Recommended Posts

I came to you guys a few months ago in asking for advice one a computer I was going to build. I took your advice and put it on the back burner to see what will all be coming out this year. Haswell seems impressive but I was wondering by all of the documentation released on it by now is it worth it for me? I want to get the most power out of my computer. Will haswell be better then the previous high end ivy bridge processors? I don't care about power consumption as much as power since it's my desktop. If it were my laptop it would be the opposite. I do high end gaming and I develop apps/small games.

TL;DR: Basically I want to know if you would recommend haswell for a power user who programs apps, runs multiple virtual machines / does high end gaming (obviously would get a high end video card so the gaming doesn't apply as much) and wants the most power possible for a desktop user.

Link to comment
https://www.neowin.net/forum/topic/1130518-will-haswell-be-worth-it/
Share on other sites

Someone asked an Intel engineer this:

"As a person who just bought an Ivy-bridged based system, is there anything you can tell me to convince myself to save up for a haswell or broadwell system?"

The answer:

What do you usually do with your system? If you like to overclock, Haswell is worth it (can't tell you why but read the Haswell Anandtech preview very carefully for buried treasure). On-die graphics is improving quite a bit as well. If you're into energy efficiency or even more graphics, Broadwell. I think the tech community will be very pleasantly surprised with Broadwell. But I'm biased, so we'll just going to have to prove it the hard way.

Source:

http://www.reddit.com/r/IAmA/comments/15iaet/iama_cpu_architect_and_designer_at_intel_ama/

What are the specs of the computer you're running now?

I don't see it being that much better. Depends on the price mostly. It will most likely be faster, but will it be worth the premium? Probably not. I don't really see why anyone need more power than what's available now unless you NEED or for specific applications. I run 2-3 VMs regularly on an i5-2500 with 8GB of RAM and they run just fine. You'll get the best performance from getting an SSD and 8-16GB or RAM.

I'm sceptical. The biggest disappointment is that they'll only go up to quad-core, which feels like a real missed opportunity. Ivybridge was a huge let-down and games performance seems to have taken a back seat for Intel, which is what I'm most interested in.

As always, we'll know more when the first proper reviews come out. If only AMD was more competitive.

Well I have a first gen i7 right now with 12 GB of DDR3. I'm going to upgrade eventually b/c my fiance has an ancient computer (running xp that came on it 6-7 years ago). So it's not so much that I NEED the upgrade, it's that when I do upgrade I'm wondering if it's smart to invest in a haswell processor or to just get an ivy-bridge. Will the haswell processors be faster even though intel is focusing solely on the power reduction areas of the cpu?

EDIT: That's one of my disappointments as well theyarecomingforyou. I was hoping for at least a 6-core hyper threaded version to come out for the enthusiast brand (which I hear actually has less features planned then the non-enthusiast version).

1st gen i7 is still very fast, why do you need to upgrade to ivy bridge or haswell, surely that is fast enough? My 1st gen core i5 laptop is fast enough at the moment on win 7. Broadwell should bring large performance increases and it is a SoC so motherboards will be smaller, lower heat levels due to less chips.

I can't see intel releasing any chips with tons of cores as those users may not upgrade for years after buying one of those chips. If AMD was more competitive then intel would be forced to.

Well I have a first gen i7 right now with 12 GB of DDR3. I'm going to upgrade eventually b/c my fiance has an ancient computer (running xp that came on it 6-7 years ago). So it's not so much that I NEED the upgrade, it's that when I do upgrade I'm wondering if it's smart to invest in a haswell processor or to just get an ivy-bridge. Will the haswell processors be faster even though intel is focusing solely on the power reduction areas of the cpu?

EDIT: That's one of my disappointments as well theyarecomingforyou. I was hoping for at least a 6-core hyper threaded version to come out for the enthusiast brand (which I hear actually has less features planned then the non-enthusiast version).

Like I said in the other post, I'm giving my fiance my computer now and getting myself an upgrade. She just does web surfing and casual gaming so it will be perfect for her :)

Exactly the above. Because AMD can't offer anything of worth at the moment, Intel will keep their aces (if they have them) up their sleeve for some time. Average Joe still doesn't know how to fully load i3-3220 most of the time. But if one does, there's all the rest available.

As for the enthusiast part, enthusiasts cried out loudly for Ivy Bridge-EP/EX, which at one point was said to not exist. That's what enthusiasts will get, in due time and that's where 10 hyperthreaded cores are going to be.

