First Build Since 2009, Please Give Some Guidance

[Essence_of_War]

My thought there was that in terms of future-proofing, more software will likely use more cores as time goes by. Also the current main reason for the new build is the multi-threaded trading optimizations, so more cores can't hurt. My wallet hurts, though.

You paid $1k for a CPU just to "future-proof"? Return it. Future-proofing is a concept that exists only when the more expensive part is still actually good value for money.

I agree with lehtv. You doubled the cost of your CPU and didn't get anywhere close to doubling the performance. If you haven't opened it yet, I'd return it for the 5820 or 5930, and pocket the $500.

 

[James999]

You paid $1k for a CPU just to "future-proof"? Return it. Future-proofing is a concept that exists only when the more expensive part is still actually good value for money.

"Just to" future-proof? Not exactly, no.

1) To get the max performance from an enthusiast-level processor for the multi-thread optimization work that I am building this system for;

2) To allow 8 overclocked cores, which will likely be useful in more and more games, Photoshop, and other software over time;

3) To retain 40 PCIe lanes for eventual addition of M.2 storage (and for the eventual addition of multiple GPUs if I go game-crazy again at some point in the future)

4) To future-poof the system as much as possible today, looking forward over 5+ years before I need to build a system again.

Value is in the eyes of the beholder. I am comparing this CPU to a Xeon at twice the price, so in that sense it is a bargain, can be overclocked to approximate Xeon performance, etc.

 

[Essence_of_War]

2) To allow 8 overclocked cores, which will likely be useful in more and more games, Photoshop, and other software over time;

Unless you spend a lot of time doing CFD, PIC, or similar work, this is by no means clear.

3) To retain 40 PCIe lanes for eventual addition of M.2 storage (and for the eventual addition of multiple GPUs if I go game-crazy again at some point in the future)

The 5930k does that at half the cost.

4) To future-poof the system as much as possible today, looking forward over 5+ years before I need to build a system again.

Future-proofing is not a thing. The only way to future proof, is to save money today and spend it in the future. There is always something cool/sexy/better right around the corner in the hardware world. You can't stop that tide by overspending now.

 

[James999]

Future-proofing is not a thing. The only way to future proof, is to save money today and spend it in the future. There is always something cool/sexy/better right around the corner in the hardware world. You can't stop that tide by overspending now.

While I agree with you, I want to build a system right now. I debated putting the build off for 2-3 months, knowing that bigger, better, faster, cheaper is always right around the corner. For me it comes down to building now and having the best system I can afford. Maybe "future-proofing" is the wrong term. Perhaps "now-maxing" would be better for my situation.

I'm still somewhat back-and-forth about whether I should have grabbed the 5930K instead, so can't really disagree with that comment.

 

[Charlie98]

Maybe "future-proofing" is the wrong term. Perhaps "now-maxing" would be better for my situation.

Meh... you got what you wanted and are within your budget. Nothing exceeds like excess. Absolute performance has a cost, even if it is not generally realized. I have an i5 in my business PC... excessive? Probably, but I paid the money anyway. (That is not to say it was a $1K CPU, either. )

 

[RaistlinZ]

Meh... you got what you wanted and are within your budget. Nothing exceeds like excess. Absolute performance has a cost, even if it is not generally realized. I have an i5 in my business PC... excessive? Probably, but I paid the money anyway. (That is not to say it was a $1K CPU, either. )

Aye. This is what the OP wanted, and it's a sweet system. Let's let him enjoy it. :biggrin:

 

[Danrr]

If I could afford it, I'd buy everything top of the line.

That's why luxury products exist, you can get a cheaper one that can do the job but it's not best out there, like sport cars.

 

[lehtv]

If I could afford it, I'd buy everything top of the line.

That's why luxury products exist, you can get a cheaper one that can do the job but it's not best out there, like sport cars.

Sports/luxury cars are status symbols, and it is precisely their excessive cost that makes them status symbols. Not so with processors; a top of the line processor isn't expensive so that you could wave it around and let people see how successful you are. It's expensive because there are people who have - pardon me, but 'tis the truth - thick enough skulls that they don't care, and don't want to care, that they're grossly overspending just to get a marginal, temporary improvement in performance, and zero gain in status.

Intel don't sell processors to benefit the customer, they sell them to make money. A customer who manages to rationalize the purchase of a $1000 processor is just playing straight into Intel's pocket; it is precisely these customers who allow Intel to get away with charging $1000 for a marginally better processor in the first place. Although if the rationalization is something like "I just want to give Intel money", then that's perfectly fine.

 

[Termie]

Sports/luxury cars are status symbols, and it is precisely their excessive cost that makes them status symbols. Not so with processors; a top of the line processor isn't expensive so that you could wave it around and let people see how successful you are. It's expensive because there are people who have - pardon me, but 'tis the truth - thick enough skulls that they don't care, and don't want to care, that they're grossly overspending just to get a marginal, temporary improvement in performance, and zero gain in status.

Intel don't sell processors to benefit the customer, they sell them to make money. A customer who manages to rationalize the purchase of a $1000 processor is just playing straight into Intel's pocket; it is precisely these customers who allow Intel to get away with charging $1000 for a marginally better processor in the first place. Although if the rationalization is something like "I just want to give Intel money", then that's perfectly fine.

