Multicore (12+) build machine

KyussBishop

I'm doing programming work with a project that has insanely long compile times, and requires frequent full-recompiling (which is wrecking my workflow), so I'm looking at putting together a powerful build machine that will be able to handle this, but put together at a decent value.


For this, I am probably going to go with dual (or maybe quad, depending on cost) 6/8 core Xeon's, and I'm wondering if anyone has experience putting together something like this (particularly for programming), who can advise me on:
What the best value Xeon is?

What is better value/performance wise; a dual processor setup with more expensive and/or higher-core/Ghz processors, or a quad processor setup with less expensive and/or lower-core/Ghz processors? (factoring in things, like cost of quad motherboard vs dual and such)

Where are good places for buying this equipment, that are good value? (and places that are good for reading up on putting together this kind of setup too)

Any other considerations I should be making? (I assume bog-standard memory is good enough and would not be a potential bottleneck)

What is a good way to keep this future proof? At some stage, will probably use it as a gaming computer, so don't want any limitations that e.g., limit what future graphics cards I can throw into it, or prevent me upgrading memory. For something this powerful, I should probably want to be using it almost a decade from now.


I'd put the budget at €1500-2000, but would be happy to trim it lower, or go higher, if it means significantly better value or speeds for me; stuff like graphics and any other needed extras, will be bought as cheap as possible.

BloodBath

To be honest you could end up spending a fortune there for only a small increase over a strong single processor like the 4930k. 6 cores/12 threads should give you a lot of power. Stick a good cooler on that and overclock it and you will have even more.

Dual cpu boards alone are around €400-500. 2 x 6 core xeons are going to be another €1000 or more. 2 x 8 core xeons is going to be around €1500 That's over €1500-€2000 for the cpu's and board alone.

You should fill this in.

1. What is your budget? [€xxx]

2. What will be the main purpose of the computer? [Gaming/Video editing/3D Modelling/HTPC/Internet] (If gaming include which games)

3. Do you need a copy of Windows? [Yes/No]

4. Can you use any parts from an old computer? [Hard drive/DVD Drive/Case/PSU/etc.] (If possible state brand and model of the parts to ensure compatibility)

5. Do you need a monitor? [Yes/No]

5a. If yes, what size do you need. [19'/20'/22'/24'/etc.]

5b. If no, what resolution is your current monitor and do you plan to upgrade in the near future? [1920x1080/1440x900/etc.] [Yes/No]

6. Do you need any of these peripherals? [Keyboard/Mouse/Wireless Card/Card Reader/Speakers/etc.]

7. Are you willing to try overclocking? [Yes/No]

8. How can you pay? [Bank Transfer/Credit Card/Laser]

9. When are you purchasing? [In x days]

10. If you need help building it, where are you based? [South Dublin City/Cork City/Kerry/etc.]

KyussBishop

Well, I'll be using it for (very) frequent (re)compiling of code, which is one of the rare things that really does spread spectacularly well across multiple cores/processors, so I'd like to try find the best mix of that; less processors/cores and more Ghz, may get me less bang for the buck than more processors/cores at a lower Ghz, and such (so am curious about finding the right configuration/mix/balance of Ghz/core#/processor# for my needs).

I'll be looking to build this myself as well (have done this plenty before with past computers), so focusing solely on motherboard/processors(/memory perhaps too) keeps it simpler, as I can figure out the rest of the peripherals easily enough.
This includes buying it all myself as well - I'd like to find out about different places I can buy this kind of equipment, that I can shop around from (standard consumer places don't seem to sell this level of equipment).

So:
1: I'd put my budget at €2000 upper-end (but willing to go above it, if the best value setup is just a little ahead of my budget) - can consider this as just for the motherboard/processors, to keep it simple
2: Compiling code primarily at first (may throw a decent graphics card in it down the line, and I would like to keep it as future-proof as possible); more cores/processors the better
7: If overclocking can save me a few bob, I may do that.

