should I abandon the AMD ship? (Updated)

[tential]

Update: I am the proud owner of a i5 4690K and ASRock Z97 Extreme 4. Thanks for all the advice guys.

Literally the only logical choice to come out with if you're a person trying to spend a reasonable amount of money to get performance that will last generations to come.

I'd personally take the 4790k but that's because I'd never owned a top of the line intel nonextreme processor before.

 

[Cheesesticksandbeer]

Literally the only logical choice to come out with if you're a person trying to spend a reasonable amount of money to get performance that will last generations to come.

I'd personally take the 4790k but that's because I'd never owned a top of the line intel nonextreme processor before.

I almost did pick it up but decided to get the protection plan with microcenter and an i5 instead.

 

[crashtech]

Great hypothetical. The next logical question from a smart but not overly tech savvy computer shopper (assuming we're working that angle) is...

How many Bogomarks does it take to have a pleasing computing experience for the next X years? I assume someone looking at $500 pre-built PC's is trying to spend as little as possible (and Google hasn't tried to sell them a Chromebox yet lol).

Pleasing? I don't know. Obviously everyone has their own definition or there would be a lot less disagreement. If we are going by the "good enough" model that seems to have taken hold, one of the faster Core2 Duos retrofitted with an SSD and 4GB RAM might do the trick. That A4 in the earlier example has similar computing power and it is being sold today. But at least the Core2 will leave you with some bucks left in your pocket.

 

[crashtech]

Update: I am the proud owner of a i5 4690K and ASRock Z97 Extreme 4. Thanks for all the advice guys.

An interesting choice, given the excellent OC you had achieved on your 6300. Though the numbers will say your new choice is faster, I had wondered if it would be enough to actually be discernible in games and whatnot. I hope you will report in with benches from both systems, and some subjective impressions as well.

 

[Ramses]

one of the faster Core2 Duos retrofitted with an SSD and 4GB RAM might do the trick. That A4 in the earlier example has similar computing power and it is being sold today. But at least the Core2 will leave you with some bucks left in your pocket.

That's exactly what I see occurring with "normal" folks way more often than not. About that era CPU, the Core2, if not older. Or something AMDish from along then. Honestly I'm a "power user" I guess, and my old 955BE was fine, I put it in a new case with more ram and an ssd and sold it to my boss to replace some ancient Intel Dell box for half the cost of what the locals wanted to sell him, he's thrilled with it. Good enough is in the eye of the beholder I suppose. I'd hate to be trying to sell PC's these days.

 

[Cheesesticksandbeer]

An interesting choice, given the excellent OC you had achieved on your 6300. Though the numbers will say your new choice is faster, I had wondered if it would be enough to actually be discernible in games and whatnot. I hope you will report in with benches from both systems, and some subjective impressions as well.

Yeah im not rushing the build, im going to bench the hell out of my 6300 and then switch and bench. Should really provide me a good comparison. Going to bench about 8 games and the normal synthetic ones. 6300 @ 4.6 vs i5 4690k stock and whatever O/C i get.

All same hardware but MB/CPU should be a good comparison.

 

[Cheesesticksandbeer]

An interesting choice, given the excellent OC you had achieved on your 6300. Though the numbers will say your new choice is faster, I had wondered if it would be enough to actually be discernible in games and whatnot. I hope you will report in with benches from both systems, and some subjective impressions as well.

The large factor was GPU utilization on benchmarks. My cf 7970s is really the logic for the jump. 6300 in some benchmarks show poor utilization compared to i5. I really want to test this myself.

 

[MiddleOfTheRoad]

A hypothetical:

We have CPU A that costs $50 and scores 1000 Bogomarks, and CPU B that costs $100 and scores 1500 Bogomarks:

CPU A scores 20 Bogomarks per dollar, and CPU B scores only 15 Bogomarks per dollar. Looking at this alone, CPU A gives 33% more performance per dollar.

But then we put them into a complete system that costs $500, aside from the CPUs. System A costs $550, and System B costs $600:

It's still a dumb way to compare especially in regards to value. Why would anyone not simply try to buy the most powerful chip at the $50 price point? Or the the best chip for $100? Because that is how people shop when they are building systems around value.

Most people stretch their budget to get the best hardware they can afford..... You keep grasping at straws to find a way to make these flawed system build numbers actually prove meaningful. But that is impossible -- because the testing methods were broken from the start.

These system builds are inherently Apples To Oranges comparisons because of all the individual components that impact the results. Introducing a bunch of variables is an awful way to benchmark.

 

[Ramses]

It's still a dumb way to compare especially in regards to value. Why would anyone not simply try to buy the most powerful chip at the $50 price point? Or the the best chip for $100? Because that is how people shop when they are building systems around value.

