Welcome to the dark ages

[Gigantopithecus]

Software needs to seriously catch up to hardware.

This! But outside of HPC, what's to improve upon?

 

[Pijoto]

PC hardware lasts longer than ever (esp. on the CPU side). While some may find that "dark ages", it also saves you $ from unnecessary upgrades. Software needs to seriously catch up to hardware.

Most games still use only dual cores, once software catches up to using quad, everyone who bought quad-cores will magically find their CPUs to be twice as good...no point in upgrading for another 3-4 years if you have a decent quad, huh? Personally, I'm thinking of upgrading to Phenom II X6, and just disabling 2 or 3 cores till games start using the extra cores, and overclock the rest.

 

[nyker96]

I think the BD launch feel to me like a unfinished product. I hope w/ some engineering work on the core, it can better itself fast in the next few years into something competitive. I think I was a bit overly critical of it since the expection was high for me. But it feel like something that hasn't been finished yet. The module idea is good, if combined w/ a gpu side the floating point calculation can be better. plus if they can optimize the each core by 10-15% each iteration, well you got 8 of these which means 8ximprovements. With that extra 5-10% from win8 scheduler, I'm hoping the next few iteration will make us forget this failed launch like PhenomI when PhII came out and fixed things. But all in all, this chip needs to go back to drawing board.

 

[Gikaseixas]

Intel will keep pushing out improvements for it’s already amazing line up, they're a forward looking company and they know that being complacent will put them in the same position they found themselves a few years ago (Athlon glory days). Also, AMD is used to work with much slimmer margins than Intel so don't be so quick to consider them out of the race. They can improve power consumption, tweak single thread performance and voila, have a chip that can give Intel a run for its money. Competition fuels improvements ,make no mistake…

 

[3DVagabond]

Intel will keep pushing out improvements for it’s already amazing line up, they're a forward looking company and they know that being complacent will put them in the same position they found themselves a few years ago (Athlon glory days). Also, AMD is used to work with much slimmer margins than Intel so don't be so quick to consider them out of the race. They can improve power consumption, tweak single thread performance and voila, have a chip that can give Intel a run for its money. Competition fuels improvements ,make no mistake…

I'm not so sure that Intel got complacent last time as much as they just made a wrong design decision. Luckily their mobile division was onto something good and it translated well to the desktop. Since then they've been able to do no wrong building on that.

I mentioned before, but the person I said it to just dismissed it (not you), AMD has other products to sell. The $200-$300 CPU segment is not going to make or break them. Sure, they would really like to sell a lot of processors in that category, but not doing so isn't going to bankrupt them. They probably can't afford to lay any more eggs though right now. Trinity has to be good. Southern Islands has to be good. They took a big swing and a miss on BD. We'll see if this "magic re-spin" we've heard about brings any appreciable improvement. It's not going to get them into the same league as Intel, but maybe they can at least beat their last arch.

I'm in agreement with your overall assessment. Just in case it didn't read that way.

 

[Riek]

How is AMD not competitive with Intel?
If intel increased the price of their lineup with 50$ they would become extremely unattractive compared to the other choices. so no Intel is not without competition... They do have a performance advantage, but they are not without competition given they can't ask what, where and how they want it. The real question will be how will they be positionned over a year... will the picture change alot? If by then we still pay the same for 2600K performance. (e.g. intel makes its newer and faster processor more expensive instead of the same price regions) then we can speak about lack of competition. This is not yet the case for today.

 

[amenx]

Intel still has to compete with themselves. Their products can't get too expensive because people simply won't buy them. There's no reason to. If Intel wants to continue to move products, they have to obsolete their old products and convince people it's worth the money.

I agree. People will just stick to what they have - no point in upgrading if it doesnt bring a worthwhile improvement.

 

[Masahiro]

While I agree that BD's launch was very poor and as a result Intel can breathe a sigh of relieve, I agree moreso with people saying they still have to compete with themselves. They've already set in place a corporate mantra with the tick tock release cycle so I'm fairly certain that regardless they're going to follow it.

