Hieb

If you can afford it there's no good reason not to get it. The i7-5820K is an absolute BEAST of a CPU. However an i5-4690K or i7-4790K would have zero problem playing every game in existence for the next few years. I mean if you're gonna be spending $300 on a motherboard anyways and getting a 4790K, why not spend the extra $50-100 for a six core?
 
However no, it's not really standard. It's on a different path than Z97.
 
Entry-Level: H61 -> H81
Business (also entry-level): B65 -> B75 -> B85
Mainstream: H67 -> H77 -> H87 -> H97
Performance: Z67 -> Z77 -> Z87 -> Z97
Enthusiast/Pro: X59 -> X79 -> X99
 
So yeah X99 is newer, but it's not an evolution of the previous mainstream/performance segment. It's on a whole other level.

WAT
 
In Canada the A10-7850K and A10-7870K are like... $170. Really not worth it considering the A8-7650K performs within a very close margin while costing only $110... plus you can basically get an 860K or G3258 paired with an R7 250X or GTX 750 for just slightly more, and it will perform a lot better due to GDDR5..
 
If I could get a 7850K plus a board for $110... ooooh man.

there are way more factors than just ipc
read about them here:
https://en.wikipedia.org/wiki/Instructions_per_cycle
in the "computer speed" section
 
the best way is to actually use the CPUs in the real world and look at benchmarks for the program/game you want to know

8350 = 4690 in editing
9590 is wasteful.

honestly, a single 970 can do 1440p fairly well. I would throw another in there if you're not happy with the performance.

Buy another 970. Just make sure your PSU can handle it. 

The R9 390 gets about ~10% better framerates but Nvidia's cards have a bit more features... Shadowplay is better than AMD's alternative and the Nvidia control panel offers a bit more... like adjusting number of pre-rendered frames for example. Up to you depending if you need the features or just want the highest fps

less efficient code = more instructions needed to get stuff down = cpu intensive o_0

For every frame your video card renders, the CPU first has to do a bunch of processing for it... the CPU processes animations, where all the objects are and then tells the GPU. The CPU also processes a lot of effects... it's involved heavily in shadows, lighting and particle effects (like explosions). So the more complex all of these things become (or the less optimized a game is), the more work the CPU has to do for each frame. And because the CPU does this for every frame, each frame adds more CPU load. So playing at 30 FPS will have roughly half the CPU requirements of playing at 60 FPS (although in some games physics calculations, AI, etc. are not tied to framerate, so then the CPU usage doesn't scale linearly)
 
So higher settings and higher framerates means more things for the CPU to process...
 
So if a certain CPU reaches its limits at, let's say, 45 FPS at high settings @ 1080P in a game with a GTX 760, then that same CPU will also reach its limits at 45 FPS at high settings @ 4K with a GTX 980 Ti.
 
This is why I don't like people saying X CPU will bottleneck Y GPU, because it depends. An i5-6600K could bottleneck a GTX 980 Ti if someone is playing triple A titles at 1080P, looking for 144Hz gameplay... playing at ~144 FPS requires a lot of CPU in big titles. Meanwhile if someone is playing at 4K, a GTX 980 Ti could probably be supported without any issues by a high end i3 (like an i3-4360) or low end i5 (like an i5-4440), because framerates at 4K are gonna be targeted around 60 FPS, which isn't too terribly demanding.
 
This is also why often times you'll see low-resolution benchmarks in CPU tests (although this is less common now since ever since Tek Syndicate's video about the FX-8350 people bitch about "real world" benchmarks, which serve a different purpose... but I digress). Because as you lower the resolution, it makes it easier for the video card to pump out more frames, and this allows you to find the point where the CPU spends all of its time processing frames (and physics, etc.) and cannot handle anymore... and at this point lowering the resolution further or adding a better graphics card won't improve the FPS because the CPU has reached its limit.
 
So long story short:
 
- CPU processes things for each frame (basically prepares it for the GPU, telling it what to draw)
- Particle effects, lighting effects, shadows (and ambient occlusion) typically have significant CPU involvement
- Then the GPU builds the picture out of pixels
 
Higher settings = more CPU involvement per frame
More frames = more frequently needs CPU to prepare frames

As always it would depend 100% on the specific game and settings you're playing at... are you looking for 1080P 144Hz gameplay? Then the i5-4590 will probably bottleneck in a lot of games, as at framerates that high there'd be tremendous CPU load. Looking at 4K 60 Hz gameplay? Then that CPU could drive that without any issues.

Well I use a Gigabyte Z97X-SLI... it does the job fine, no real complaints... only real downside to it is that it's not as wide as standard ATX boards (doesn't reach the third set of standoffs), so it flexes alot when trying to connect or disconnect the 24-pin ATX power connector.
 
