as far as integration goes. There may be some advantages to using different materials, but certainly getting from 100 to 10 nm was easy compared to getting from 10 nm to anywhere else. OTOH there's plenty of room for innovation in architecture.

It's getting harder to shrink, but it doesn't even need to change materials until like 2nm. There are 3 or 4 techs that are cooking for it that will help massive amounts.

The biggest near term one is T-FETs, which will allow super low voltage operation(like .3v or less) and high switching frequency.
 
Back
Top Bottom