If they plan to stay ahead they need to keep their research and development division up and running with as much funding as possible to keep the new ideas coming. If not they will rest on their laurels and that's when AMD would be able to strike back hard.

As for no point in upgrading if it doesn't bring a worthwhile improvement, performance is but one facet of customer expectations today. The idea of less power hungry processors intrigues many people (myself included).

 

[keyser fluffy]

I agree there is or is going to be a 'lull', I think there are a few problems:

• clock speeds aren't increasing any more
• the industries answer to this was multiple cores
• diminishing returns of multiple cores, because concurrency is hard, I like the analogy of pistons in a car engine, because 4 is enough
• we've hit the 'heat cliff' (Jim Gray quote) of silicon - not many people want a CPU to use 130+watts etc

in terms of gaming:

• diminishing returns of multiple cores, because concurrency is hard
• the 8th gen consoles are delayed because.. inset guess here: we already have enough processing power for 3D games, realism != fun, concurrency is hard...

It's not much to do with Bulldozer.

 

[Lonyo]

If anything, we're about to enter a new age of competition where Intel needs to be even more competitive.

Where Intel is a DOMINANT leader is in the high end CPU market.

Where we seem to be heading is having more devices, each less powerful, and that's the lower end. Many mobile devices, where AMD has product offerings (of a sort) and where ARM is/is heading (scaling up to be more powerful).

Any shift "away from" single high power machines (in the consumer market) means that Intel's high end CPU dominance doesn't count for much.
Then you have the server/HPC market, where people are looking at GPUs, and they are starting to be used more and more. You get super high end HPC mixed servers with many CPUs + many GPUs, and if people exploit GPUs further, that again moves people away from (Intel) CPUs.

Just because AMD isn't competing in the high end of the consumer CPU market doesn't mean Intel has a hope in hell of sitting back and making the most of their dominant position in that one market. The future is coming, and it won't be televised, it will be streamed over the internet onto a multitude of devices which don't run a high end CPU.

 

[StrangerGuy]

Intel also needs to compete with its old backlog of CPUs...The average users tend not to buy new PCs when even their 5 year old C2Ds are more than good enough for office work and the like.

 

[Gikaseixas]

I'm not so sure that Intel got complacent last time as much as they just made a wrong design decision. Luckily their mobile division was onto something good and it translated well to the desktop. Since then they've been able to do no wrong building on that.

I mentioned before, but the person I said it to just dismissed it (not you), AMD has other products to sell. The $200-$300 CPU segment is not going to make or break them. Sure, they would really like to sell a lot of processors in that category, but not doing so isn't going to bankrupt them. They probably can't afford to lay any more eggs though right now. Trinity has to be good. Southern Islands has to be good. They took a big swing and a miss on BD. We'll see if this "magic re-spin" we've heard about brings any appreciable improvement. It's not going to get them into the same league as Intel, but maybe they can at least beat their last arch.

I'm in agreement with your overall assessment. Just in case it didn't read that way.

I agree with your assessment too , but I still believe they got caught off guard back then.

Intel had parallel teams working as everyone knows and that paid off for them. Again, they chose poorly (bad decision) but maybe they truly believed in long pipeline design of P4 and decided to pursue that road with dual core cpus and things got out of hand because P4 was never engineered to be a multi core processor. Added cores = added heat

 

[Masahiro]

Where we seem to be heading is having more devices, each less powerful, and that's the lower end. Many mobile devices, where AMD has product offerings (of a sort) and where ARM is/is heading (scaling up to be more powerful).

Any shift "away from" single high power machines (in the consumer market) means that Intel's high end CPU dominance doesn't count for much.
Then you have the server/HPC market, where people are looking at GPUs, and they are starting to be used more and more. You get super high end HPC mixed servers with many CPUs + many GPUs, and if people exploit GPUs further, that again moves people away from (Intel) CPUs.