Otherwise it's a great board considering it was quite cheap

Just to clear the air on a couple things:
 
1) Core is a very ambiguous term, especially in recent years since AMD's Bulldozer family of architectures... afaik there's no strict requirement or classification for a core, it's just the accepted term for the 'pipeline' in a CPU to execute instructions for a thread.
 
2) Hyper-threading effectively makes the core two cores. Obviously it doesn't double its processing power, however it allows the core to complete instructions from two threads within a single clock cycle... meaning that an i3 can in fact simultaneously process 4 things, just like an Athlon can.
 
The i3 will outperform the Athlon in almost every game out there... games aren't a perfectly parallel workload, and if they were the i3 and Athlon would perform pretty close to each other, I'd think. Games are a mixed multi-threaded, bursty workload... the load isn't consistent across all the threads, and it's not a seamless stream of tasks, it's a bunch of tasks spread out where performance is dependent on how quickly they are completed. This is why single-threaded performance continues to make such a big impact even though most games released over the past few years can use four threads or more.

Just to clear the air on a couple things:
 
1) Core is a very ambiguous term, especially in recent years since AMD's Bulldozer family of architectures... afaik there's no strict requirement or classification for a core, it's just the accepted term for the 'pipeline' in a CPU to execute instructions for a thread.
 
2) Hyper-threading effectively makes the core two cores. Obviously it doesn't double its processing power, however it allows the core to complete instructions from two threads within a single clock cycle... meaning that an i3 can in fact simultaneously process 4 things, just like an Athlon can.
 
The i3 will outperform the Athlon in almost every game out there... games aren't a perfectly parallel workload, and if they were the i3 and Athlon would perform pretty close to each other, I'd think. Games are a mixed multi-threaded, bursty workload... the load isn't consistent across all the threads, and it's not a seamless stream of tasks, it's a bunch of tasks spread out where performance is dependent on how quickly they are completed. This is why single-threaded performance continues to make such a big impact even though most games released over the past few years can use four threads or more.

I haven't personally used the G3258 but from what I've gathered reading various articles, reviews, and watching performance tests on YouTube... when the G3258 struggles, it does so because of only being able to run two threads, so it can't schedule things very efficiently. The G3258 usually doesn't just (if at all) have low average framerate, rather it suffers from stuttering, freezes and periodic big fps drops. Overclocking doesn't fix these things as far as I know.
 
I mean just to give you an idea, even when overclocked to like 4.5GHz the G3258 is usually outperformed by an i3
 
In any games that will run fine on two thread CPUs (such as World of Warcraft, older games, I suspect most mobas would do well...) the G3258 will generally be ample fast enough to drive the GTX 670, and the overclock will certainly help with that... but in more robust modern titles like GTA V or the likes, the G3258 isn't gonna provide particularly smooth gameplay regardless if it's OCed or not.

I haven't personally used the G3258 but from what I've gathered reading various articles, reviews, and watching performance tests on YouTube... when the G3258 struggles, it does so because of only being able to run two threads, so it can't schedule things very efficiently. The G3258 usually doesn't just (if at all) have low average framerate, rather it suffers from stuttering, freezes and periodic big fps drops. Overclocking doesn't fix these things as far as I know.
 
I mean just to give you an idea, even when overclocked to like 4.5GHz the G3258 is usually outperformed by an i3
 
In any games that will run fine on two thread CPUs (such as World of Warcraft, older games, I suspect most mobas would do well...) the G3258 will generally be ample fast enough to drive the GTX 670, and the overclock will certainly help with that... but in more robust modern titles like GTA V or the likes, the G3258 isn't gonna provide particularly smooth gameplay regardless if it's OCed or not.

I haven't personally used the G3258 but from what I've gathered reading various articles, reviews, and watching performance tests on YouTube... when the G3258 struggles, it does so because of only being able to run two threads, so it can't schedule things very efficiently. The G3258 usually doesn't just (if at all) have low average framerate, rather it suffers from stuttering, freezes and periodic big fps drops. Overclocking doesn't fix these things as far as I know.
 
I mean just to give you an idea, even when overclocked to like 4.5GHz the G3258 is usually outperformed by an i3
 
In any games that will run fine on two thread CPUs (such as World of Warcraft, older games, I suspect most mobas would do well...) the G3258 will generally be ample fast enough to drive the GTX 670, and the overclock will certainly help with that... but in more robust modern titles like GTA V or the likes, the G3258 isn't gonna provide particularly smooth gameplay regardless if it's OCed or not.

I haven't personally used the G3258 but from what I've gathered reading various articles, reviews, and watching performance tests on YouTube... when the G3258 struggles, it does so because of only being able to run two threads, so it can't schedule things very efficiently. The G3258 usually doesn't just (if at all) have low average framerate, rather it suffers from stuttering, freezes and periodic big fps drops. Overclocking doesn't fix these things as far as I know.
 