The rest I'd put as 'other peripherals', that I can figure out easily enough.

rd1izb7lvpuksx

Are you using CI? Is your build server distributable? I have a build farm for a Java/C++ shop, and we're using using multiple 2-socket Dell servers with Xeons.

uberpixie

KyussBishop wrote: »

I'm doing programming work with a project that has insanely long compile times, and requires frequent full-recompiling (which is wrecking my workflow), so I'm looking at putting together a powerful build machine that will be able to handle this, but put together at a decent value.

What are you currently using hardware wise?

BloodBath

Well my point is 2 x 6 core xeons will only only run at 2.6ghz (3.1 turbo). The 4930k overclocked to 4.5 - 4.7ghz is going to be around 65-75% faster than a single xeon and not much slower than 2. The 8 core xeons run at 2ghz. I think an overclocked 4930k offers by far the better value.

You could build something like this and still have enough left to get a good graphics card and whatever else you need.

Item|Price
Intel Core i7-4930K, 6x 3.40GHz, bx|€509.16
16GB-Kit G.Skill PC3-10667U CL9-9-9-24|€134.39
Samsung SSD 840 EVO Basic 120GB SATA 6Gb/s|€92.07
WD Caviar Blue 1TB 6Gb's|€53.74
Corsair Professional Series HX850 80PLUS Gold 850W|€127.37
Corsair Hydro Series H100i|€98.99
Nanoxia Deep Silence 140 - 1100|€9.05
Nanoxia Deep Silence One Dark Black, ATX, ohne Netzteil|€94.19
ASRock Fatal1ty X79 Professional, Sockel LGA2011, ATX|€235.08
Shipping|€18.99
Total|€1373.03

KyussBishop

Beatrice Witty Womb wrote: »

Are you using CI? Is your build server distributable? I have a build farm for a Java/C++ shop, and we're using using multiple 2-socket Dell servers with Xeons.

What is CI? Unfortunately, it's not distributable, because the project depends upon Visual Studio (and I have no control over that); I've looked at using IncrediBuild to distribute compilation (distcc can't work with VS), but at the amount that costs (and the 'rents' it costs in renewal in the future), I think it's better just to build a machine.

That setup you have is interesting/similar; what are the specs of those servers, and how much did it put you back?

KyussBishop

uberpixie wrote: »

What are you currently using hardware wise?

Just a bog-standard laptop/computer, with an i7-940QM (not very powerful), and a Q9650; I need multiples of the current speed I get with these though, which means (since it's not distributable) a computer with multiple processors and more than quad cores.

KyussBishop

BloodBath wrote: »

Well my point is 2 x 6 core xeons will only only run at 2.6ghz (3.1 turbo). The 4930k overclocked to 4.5 - 4.7ghz is going to be around 65-75% faster than a single xeon and not much slower than 2. The 8 core xeons run at 2ghz. I think an overclocked 4930k offers by far the better value.

You could build something like this and still have enough left to get a good graphics card and whatever else you need.

Item|Price
Intel Core i7-4930K, 6x 3.40GHz, bx|€509.16
16GB-Kit G.Skill PC3-10667U CL9-9-9-24|€134.39
Samsung SSD 840 EVO Basic 120GB SATA 6Gb/s|€92.07
WD Caviar Blue 1TB 6Gb's|€53.74
Corsair Professional Series HX850 80PLUS Gold 850W|€127.37
Corsair Hydro Series H100i|€98.99
Nanoxia Deep Silence 140 - 1100|€9.05
Nanoxia Deep Silence One Dark Black, ATX, ohne Netzteil|€94.19
ASRock Fatal1ty X79 Professional, Sockel LGA2011, ATX|€235.08
Shipping|€18.99
Total|€1373.03

I've been researching a bit based on this, and I might go for a dual setup of a processor comparable to (if not the same as) that, so that's a good processor recommendation.

However, I've been reading up on Quick Path Interconnect, which seems to be important for multi-processor setups; I can't seem to find a straight answer on this while googling, is QPI (possibly with multiple channels) a requirement for dual processor setups, or just an optimization, or does it depend upon the motherboard? (it doesn't strike me as the kind of thing code compilation would benefit a whole lot from)

If QPI isn't a requirement, I will probably go with a dual setup of that processor.