The 5960X is precisely 33% more powerful than a 5930K at the same clock speed. I'm sure the OP took that into consideration and decided it was worth spending 80% more to get it. This cannot be compared to past years when the Extreme processor has been fundamentally identical to a lower-end processor at half the cost. The 5960X is not just "marginally" better. If time is money, 33% faster is worth a lot of money.

 

[James999]

It's expensive because there are people who have - pardon me, but 'tis the truth - thick enough skulls that they don't care, and don't want to care, that they're grossly overspending just to get a marginal, temporary improvement in performance, and zero gain in status..

Envy is ugly.

Intel don't sell processors to benefit the customer, they sell them to make money.

Actually it is both. This is how free markets work.

 

[Charlie98]

Actually it is both. This is how free markets work.

Yup. Something is only worth what someone is willing to pay for it.

 

[mfenn]

Intel don't sell processors to benefit the customer, they sell them to make money. A customer who manages to rationalize the purchase of a $1000 processor is just playing straight into Intel's pocket; it is precisely these customers who allow Intel to get away with charging $1000 for a marginally better processor in the first place. Although if the rationalization is something like "I just want to give Intel money", then that's perfectly fine.

I thought about it a little more, and since the OP is using the CPU to do analysis for stock purchases, he can presumably do more (or more timely) analysis with the faster CPU. Depending on the volumes he's working with, he might make back that $500 in a month.

 

[lehtv]

The 5960X is precisely 33% more powerful than a 5930K at the same clock speed. I'm sure the OP took that into consideration and decided it was worth spending 80% more to get it. This cannot be compared to past years when the Extreme processor has been fundamentally identical to a lower-end processor at half the cost. The 5960X is not just "marginally" better. If time is money, 33% faster is worth a lot of money.

How can you be so sure he took that into consideration? Just because you would've doesn't mean he did. Plenty of people with lots of money care little about "is this a good deal" because they don't need to - it makes no difference in the end if it's not as good of a deal, they'll still have enough money for whatever else they need and want in the short term, and they can feel good about having the best product, even if it cost an arm and a leg.

But perhaps I should've done my research better, I didn't realize it was 33% faster. Regardless, two of my points stand - (1) comparing top of the line processors to expensive sports cars is silly because processors aren't status symbols, and (2) the reason the best consumer processor is so ridiculously expensive is because there are enough customers don't care about value for money that Intel can get away with charging so much for it. Basically the reason is the same as with top of the line graphics cards, and almost top of the line anything in the hardware world. Trying to rationalize such purchases with anything other than "I buy the best because I can afford to" is pointless.

Envy is ugly.

Envy has nothing to do with this, although I'm not surprised that someone whose decision I've just criticized would criticize me right back.

Actually it is both. This is how free markets work.

Their products benefit the customer only insofar as it helps them make more money, just as with any corporation. Benefiting the customer serves the purpose of benefiting Intel, but often corporations will take advantage of the customer as well.

I thought about it a little more, and since the OP is using the CPU to do analysis for stock purchases, he can presumably do more (or more timely) analysis with the faster CPU. Depending on the volumes he's working with, he might make back that $500 in a month.

I thought he said

The trading optimizations are for hobby trading, not my day job, so not mission critical.

 

[James999]

How can you be so sure he took that into consideration? Just because you would've doesn't mean he did. Plenty of people with lots of money care little about "is this a good deal" because they don't need to

Well let me assure you that I did take into consideration the cost. I have enough money to build a $2,300 machine that will suit my needs, but I work as a math professor, not a hedge fund manager. I paid $300 more (not $500) for a processor that has 2 more cores and 5MB more cache. This will speed up the optimization work I'm doing. It will also, in all likelihood, extend the useful life of my system.

This was not a status-driven decision, as you previously implied. My last build was more than five years ago. Hopefully my next build will be ten years from now. This tradeoff between going cheaper now with more frequent upgrades, or going all-in now (within reason) and with no need to upgrade for an extended time is something anyone doing a new build has to sort out for themselves.

But perhaps I should've done my research better, I didn't realize it was 33% faster.

Hmm.

Envy has nothing to do with this, although I'm not surprised that someone whose decision I've just criticized would criticize me right back.

To be accurate, your criticism was directed at my "thick skull". This was a personal attack, not a criticism of the system build. In my experience, when someone resorts to personal attack, it is often a sign of emotionality entering the conversation, instead of rational argument.

Finally, the objective of hobby trading is still to make money, just not as a full-time job. And, yes, the ability to more quickly analyze/optimize a trade set-up does sometimes translate into greater profit. It is also much less annoying than waiting 20 minutes to get the results back from your computer, especially when you may need to tweak your code numerous times to finish refining an idea.

I will happily discuss facts, even in the face of differing, even strong opinions. But I would much rather discuss a build based on its merits (and costs) than trade insults.

 

[lehtv]

A very fair and level headed reply. What I said wasn't intended as a personal attack, as it was not directed at you personally. Though I can see how it could be interpreted that way, apologies for that. Good luck with your build :thumbsup:

 

[James999]

A very fair and level headed reply. What I said wasn't intended as a personal attack, as it was not directed at you personally. Though I can see how it could be interpreted that way, apologies for that. Good luck with your build :thumbsup:

Not a problem, and thank you for the well wishes.