Most people stretch their budget to get the best hardware they can afford..... You keep grasping at straws to find a way to make these flawed system build numbers actually prove meaningful. But that is impossible -- because the testing methods were broken from the start.

These system builds are inherently Apples To Oranges comparisons because of all the individual components that impact the results. Introducing a bunch of variables is an awful way to benchmark.

That why I steered your debate toward (really crappy) pre-built consumer systems. Because enthusiasts are the only ones building systems and reading per-component benchmarks. In that context, you are right, in the pre-built buy at Best Buy world, the other view is right. That's why I dismiss them out of hand, because they aren't my reality, or yours most likely. But they exist.

Ta-da...

 

[crashtech]

It's still a dumb way to compare especially in regards to value. Why would anyone not simply try to buy the most powerful chip at the $50 price point? Or the the best chip for $100? Because that is how people shop when they are building systems around value.

Most people stretch their budget to get the best hardware they can afford..... You keep grasping at straws to find a way to make these flawed system build numbers actually prove meaningful. But that is impossible -- because the testing methods were broken from the start.

These system builds are inherently Apples To Oranges comparisons because of all the individual components that impact the results. Introducing a bunch of variables is an awful way to benchmark.

I don't know how to better explain the simple arithmetic that I used to explain a very simple concept. Sorry about that, I'll try harder next time so that we can avoid these little misunderstandings.

 

[escrow4]

Pleasing? I don't know. Obviously everyone has their own definition or there would be a lot less disagreement. If we are going by the "good enough" model that seems to have taken hold, one of the faster Core2 Duos retrofitted with an SSD and 4GB RAM might do the trick. That A4 in the earlier example has similar computing power and it is being sold today. But at least the Core2 will leave you with some bucks left in your pocket.

Core 2 is rubbish in 2014. Buy an i3 4330 and it will slaughter it whilst sipping power. A G1820 is as fast and that is the beginning of Intel's stack. A random buyer will likely buy some craptop though and then wonder why it is so rubbish. RAM and an SSD won't help if you have an ageing architecture, and yes you can spot the difference.

 

[KingFatty]

I don't know how to better explain the simple arithmetic that I used to explain a very simple concept. Sorry about that, I'll try harder next time so that we can avoid these little misunderstandings.

I wanted to thank you for the explanation. It induced forehead-slapping.

But I get it now, thanks to your example. If you take it to extreme, I could see how you could build two diamond-encrusted super systems, say they both cost $10K each. The weird part is that one could have twice the performance of the other, due to a CPU price difference that becomes negligible.

Like if you compare two houses that both cost $500K. One house has a air-conditioning unit that is super-awesome for like $20K, but the other house has a super-crappy A/C unit that is only $10K. The house that costs $520K total will be awesome to live in, and way better than the $510K house.

Ultimately it's psychology too - when you are buying a $500K house, you are spending so much already, you want to spend extra to have the super nice A/C unit. Same with computer, if you know you are blowing a lot of money on the whole system, it's easier to swallow the individual price difference between components.

With CPUs, the analogy kind of breaks down, because the CPU pretty much drives the value of the entire system. It can really drag down an otherwise nice/expensive system that has an SSD and dual video cards etc. That's why I used the house/air-conditioner analogy, because if you mismatch an under-powered A/C unit to a house that's too big for it, it really makes the house miserable in the middle of summer, and you'll be willing to go to extremes to upgrade that system, relative to the big investment in the entire house.

 

[SlowSpyder]

Update: I am the proud owner of a i5 4690K and ASRock Z97 Extreme 4. Thanks for all the advice guys.

:thumbsup: Let us know what you think after you get it up and running!

 

[Ramses]

I wanted to thank you for the explanation. It induced forehead-slapping.

But I get it now, thanks to your example. If you take it to extreme, I could see how you could build two diamond-encrusted super systems, say they both cost $10K each. The weird part is that one could have twice the performance of the other, due to a CPU price difference that becomes negligible.

Like if you compare two houses that both cost $500K. One house has a air-conditioning unit that is super-awesome for like $20K, but the other house has a super-crappy A/C unit that is only $10K. The house that costs $520K total will be awesome to live in, and way better than the $510K house.

Ultimately it's psychology too - when you are buying a $500K house, you are spending so much already, you want to spend extra to have the super nice A/C unit. Same with computer, if you know you are blowing a lot of money on the whole system, it's easier to swallow the individual price difference between components.

With CPUs, the analogy kind of breaks down, because the CPU pretty much drives the value of the entire system. It can really drag down an otherwise nice/expensive system that has an SSD and dual video cards etc. That's why I used the house/air-conditioner analogy, because if you mismatch an under-powered A/C unit to a house that's too big for it, it really makes the house miserable in the middle of summer, and you'll be willing to go to extremes to upgrade that system, relative to the big investment in the entire house.

Unless you live where I do and never need AC.

