Intel Core Family

[jigger]

If your aim is to build a system for benchmarking DX11 games at 1920x1080 then the highest binned Intel quad core or maybe hex core paired with a Titan is the way to go. Move away from that scenario and have a lot more options.

Intel are selling b0rked chips. They need to stop that IMO.

 

[passifid]

i'm always amazed with the amount of 144 and 160+hz gamers on here.
Unless you do those talking about FPS gains is 100% pointless as both will do 60fps in their sleep.

After that is for content creator or general use. You again wouldn't a difference in general use. And if you are a pro content creator you likely use less than 5 specific applications and you should do it on a case by case basis.

Because of solder, higher overall power and less security flaws its likely my next processor will be AMD for the first time ever. But honestly if the 8700k was £250 enough to make them the same price for a system it would really be a toss up.

 

[Martini1991]

*removed all but the important part because the forum is caching quotes and readding them after removing them and making a mess of the whole thing.*




Your slide reads like this:

8700K @ 4Ghz = 1325
2600X @ 4Ghz = 1384 (+4.5%)

So in MT Ryzen 2 has 4.5% higher IPC, if you read what i wrote you will see that i guesstimated 6%, i wasn't far off. 1.5% percent off.

Your reply doesn't actually address what I've said (Further reason why any reasonable debate with you falls flat).
You've just used it to promote your own agenda and bang on about the "Ryzen's higher IPC in MT".

 

[humbug]

Your reply doesn't actually address what I've said (Further reason why any reasonable debate with you falls flat).
You've just used it to promote your own agenda and bang on about the "Ryzen's higher IPC in MT".

Its the same thing that i have been saying all the way through, from start to finish i'm saying the same thing, what exactly is your argument with me? your disagreement?

 

[Martini1991]

Its the same thing that i have been saying all the way through, what exactly is your argument with me? your disagreement?

I don't care anymore.
Maybe I will in a few more days. But I'm done for today.

 

[humbug]

I don't care anymore.
Maybe I will in a few more days. But I'm done for today.

This is what started all this, all the way through i maintained the same argument, even you provided material to backup exactly what i said, you don't disagree with this, so what exactly is your problem?

Using one singular thread yes, as you say "single core IPC", because of better multicore scaling the IPC is on par if not even slightly ahead, certainly clock for clock in Cinbenench Ryzen is a few % ahead in MT.

Tell me what in the real world outside of benchmarking applications uses one sole core? nothing.

 

[ArmitageShanks]

There is no such thing as an application written to use solely one compute thread, other than one or two benchmarking application for academic purposes.

It's a single core performance metric. Your misguided notions do not change that fact. And yes, single-threaded applications do still exist because writing a multithreaded application is not trivial (I do it for a living). Renderers are actually one of the easiest applications to multithread as the work can be easily chunked. That's not the case with realtime applications.

Ever see an application cause one of your cores to hit 100% while the rest are completely idle?

 

[MagicBoy]

Can you even explain that? why would IPC measurement have to be restricted to single core, if a CPU with 12 compute threads is faster than another CPU also with 12 compute threads then does it also not have higher IPC?

Yup. TL;DR - I'm right, you're wrong. My computer science degree trumps your crazy AMD fanboi tendencies as it's a real world qualification.

 

[humbug]

It's a single core performance metric. Your misguided notions do not change that fact. And yes, single-threaded applications do still exist because writing a multithreaded application is not trivial (I do it for a living). Renderers are actually one of the easiest applications to multithread as the work can be easily chunked. That's not the case with realtime applications.

Ever see an application cause one of your cores to hit 100% while the rest are completely idle?

Ok, self proclaimed expert, talk is cheap, anyone can say anything they like to bolster their arguments, i put it to you that is exactly what you are doing, so name 5 conmen applications used today that are restricted to one singular thread.

To make the point i'll easily name 10 that use all available threads.

Substance Designer.
Rhino 3D.
Blender.
World Machine.
Z Brush.
Cinema 4D.
Handbreak.
7Zip.
WinRar.
iClone.
Shadermap.
VUE xStream.

I went a little over there, never mind, there's so many to chose from.