Gumbi

An overclocked 4930k will have over twice the clock speed, and be ~30% faster per clock speed, than the CPU you're currently running. In addition, it has 6 cores/12 threads, as opposed 4 cores/8 threads. It would be several times faster than your current setup easily. I don' think there's any need for a pair of them at all.

rd1izb7lvpuksx

KyussBishop wrote: »

What is CI? Unfortunately, it's not distributable, because the project depends upon Visual Studio (and I have no control over that); I've looked at using IncrediBuild to distribute compilation (distcc can't work with VS), but at the amount that costs (and the 'rents' it costs in renewal in the future), I think it's better just to build a machine.

That setup you have is interesting/similar; what are the specs of those servers, and how much did it put you back?

CI is continuous integration, a method of reducing the impact of many commits on large projects. If you need to retain your current build scheme then a single box is the way to go. If you can distribute, go for that. We spent about €1,500 per box and we have eight boxes at the minute. I'm not sure on the specs, but we have no issue with latency.

BloodBath

An overclocked 4930k will be about 4-5 times faster than a stock speed q9650.

srsly78

KyussBishop wrote: »

Just a bog-standard laptop/computer, with an i7-940QM (not very powerful), and a Q9650; I need multiples of the current speed I get with these though, which means (since it's not distributable) a computer with multiple processors and more than quad cores.

You should really look at improving your build process first. Throwing hardware at the problem is an inefficient solution on it's own. For example, what version of visual studio are you using? Some versions will only use 1 core...

Why are you doing a full rebuild all the time? Why aren't you using Jenkins CI with multiple slaves? Is your project one monolithic blob of code? It's better to break stuff up into multiple smaller projects.

gctest50

Could you try out a few different "sizes" of web based servers to get an idea of cores vs. speed etc ?

http://goparallel.sourceforge.net/personal-supercomputing-with-windows-parallel-studio-part-1/
http://goparallel.sourceforge.net/personal-supercomputing-with-windows-parallel-studio-part-2/

KyussBishop

Gumbi wrote: »

An overclocked 4930k will have over twice the clock speed, and be ~30% faster per clock speed, than the CPU you're currently running. In addition, it has 6 cores/12 threads, as opposed 4 cores/8 threads. It would be several times faster than your current setup easily. I don' think there's any need for a pair of them at all.

I need more; as I said, I'm doing compilation of code, and I'm looking at up to an hours compile time right now. Twice my current speed, will be 30 mins (still too long), four times that will be 15 minutes (approaching acceptability), and eight times would hopefully be within 7-10 (good enough for regular compiling).

BloodBath

You won't get a board that supports dual 4930k's with overclocking. The only one's I've seen support xeons only.

KyussBishop

srsly78 wrote: »

You should really look at improving your build process first. Throwing hardware at the problem is an inefficient solution on it's own. For example, what version of visual studio are you using? Some versions will only use 1 core...

Why are you doing a full rebuild all the time? Why aren't you using Jenkins CI with multiple slaves? Is your project one monolithic blob of code? It's better to break stuff up into multiple smaller projects.

I don't control the build process (it's not my project), and it's about as efficient as it will get within Visual Studio, as it's a very large project; it doesn't require a full recompile every time (it's modular and dynamically linked - and overall heavily optimized), but I will be making regular changes to source that triggers a recompile of most of the project, due to dependancies.

I've researched distributed builds, and other methods of speeding this up, and the only compatible method is IncrediBuild, which would cost me the bulk of the price of a build machine (with less of the benefit).

So, due to the specific needs of this project, a heavy duty build machine is needed for me to get work done efficiently (and I'd say I'll be working on this for many years, so it pays off).

KyussBishop

BloodBath wrote: »

You won't get a board that supports dual 4930k's with overclocking. The only one's I've seen support xeons only.

Okey; and do you know if dual Xeon's require support for QPI? The Xeon 1650 version 2, seems pretty close in spec/price to the 4930k (though a little hard to get), but does not have any QPI channels.

BloodBath

You're not going to get the kind of performance you're looking for even with dual 8 core xeons. You might get around 12-13 mins compile time with the 4930k. I'd say you won't get much better than 9-10 mins with dual 8 core xeons @ 2ghz which would cost €1300 more.

Afaik you need qpi for dual processors. That's the same reason the 4930k can't be used. It doesn't have a qpi link.