I'm not on the highest end of systems by a good long shot, but I also find nothing I cannot run. When I did have issues with my old Core 2 Duo, I upgraded my system. I've been running probably the most hated of any processor I've ever known, an FX-6100. It's does great, so... is anything more than even a core i5 worth it? No, and not just according to me, many sites say even buying a Core i7, unless for some highly specialized processor intensive task, is not worth it.

I don't see it being that much better. Depends on the price mostly. It will most likely be faster, but will it be worth the premium? Probably not. I don't really see why anyone need more power than what's available now unless you NEED or for specific applications. I run 2-3 VMs regularly on an i5-2500 with 8GB of RAM and they run just fine. You'll get the best performance from getting an SSD and 8-16GB or RAM.

It's what I've been saying since (surprisingly) Sandy Bridge - Haswell is for outliers (and far outliers, at that); that is, in fact, why it will appear outside the general core LGA1155 socket first. (Either LGA2011 or its replacement - basically the workstation/server market - currently occupied by Sandy Bridge-E.)

To be honest, there's a lot of untapped potential merely in Ivy Bridge - and I'm talking i3 through i5-K - teamed with a Z77 motherboard, 16GB of RAM and a decent midrange GPU (AMD's HD67xx/77xx and above or NVidia GTX550Ti (Fermi) or GTX650Ti (Kepler) - in not one case should the GPU set you back $200, even retail).

An SSD need only be for the OS and the most-used directory - for gamers, typically it's the folder on which Steam is installed; with the addition of targeting for installs (either for Steam itself or games purchased/installed via Steam), smaller SSDs are more practical than ever, especially considering recent outbreaks of pricing sanity for the 128GB SSD range (now generally in the $0.75/GB range, even for known-quality SSDs from Samsung [830 and 840], Crucial [m4], and Intel[33x/52x]) - use your existing mechanical drive as data-only.

I'm sceptical. The biggest disappointment is that they'll only go up to quad-core, which feels like a real missed opportunity. Ivybridge was a huge let-down and games performance seems to have taken a back seat for Intel, which is what I'm most interested in.

As always, we'll know more when the first proper reviews come out. If only AMD was more competitive.

Please.

It's not that performance has taken a back seat at Intel - it's that performance has been hobbled by multi-platform development.

When you develop a game - ANY game - for multiple platforms, the defining platform will ALWAYS be the LEAST powerful - not the most powerful.

Look at Supreme Commander (and the add-ons/reworks thereof) vs. Supreme Commander 2 - the first was not merely PC-exclusive, but multi-core optimized; the latter was, however, neither.

Supreme Commander takes advantage of multiple cores in a way that no game preceding it - and few games since - can match. (How is Intel to blame for that?)

The issue with general-purpose applications is, if anything, worse - FAR worse. (For example, how many x64 general-purpose [productivity] applications exist, despite no x32 general-purpose CPU - outside of Atom - even being manufactured by either Intel or AMD?)

What was all the criiticism of Crysis 2 about? That it was hobbled by being multi-platform; basically, the least-powerful platform defined the entire game. (Crytek DID overcome quite a bit with the 1.9, HDR, and DX11 Ultra upgrades - however, that least-powerful-platform hobbling was a very high hurdle to clear - I'm quite honestly surprised they managed to clear ANY of it.)

Still, gaming remains like workstation application usage - one for outliers pretty much; not the general PC masses.

Haswell will definitely bring some improvements. Among others GPU and AVX2 come to mind.

Will Haswell be extremely revolutionary? No! (at least not for the consumer...)

Will it make sense to buy Ivy Bridge after Haswell has been released? No! (at least in the general case you shouldn't buy into an outdated platform)

Is it necessary to wait for Haswell? No! (if you need the power/a new system now, just buy it!)

Haswell will definitely bring some improvements. Among others GPU and AVX2 come to mind.

Will Haswell be extremely revolutionary? No! (at least not for the consumer...)

Will it make sense to buy Ivy Bridge after Haswell has been released? No! (at least in the general case you shouldn't buy into an outdated platform)

Is it necessary to wait for Haswell? No! (if you need the power/a new system now, just buy it!)

AVX2 is not new - it came with Sandy Bridge, and has been usable ever since. (AVX2 builds on AVX, which was standard fare starting with Nehalem.)

GPU improvements will be primarily at play in the portable space - which is not the desktop space. (Discrete still rules on desktops.)