I mean just to give you an idea, even when overclocked to like 4.5GHz the G3258 is usually outperformed by an i3
 
In any games that will run fine on two thread CPUs (such as World of Warcraft, older games, I suspect most mobas would do well...) the G3258 will generally be ample fast enough to drive the GTX 670, and the overclock will certainly help with that... but in more robust modern titles like GTA V or the likes, the G3258 isn't gonna provide particularly smooth gameplay regardless if it's OCed or not.

Games will cache extra data in VRAM when it's available to reduce how often it needs to juggle things around between HDD, system RAM and VRAM. It doesn't always require this though. For example I can play GTA 5 on my GTX 760 2GB with High textures just fine... no massive framedrops or anything, and VRAM usage is in the neighbourhood of 1800-2000MB. But if you played on the exact same settings with the 4GB version you'd see VRAM usage well above 2GB just because it's there. Basically games will use extra VRAM if it's available just in case.
 
Also running an extra monitor just for desktop use doesn't add a significant amount of VRAM requirements, so don't worry about that.
 
GTX 980 Ti will be enough. Go nuts.

A reference card is a GPU with a PCB (motherboard of the GPU) not manipulated from the original design from Nvidia.
MSI, ASUS, Gigabyte, etc. often change the PCB a bit to fit for their coolers or whatever, but sometimes they keep it original (I think EVGA does that sometimes).
 
 
So a reference card is a card with the original non manipulated PCB and the blower style cooler and some cards have a different cooler, but still the reference PCB.
 
Either way, I believe the MSI GTX 970 is not a reference PCB and definitely not a reference cooler. The EKWB cooler probably wont fit on it (but you can contact them for some more info)

I would be fine with that but with 75 or 90Hz :3

Sees price. Exits out of,page.
They should make a 34-40inch ones that like um $1500-$2000. Still very expensive but much more affordable.

No way in hell am I getting a GSync monitor lol... cheapest one is $650 for a 1080P G-Sync monitor... no thanks. I can get a 1440P 144Hz freesync monitor for less than that (and I plan to even though I wont be able to use FreeSync).

the G3258 is shitty due to it stuttering in modern games....
 
I have posted shitloads of counter-evidence when it comes to "how awesome the G3258" is.... it is NOT A DAILY DRIVER CPU... ITS A OVERCLOCKING TOY...
 
Do not ever, ever, ever try to tell anyone anything else.
 
Athlon 860k / Intel Core i3 is the "lowest" you should go.
 
benchmarks doesnt show you stutters. They show you average FPS numbers. They do not show you frame pacing. Only FCAT does that, and there is PLENTY of test proving how the G3258 isnt nearly that good when paired with a high end GPU. Austin arent intentionally misinforming, he just isnt showing the whole picture, and i do not blame him for that....

I KNOW' again here is my testing of the G3258:
Stuttery mess...but the 860K isnt really better.
i think you seriously have NO IDEA who you're talking too...10 000+ posts made in the CPU section of this forum son.
i've owned AMD FX...tested with 2 modules (FX-4300) 3 modules (FX-6300) and 4 modules (FX-8350) both at stock and overclocked...tested haswell with 2 cores, 2 cores with HT, 4 cores, 4 cores with HT, overclocked and stock clocks...tested everything so i DO have a VERY good and accurate idea of how these chips perform in modern games.
Here is for example the testing i did with watchdogs when it came out (only intel CPU's though) i know i have no life..15 000 views on this video though!

The Athlon 860K perform about the same as the FX-4300...slightly better throughput but nothing you would notice and the L3 cache on the FX-4300 makes it a slightly better performer overall...this site is VERY good it's user submitted results and this has also been proven to be accurate:
http://cpu.userbenchmark.com/Compare/AMD-Athlon-II-X4-860K-vs-AMD-FX-4300/3265vs2879

Fair enough, the card is overpriced... AMD with this generation of cards is really hoping to pull in some extra dollars for 'premium' products with the Fury lineup. It should cost the same as the Fury non-X imo.
 
 
 
Agree completely... been saying this since as soon as the Fury lineup was announced, I think the lineup should have looked like this:
 
R9 390X - Fiji XT (Fury X)
R9 390 - Fiji (Fury)
R9 380X - Hawaii XT (R9 290X)
R9 380 - Hawaii (R9 290)
R9 370X - Tonga XT (R9 M295X)
R9 370 - Tonga (R9 285)
R7 360X - Pitcairn XT (R9 270/HD 7870)
R7 360 - Pitcairn (R7 265/HD 7850)
R7 350X - Bonaire XT (R7 260X)
R7 350 - Bonaire (R7 260)