Even a pair of the fastest 8 core xeons that run at 2.7ghz and cost €1500 each are only going to give you about 50-60% more performance than the 4930k running at 4.5ghz.

It's just estimates anyway. There's no benchmarks directly comparing the 2. I'm just roughly working out figures based on the max performance of those cpus. Depending on how the compiler handles multiple threads and clock speeds is a factor as well. Assuming it can make the most of both choices the figures shouldn't be too far off.

-edit- This is the closest thing I could find to benchmarks comparing them. The top €2000 xeon is about the same performance as an overclocked 4930k so 2 of those may give you around double the performance but that would be €4000 on cpu's alone.

http://www.cpubenchmark.net/cpu_list.php

KyussBishop

Ya I've been checking out that site earlier, but unfortunately they are missing all of the just-released (September) v2 Xeon's, so it's possible to do better, just no handy site to determine likely performance.

A dual setup of the best 8 core processor on that site though, would give 3x the speed of the 4930 (though 4 times the price); either way, I'll likely have to relax my budget somewhat.

EDIT: Here is their current set:
https://en.wikipedia.org/wiki/List_of_Intel_Xeon_microprocessors#.22Ivy_Bridge-EP.22_.2822_nm.29_Efficient_Performance_2

KyussBishop

Dual Xeon 2687W's (8 core), may work well for me; ignoring price for a moment, how do they strike you performance wise? They seem a bit of an oddity, in having high power requirements, but think they'd do the job.

BloodBath

KyussBishop wrote: »

Ya I've been checking out that site earlier, but unfortunately they are missing all of the just-released (September) v2 Xeon's, so it's possible to do better, just no handy site to determine likely performance.

A dual setup of the best 8 core processor on that site though, would give 3x the speed of the 4930 (though 4 times the price); either way, I'll likely have to relax my budget somewhat.

EDIT: Here is their current set:
https://en.wikipedia.org/wiki/List_of_Intel_Xeon_microprocessors#.22Ivy_Bridge-EP.22_.2822_nm.29_Efficient_Performance_2

The v2's are about 10% faster clock for clock.

That was a stock speed 4930k though. It has about 25-30% overclocking headroom which would push it's score/performance pretty even with that 3ghz 8 core xeon.

Those cpu's are near €2000 each. Is it really worth an extra €4000 for double the performance?

The cheaper options don't come close. I'll stick by the opinion that the best value to performance by a long long way would be a single overclocked 4930k.

Gumbi

BloodBath wrote: »

The v2's are about 10% faster clock for clock.

That was a stock speed 4930k though. It has about 25-30% overclocking headroom which would push it's score/performance pretty even with that 3ghz 8 core xeon.

Those cpu's are near €2000 each. Is it really worth an extra €4000 for double the performance?

The cheaper options don't come close. I'll stick by the opinion that the best value to performance by a long long way would be a single overclocked 4930k.

Basically this. Easily 4.4 - 4.6ghz on good cooling.

KyussBishop

Okey, I may go with that then, and just upgrade my main desktop (which I need to do anyway); if it doesn't cut down the compile times enough for me though, may still have to look at a more expensive dual processor setup, but makes sense to at least try this out first (won't go to waste if I don't end up using it as main build machine).

Thanks for the help/advice BloodBath/all.

Cuddlesworth

From someone who actually has dual Xeons(x5650) and access to some i7 workstations, the cost for dual procs rules out the marginal benefit they give. A over clocked i7 as said above is a much better purchase.

Can you not offload compute time to Azure now at a low cost using TFS?

srsly78

KyussBishop wrote: »

I don't control the build process (it's not my project), and it's about as efficient as it will get within Visual Studio, as it's a very large project; it doesn't require a full recompile every time (it's modular and dynamically linked - and overall heavily optimized), but I will be making regular changes to source that triggers a recompile of most of the project, due to dependancies.

I've researched distributed builds, and other methods of speeding this up, and the only compatible method is IncrediBuild, which would cost me the bulk of the price of a build machine (with less of the benefit).

So, due to the specific needs of this project, a heavy duty build machine is needed for me to get work done efficiently (and I'd say I'll be working on this for many years, so it pays off).