 

[MagicBoy]

What does a list of multithreaded applications have to do with a theoretical architecture efficiency measurement?

 

[sideways14a]

Yup. TL;DR - I'm right, you're wrong. My computer science degree trumps your crazy AMD fanboi tendencies as it's a real world qualification.

Is this the best time for me to mention most of the comp science students i see go through here would have trouble tying there shoelaces?

 

[humbug]

What does a list of multithreaded applications have to do with a theoretical architecture efficiency measurement?

Your serial box computer science qualification not teach you that? You spent £50 over the price of the 4 core 4 thread i5 7600K with a fancy name on a platform that costs twice as much as the X270 platform your i5 comes from and you got nothing for it.

To answer your question, everything, i know you would rather it be ignored but nothing is tied to singular thread performance, not since about 2003.

Here is your 7600K 7740X in action... this is what it has to do with.

Please don't presume to lecture me because you claim to have a qualification, even if you do its meaningless and your choice in hardware doesn't put you in good stead for real knowledge.

 

[8 Pack]

5 pages for no reason... Dear me..

 

[David Bisset]

It's a single core performance metric. ...

No, it's a vague concept that has never been an actual metric as it is workload driven and is just used as an arguing point. Being workload driven, to claim it is either single core only or multi core only are both daft options.

Amusingly, 'IPC' varies quite a lot with clock speed too even though it should be the same, as too many other aspects of the system and task are relevant. Some tasks give roughly linear response but many don't. Anyway, I'll leave you guys to your argument as everyone will pick the meaning they want to prove what they want.

 

[humbug]

No, it's a vague concept that has never been an actual metric as it is workload driven and is just used as an arguing point. Being workload driven, to claim it is either single core only or multi core only are both daft options.

Amusingly, 'IPC' varies quite a lot with clock speed too even though it should be the same, as too many other aspects of the system and task are relevant. Some tasks give roughly linear response but many don't. Anyway, I'll leave you guys to your argument as everyone will pick the meaning they want to prove what they want.

Right, This argument is about the meaning of IPC and the truth is there are so many real life variables that need to be qualified to it it becomes meaningless.

This is just my personal opinion but the whole concept of IPC when applied to modern architectures is flawed, the performance you get per clock depends on a huge array of variables and those variable do not compare across different vendors.

IPC is a concept that belongs, ironically, to a time when CPU's only had one core and didn't do much beyond counting zeros and ones.

 

[ArmitageShanks]

No, it's a vague concept that has never been an actual metric as it is workload driven and is just used as an arguing point. Being workload driven, to claim it is either single core only or multi core only are both daft options.

Amusingly, 'IPC' varies quite a lot with clock speed too even though it should be the same, as too many other aspects of the system and task are relevant. Some tasks give roughly linear response but many don't. Anyway, I'll leave you guys to your argument as everyone will pick the meaning they want to prove what they want.

IPC is an average figure quoted by manufacturers, but it does not change with clock speed. Single core performance (on the same instruction architecture) is measured by clock speed multiplied by the IPC. The IPC is how many instructions the architecture can do per clock cycle - it's in the name.

This is why Intel and AMD quote the IPC for families of CPUs, not individual CPU models. IPC is not a benchmark in itself, but it certainly does affect real world performance.

 

[humbug]

IPC is an average figure quoted by manufacturers, but it does not change with clock speed. Single core performance (on the same instruction architecture) is measured by clock speed multiplied by the IPC. The IPC is how many instructions the architecture can do per clock cycle - it's in the name.

This is why Intel and AMD quote the IPC for families of CPUs, not individual CPU models. IPC is not a benchmark in itself, but it certainly does affect real world performance.

True but don't you think they do this because its a nice easy way to describe the differences between their own architectures?

AMD certainly did this comparing Zen with Excavator, but i think if they pulled up Techspots @ 4Ghz IPC comparison Intel would point out that its not quit as simple as that, and they would be right.

 

[ArmitageShanks]

Ok, self proclaimed expert, talk is cheap, anyone can say anything they like to bolster their arguments, i put it to you that is exactly what you are doing, so name 5 conmen applications used today that are restricted to one singular thread.