 

[Ramses]

:thumbsup: Let us know what you think after you get it up and running!

Month from now he's going to be looking for more GPU to match that CPU. :biggrin:

 

[MiddleOfTheRoad]

Month from now he's going to be looking for more GPU to match that CPU. :biggrin:

No doubt -- I'm just as guilty. It really does border on obsession for some of us. I just moved to liquid cooling this week on my backup machine. Probably overkill -- but did it anyway.

 

[Ramses]

No doubt -- I'm just as guilty. It really does border on obsession for some of us. I just moved to liquid cooling this week on my backup machine. Probably overkill -- but did it anyway.

I tried to explain all the case fans in my last box to my wife, she's like "but it never gets hot here?"

 

[Cheesesticksandbeer]

So I had bent pins on the asrock board. Took it back and also took back my i5. I now have the same board, however, I said screw it and got the i7 instead! Windows installing now!

 

[crashtech]

Now you're talking. 4790K, I presume?

 

[Cheesesticksandbeer]

Now you're talking. 4790K, I presume?

Thats correct, after more research i felt with the increase in clocks alone it was likely worth the 80$ more. Hopefully I also get a good oc, but i dont feel as pressured to push it as I would with the i5.

 

[Ramses]

Buying with the expectation of the overclock as icing rather than cake is the wise man's way.

 

[VirtualLarry]

Core 2 is rubbish in 2014. Buy an i3 4330 and it will slaughter it whilst sipping power. A G1820 is as fast and that is the beginning of Intel's stack. A random buyer will likely buy some craptop though and then wonder why it is so rubbish. RAM and an SSD won't help if you have an ageing architecture, and yes you can spot the difference.

I don't know that I would call Core2 exactly "Rubbish". Aside from power consumption, it's still faster than most of the Bay Trail-M/D systems, even the quad-core ones, in web browsing "snappiness" (due to poor ST performance of Silvermont, compared to Core2).

I bought a couple of DC5800 refurbs, 2GB RAM, 2.33Ghz Core2 CPU (unsure which one, but probably the older one with 4MB L2 cache), and Win7 32-bit. With a semi-modern 80/160GB HDD (WD Blue, I think), it was actually quite snappy during web browsing. I didn't load it down with huge amounts of tabs during testing, but what little testing I did (with a wired connection, 30Mbit internet), it wasn't sluggish in any way that could discern.

Now, I wouldn't consider a system like that for modern gaming, but for web browsing, it's entirely adequate, and in most cases, better than Atom.

Now, if Bay Trail systems were cheap enough, to also include SSDs (in budget desktops and laptops - aka "craptops"), then there might be an argument for Atom. But as long as they keep throwing in 5400RPM HDDs along with them, I'll take a Core2 and a 7200RPM HDD all day long.

 

[Shivansps]

The problem with the C2D today is (i have 2 in use, a E2140 and a E5200), is the IGP, the E2140 i have it with a GT520 and it hadles very good at web browsing, videos and even some light gaming, but if i need to use the IGP of its MB (Biostar P4M890-M7 TE/Via Unicrome) all that goes to hell, it can barely hold his own.

The E5200 has a G31 mb and its great, it even runs W8 very well, browsing OK, but videos? he can play them, but its all done on CPU.

If someone have one of those rare 775 Geforce 9300/9400 MB its perfect to use a C2D with, but if like most people only have crappy mbs with G3x or less igp ill say that is a problem, BT its probably better.

Not sure about G4x.

For example the only reason that a slow 2650 does not seems so slow at first sight its his IGP.

 

[crashtech]

I wanted to thank you for the explanation. It induced forehead-slapping.

It's great to know I was helpful, thanks for saying so.

The problem with the C2D today is (i have 2 in use, a E2140 and a E5200), is the IGP, the E2140 i have it with a GT520 and it hadles very good at web browsing, videos and even some light gaming, but if i need to use the IGP of its MB (Biostar P4M890-M7 TE/Via Unicrome) all that goes to hell, it can barely hold his own.

The E5200 has a G31 mb and its great, it even runs W8 very well, browsing OK, but videos? he can play them, but its all done on CPU.

If someone have one of those rare 775 Geforce 9300/9400 MB its perfect to use a C2D with, but if like most people only have crappy mbs with G3x or less igp ill say that is a problem, BT its probably better.

Not sure about G4x.

For example the only reason that a slow 2650 does not seems so slow at first sight its his IGP.

That is very true about the IGP.

 

[Cheesesticksandbeer]

Last benchmark, heaven 4.0 extreme settings, 1080p

Min: 8.5
Max: 153.0
Avg: 67.0
Score: 1688

Tweaking stuff but just ran Heaven 4.0
Min: 8.6
Max: 166.4
Avg: 71.2
Score: 1793

Thats no overclock on cpu