I have cleaned up lots of industrial size builds lately, it sounds like your build is a mess.

Recompiling dependencies: WHY? If these stay the same just compile them once then put the binary into version control. Compiling stuff like boost can take a long time, which is wasted because most people only need the headers.

Also, if dependencies change only a header change should trigger a full rebuild. A DLL changing just means you need to link again. You should post more about this on the developer forum.

Get your boss to spend money on engineers rather than hardware, it's more cost effective.

KyussBishop

Cuddlesworth wrote: »

From someone who actually has dual Xeons(x5650) and access to some i7 workstations, the cost for dual procs rules out the marginal benefit they give. A over clocked i7 as said above is a much better purchase.

Can you not offload compute time to Azure now at a low cost using TFS?

Unfortunately there is no good way to distribute the load of this particular project, without something like IncrediBuild (in fact, I think that specifically is the only thing that can in this circumstance), and all of that is out of my control.

KyussBishop

srsly78 wrote: »

I have cleaned up lots of industrial size builds lately, it sounds like your build is a mess.

Recompiling dependencies: WHY? If these stay the same just compile them once then put the binary into version control. Compiling stuff like boost can take a long time, which is wasted because most people only need the headers.

Also, if dependencies change only a header change should trigger a full rebuild. A DLL changing just means you need to link again. You should post more about this on the developer forum.

Get your boss to spend money on engineers rather than hardware, it's more cost effective.

I've been programming for well over a decade, so I know enough not to make basic mistakes like this. It is exactly frequent work with headers, that most of the project depends upon, that is requiring me to put together a powerful build machine.

The build process is as good as it will get (and the people who put it together work with a heavily optimized game engine, so they know how to streamline this stuff), with the main exception being lack of a non-commercial way to distribute the builds (it supports IncrediBuild, at which price I may as well just put together a build machine).

With the workflow being optimized for Visual Studio, there won't be a more practical way to do this, unless someone recodes a tool like distcc, to work with Visual Studio's cl.exe.

smacl

I use VS2010 for large C++ builds based on an 4ghz 8 core AMD set-up with 32gb or RAM, and it is surprisingly fast once you identify and iron out all the bottle necks. Key things to look at are

Precompiled headers are being used effectively across all libraries

/MP is enabled, for multi-core processing

Lots of available memory

Intermediate and temp files going to a ram disk

While the bottleneck for me on this is CPU at this stage, you need to make sure that it would be for your own situation prior to lashing out big bucks on a multi Xeon build. Also worth noting that if you're using the same PC for development and debugging, some build configurations such as Edit & Continue do not support multi-processor building. For this reason, I have separate rigs for building releases, carrying out automated regression testing, and building installations, none of which are that well specced. IMO, with so many template heavy libraries in use these days, for C++, not getting precompiled headers right is the main source of slow build speed.

srsly78

smacl wrote: »

I use VS2010 for large C++ builds based on an 4ghz 8 core AMD set-up with 32gb or RAM, and it is surprisingly fast once you identify and iron out all the bottle necks. Key things to look at are

Precompiled headers are being used effectively across all libraries

/MP is enabled, for multi-core processing

Lots of available memory

Intermediate and temp files going to a ram disk

While the bottleneck for me on this is CPU at this stage, you need to make sure that it would be for your own situation prior to lashing out big bucks on a multi Xeon build. Also worth noting that if you're using the same PC for development and debugging, some build configurations such as Edit & Continue do not support multi-processor building. For this reason, I have separate rigs for building releases, carrying out automated regression testing, and building installations, none of which are that well specced. IMO, with so many template heavy libraries in use these days, for C++, not getting precompiled headers right is the main source of slow build speed.

No no... his build is "perfect" and everyone he works with is an expert, didn't you read? So good they don't even need continuous integration, impressive.

smacl

KyussBishop wrote: »

I've been programming for well over a decade, so I know enough not to make basic mistakes like this. It is exactly frequent work with headers, that most of the project depends upon, that is requiring me to put together a powerful build machine.