To make the point i'll easily name 10 that use all available threads.

Substance Designer.
Rhino 3D.
Blender.
World Machine.
Z Brush.
Cinema 4D.
Handbreak.
7Zip.
WinRar.
iClone.
Shadermap.
VUE xStream.

I went a little over there, never mind, there's so many to chose from.

Most of these are not realtime applications. They deal with static workloads. Try maybe an emulator or two. Or most games for that matter.

 

[MagicBoy]

Your serial box computer science qualification not teach you that? You spent £50 over the price of the 4 core 4 thread i5 7600K with a fancy name on a platform that costs twice as much as the X270 platform your i5 comes from and you got nothing for it.

Oh dear, I appear to have touched a nerve. No need to get offensive though.

Not sure what a serial box is? Don't recall them on the shelf in PC World. I'll assume you just can't spell when angry and you meant Cornflakes .... I wasn't aware the University of Salford was selling qualifications via that method. I could have saved many years of valuable learning and knowledge had I known I could pick one up at Tesco's with a pint of milk. Could have saved £40 on that book on Microprocessor architecture as well.

As for the kit in my sig, let me explain. TL;DR - your pricing assumptions are incorrect and I need single core performance. Total cost for CPU and board was less than a new 7700K CPU.

It's used for iRacing which bottlenecks on the graphics rendering thread with triples or VR, which uses a single core only. More IPC and more clock speed is king and there's no need for any more than 4 cores at the moment. For this usage Kaby-Lake X can be a bit of a bargain. I picked the 7740X up secondhand for a lot less than the price of a new 7600K, seller had used it for less than a week to test a video editing machine before buying an i9. Motherboard was on half price offer and cost less than a similar Z270 board. If I need more cores, then I can happily upgrade to a 7820X or faster in the future.

If I'd built a Ryzen rig for iRacing then I'd get similar performance at best to my 5 year old Haswell machine. For general usage I'd happily buy a Ryzen though, no fanboi tendencies here - the 7740X rig even has a Vega in it!

 

[humbug]

Oh dear, I appear to have touched a nerve. No need to get offensive though.

Not sure what a serial box is? Don't recall them on the shelf in PC World. I'll assume you just can't spell when angry and you meant Cornflakes .... I wasn't aware the University of Salford was selling qualifications via that method. I could have saved many years of valuable learning and knowledge had I known I could pick one up at Tesco's with a pint of milk. Could have saved £40 on that book on Microprocessor architecture as well.

As for the kit in my sig, let me explain. TL;DR - your pricing assumptions are incorrect and I need single core performance. I'm

It's used for iRacing which bottlenecks on the graphics rendering thread with triples or VR, which uses a single core only. More IPC and more clock speed is king and there's no need for any more than 4 cores at the moment. For this usage Kaby-Lake X can be a bit of a bargain. I picked the 7740X up secondhand for a lot less than the price of a new 7600K, seller had used it for less than a week to test a video editing machine before buying an i9. Motherboard was on half price offer and cost less than a similar Z270 board. If I need more cores, then I can happily upgrade to a 7820X or faster in the future.

If I'd built a Ryzen rig for iRacing then I'd get similar performance at best to my 5 year old Haswell machine. For general usage I'd happily buy a Ryzen though, no fanboi tendencies here - see it's even got a Vega in it!

It does not have a Vega in it, that's Kabylake-G. what you have is actually very old nVidia IP implemented badly.

Edit you talking about the APU's?

If you got the stuff cheap, cheaper than the 7600 and the X270 then fair enough, can't really argue with that, conveniently, surely though when looking at second hand you could have picked up a 7600K for even less, same with the motherboard.
There are no benchmarks for iRacing that i can see, so i cannot confirm or deny your claims about, personally i doubt it wouldn't benefit from a proper i7 and for VR? is iRacing the only thing you play because a 4 core 4 thread i5 is one of the worst CPU's you can have for high frame rates because it fluctuates wildly causing huge dips up and down and that's the last thing you want.

@ArmitageShanks those applications are quite real.