Running a pre-Sandy desktop? If looking to upgrade, go Ivy Bridge - now.

Running a pre-Sandy portable? THEN you may want to wait for one running Haswell (GPU improvements) - otherwise, Ivy Bridge makes a lot of sense here also.

AVX2 is not new - it came with Sandy Bridge, and has been usable ever since. (AVX2 builds on AVX, which was standard fare starting with Nehalem.)

It's feature set definitely is as AVX does for example NOT support integer operations. Obviously it extends AVX hence AVX2...

AVX2 is also called "Haswell New Instructions" => it's from Haswell...

This topic is now closed to further replies.
  • Posts

    • This got me thinking, would you rather a self driving car prioritise protecting its passengers or everyone else? I'd choose the one that keeps me and my kids safest. At some point, these cars have to make those choices already, don't they? Wonder if we have a way to find out what way they lean.
    • The proportion (or number of iterations) has nothing to with this aspect of Copyright I am describing. In short, it doesn't matter how many times the manager tells you to change something or how. Your work product is always YOURS until and unless you then assign that to the person representing the client/company, usually for financial compensation -- either in salary or as a subcontract work for hire payment. if iterations determined copyright, then businesses would have learned to just keep making changes until they could claim they owned the copyright, without having to compensate the artist for their work. And that would be BAD. The only place where the amount of changes does have a role is in how much does a human modify a previous public domain work (from any source) before it is considered fair use or their own work, etc. For example, if a human makes substantial changes to a public domain (re: AI, by definition) work, then they can then claim that derivative work as their own...but NEVER the original version, of course. That's why anyone can make a movie about Dracula, for example, as long as it is based on the public domain novel, but not if they take new ideas from copyrighted movies made afterwards. As one of the people who personally advised the US Copyright Office on their recent ruling on these very issues, be assured that I specifically used the terminology precisely -- though I made it simple enough for laymen to understand it. If I made this confusing by doing so, I apologize. But, to be clear regarding your assumption that I would agree to your second statement that I quoted above -- the answer is NO. If AI does the work, no matter how much "direction" you give it, it cannot be copyrighted. All AI generated content is in the Public Domain and therefore the copyright cannot be assigned to ANYONE, even you -- until and unless substantial modifications are made to it BY A HUMAN BEING (yourself or a contracted artist/writer/etc.) and then that copyright on the derivative work is legally (in writing) transferred to you. This is a critical distinction. And it is important that people, especially AI sloppers, understand this. For example, YouTube is not paying AI slop generators for the copyright, etc. of their AI slop. What YouTube is doing is sharing AD REVENUE for permission to publish your AI slop. Copyright/ownership/rights never come into it. Importantly, that means that anyone can copy any AI slopware on YouTube, etc. and rehost it anywhere they want, even back on YouTube, and there is nothing legal that YouTube can do about it with regards to copyright protections, ownership, DMCA, etc. Anyone is legally free to use any AI slopware in any way they want. When this ruling was pending, I warned Disney legal of all of this before they did their OpenAI deal -- that it would literally dilute their entire IP portfolio forever. They ignored that warning for the PR and stock bump. But that is why, when the ruling came down last year, Disney quickly extricated themselves from that OpenAI deal, even eating the initial upfront fees -- followed closely by OpenAI ending their entire AI video generating business model. They adjusted their PR release dates to make this less obvious to shareholders, of course. Phew. I hope that this clears up the key distinctions for you and anyone reading. If you have any additional questions or even hypotheticals about AI and Copyright, please feel free to ask.
    • Each of the devices displayed on this page now has a little volume meter next to it to show if there is audio actively playing. About time.
    • Owing to the nature of Windows feature enablement updates, it was distributed over Windows Update services as a complete system upgrade rather than as an ordinary cumulative update
  • Recent Achievements

    • Collaborator
      ryansurfer98 went up a rank
      Collaborator
    • Week One Done
      Eurosoft10 earned a badge
      Week One Done
    • One Month Later
      Eurosoft10 earned a badge
      One Month Later
    • One Year In
      Skeet Campbell earned a badge
      One Year In
    • One Month Later
      Sharbel earned a badge
      One Month Later
  • Popular Contributors

    1. 1
      +primortal
      565
    2. 2
      +Edouard
      188
    3. 3
      Michael Scrip
      79
    4. 4
      PsYcHoKiLLa
      74
    5. 5
      neufuse
      71
  • Tell a friend

    Love Neowin? Tell a friend!