Sounds like a couple of days sorting out your dependencies and optimizing pre-compilation would improve build times by an order of magnitude at a fraction of the cost. As per srsly78's point, throwing hardware at what sounds like a software problem rarely solves it. There's a bucket load of good stuff on speeding up C++ builds on Stackoverflow, which is worth looking at. Before lashing out a small fortune on hardware, you should also profile your build to see where the bottleneck actually is.

Of course if you just fancy a high end beast of a machine and have money to burn, go for it. We all like our toys, nothing wrong with that, and in my thirtieth year as a full time dev I still like nice shiny new PCs. I've three dual xeon workstations here, albeit they're getting long in the tooth. These days I have different parts of the build process on different PCs, which is fast, robust and cost effective.

srsly78

You can even get jobs as a "build engineer" these days, which means optimising and working with CI a lot. Optimizing a game engine for runtime performance and optimizing a build system are two completely different things...

KyussBishop

It's a project that (accumulatively) probably has well over a hundred people working on it at the moment (including a whole build team); it's as efficient as it will get.

I've checked out a few different ways of optimizing the speed locally, and processor power is definitely the bottleneck - combine that with restrictions from the toolset the project uses (making distributed compiles costly - using Incredibuild), and the best solution for me is a faster processor (which is what the thread is about - not about the build setup).

smacl

Sorry, but not buying it...

A C++ software project built using Visual Studio that involves over a hundred people in development is obviously going to be made up of large number of smaller Visual studio projects in one or more solutions that build the components required. Anyone with any amount of build experience would tell you that the fastest way to get this built is across a number of PCs in parallel. Even if you had to rebuild the whole lot (unlikely) on a very regular basis, doing so across a large array of PCs with one or more projects per PC is going to result in a total build time that is an order of magnitude faster than what you could hope to achieve on even the fastest dual or even quad xeon set-up. The fact that such a project would be run without some kind of CI in this day and age, and that you have a build team just does not add up.

Putting one fast PC under a single developer in this scenario, even if they are a lead developer, doesn't seem liable to significantly influence overall build times.

farna_boy

Would you consider just upgrading your graphics cards and using them to process the data?

I know it is applicable in some scenarios, especially so when you can do parallel processing but I'm not sure if it would work in your case. Nvidias CUDA seems to be the basis for a lot of this (from the quick search I did) but it may be a viable alternative considering how much the Xeons will cost.

KyussBishop

I'm not asking for advice on the project guys, I asked for advice on the computer - which I got, and don't need more help with.

Honestly, if you want to offer help to someone, start by not making unfounded assumptions about what they are doing, and then ignoring the person when they tell you your assumptions are wrong (thus wasting everyones time).

smacl

KyussBishop wrote: »

Honestly, if you want to offer help to someone, start by not making unfounded assumptions about what they are doing, and then ignoring the person when they tell you your assumptions are wrong (thus wasting everyones time).

You started off with a budget of €1500-€2000 looking for a 12 core system to improve build performance on a large visual studio app. You talk about compilation causing a bottleneck as you are mucking about a lot with header files -> you're talking about a C++ project (possibly C but unlikely). You've never heard of CI, yet there are a hundred people working on this project which includes a dedicated build team. You've been told by multiple posters that high end multi-core processors will most likely not provide a reasonable return on investment, and have been provided with a number of cost effective alternatives.

The PC you're talking about is about 3-4 times over your budget and will give you at best about 60%-75% increase in build time over the PC that is inside your budget. Anyone with any nous at build optimization for large C++ projects would likely achieve a 300%-500% improvement for the same budget, and given that you haven't even heard of continuous integration, my guess is you know exactly squat about the subject of build optimization yourself.

And we're wasting your time? :rolleyes:

btkm8unsl0w5r4

Op check out this eview to see general performance http://www.tomshardware.co.uk/precision-t5600-xeon-e5-2687w-workstation,review-32805.html
I am sure such a big project would consider 10k for a machine small change. If they won't buy you a machine they shouldn't trust you to compile the code
If not the 4930k with quad channel memory is a real workhorse, I have one totally stable with all cores at 4.6ghz on a closed loop watercooler.

Gumbi

Sunny Fluffy Stockroom wrote: »

Op check out this eview to see general performance http://www.tomshardware.co.uk/precision-t5600-xeon-e5-2687w-workstation,review-32805.html
I am sure such a big project would consider 10k for a machine small change. If they won't buy you a machine they shouldn't trust you to compile the code
If not the 4930k with quad channel memory is a real workhorse, I have one totally stable with all cores at 4.6ghz on a closed loop watercooler.

May I ask what voltage?

KyussBishop

Sunny Fluffy Stockroom wrote: »

Op check out this eview to see general performance http://www.tomshardware.co.uk/precision-t5600-xeon-e5-2687w-workstation,review-32805.html
I am sure such a big project would consider 10k for a machine small change. If they won't buy you a machine they shouldn't trust you to compile the code
If not the 4930k with quad channel memory is a real workhorse, I have one totally stable with all cores at 4.6ghz on a closed loop watercooler.

That (4930k) is what I'll be going with for the moment, yes (almost identical setup actually); that should give me at least a 3-4x (5x if lucky, after overclock) performance boost over what I have now, so hopefully will work out well.

BloodBath

Maybe up the ram to 32gb. I don't know how memory intensive it is but it sounds like it might be. Maybe try and get some idea and go for more ram if you need it. You can get a different board that supports 64gb of quad channel memory either.

Could be total overkill but you will have to find that out.

KyussBishop

BloodBath wrote: »

Maybe up the ram to 32gb. I don't know how memory intensive it is but it sounds like it might be. Maybe try and get some idea and go for more ram if you need it. You can get a different board that supports 64gb of quad channel memory either.

Could be total overkill but you will have to find that out.

It shouldn't be all that memory intensive, the bottleneck definitely does seem to be CPU here (on my current setup, I never really eat far into the 8gb of memory I have).

I'm looking at similar specs as your suggestion but different components, and have a board which should support 2 quad sets (8 slots), so will just get 16gb, and (if I might have use for ramdisks or multiple VM's at some stage later - probably not) may replace that with a full 64gb 2-quad set if the memory gets much cheaper.

smacl

BloodBath wrote: »

Maybe up the ram to 32gb. I don't know how memory intensive it is but it sounds like it might be. Maybe try and get some idea and go for more ram if you need it. You can get a different board that supports 64gb of quad channel memory either.

Could be total overkill but you will have to find that out.

Given the cost of the rig, I'd go 32gb at least, probably 64gb for a bit of future proofing. Worth remembering that multicore compilation uses a separate instance of the compiler for each process, and each of these can consume quite a bit of memory if you've got complex headers in play. To keep all cores busy all the time, VS always has more compilation processes on the go than available cores. Having precompiled headers and object files on a ram disk speeds up link times over and above what you'd get with a fast SSD. My PC uses 6.8gb compiling with 8 cores, with no RAM disk and not that much else open. While I could get away with 16gb most of the time, 32gb give a bit of breathing space, and that's for a ~€1400 build.

The difference in price between 8GB and 32GB will be a few hundred euro, so on a PC costing 3k-4k it is a total no brainer.

btkm8unsl0w5r4

Gumbi wrote: »

May I ask what voltage?

Ok I have to admit I am a lazy bastard and I allow my mobo (asus formula 2011) to decide that for me. I dont think the voltage is raised, but the memory run a little slower to stay stable. The great thing about the asus mobo is that is still allow the speedstep to work, so its not always at full tilt. Temperatures are never above 70 degrees across all cores using prime 95 for a few hours.

Oh yeah OP there is a chap called blades, or bladesof glory selling a bunch of high end memory, CPU, GPUs, mobo etc. over on adverts

Gumbi

Sunny Fluffy Stockroom wrote: »

Ok I have to admit I am a lazy bastard and I allow my mobo (asus formula 2011) to decide that for me. I dont think the voltage is raised, but the memory run a little slower to stay stable. The great thing about the asus mobo is that is still allow the speedstep to work, so its not always at full tilt. Temperatures are never above 70 degrees across all cores using prime 95 for a few hours.

Oh yeah OP there is a chap called blades, or bladesof glory selling a bunch of high end memory, CPU, GPUs, mobo etc. over on adverts

I would check if I were you man, just in case it is over applying. You could well be applying 1.4v which is totally unnecessary.