That's very true. When you look at the bigger picture, high end CPUs are taking backseat to lower power CPUs used to power more mobile devices. As tablets and smartphones continue to take the consumer market by storm, CPU manufacturers are goign to continue scrambling to make new low power CPUs to power these devices.

 

[Locut0s]

The real victim in all this is going to be our wallets. As others have said I don't see Intel lying down and stopping innovation. They know better than that. BUT we very well could see a return to the old days of higher end CPUs raping our wallets. With AMD basically conceding the high end part of the market to Intel for the foreseeable future there really is little reason for Intel NOT to start to pump up prices on the higher end and rake in the profits. AND they may well need those profits to help them compete in the low end medium performance / low power market now dominated by ARM chips. With ARM chips now getting to be almost as powerful as low end x86 parts AMDs future does not look bright. Sad to say it but I could see them going out of business in a few years if they don't regroup fast and put out an innovative plan going forward. They have relegated themselves to the low end of the x86 market, a market that even Intel can play quite well in and one that is being nibbled away at from the bottom by ARM chips. I hope I'm wrong in this. :'(

 

[Locut0s]

I'm not so sure that Intel got complacent last time as much as they just made a wrong design decision. Luckily their mobile division was onto something good and it translated well to the desktop. Since then they've been able to do no wrong building on that.

I mentioned before, but the person I said it to just dismissed it (not you), AMD has other products to sell. The $200-$300 CPU segment is not going to make or break them. Sure, they would really like to sell a lot of processors in that category, but not doing so isn't going to bankrupt them. They probably can't afford to lay any more eggs though right now. Trinity has to be good. Southern Islands has to be good. They took a big swing and a miss on BD. We'll see if this "magic re-spin" we've heard about brings any appreciable improvement. It's not going to get them into the same league as Intel, but maybe they can at least beat their last arch.

I'm in agreement with your overall assessment. Just in case it didn't read that way.

Thing is their OWN road maps they have published only show a max 15% improvement in these parts! That's not going to help them much when they are behind a good 30-50%+ in many many areas. And they aren't dealing with a stationary target either, see Haswell.

 

[Locut0s]

That's very true. When you look at the bigger picture, high end CPUs are taking backseat to lower power CPUs used to power more mobile devices. As tablets and smartphones continue to take the consumer market by storm, CPU manufacturers are goign to continue scrambling to make new low power CPUs to power these devices.

Sadly this isn't actually good news for AMD either though. AMD has been sorely behind Intel in "performance / power", and let's not even look at ARM chips, ever since they release Core and their hyper focus on power consumption. And perf/power is KING in the mobile sector.

 

[exar333]

The only thing Intel is really being stingy about is clockspeed. Thankfully this is an 'enthusiast' forum and we can OC our CPUs and get the top speed that the platform allows for.

The fact that you can get a 2500k, Z-MB, and 8GB RAM all for ~$350.00 is AMAZING. CPUs are relatively cheap, motherboards are mature and cheap, and RAM is DIRT cheap. This is an amazing time to build a new system and have it perform 80-85% of the top-of-the-line 2600k/i7 990 in most tasks. What's to complain about?

I remember having to fork-out a couple hundred just a few years ago on a decent quantitiy of RAM when DDR3 came out. Components are cheap, and the sky is the limit for most OCs on good chips, Intel or AMD.

Saying this is the 'dark ages' is really naive, to be honest.

 

[piesquared]

Interesting that AMD's developer forum activity has increased substantially since the world debut of integrated GPU and CPU with their Ontario fusion part. Seems some have forgotten the prevalant conversation about a 1-2 years ago; GPGPU and OpenCL (and now Microsoft's C++ AMP). There is a big pool of untapped performance there that developers are currently targeting, and really only available to the mainstream on AMD hardware with Fusion.

 

[WhoBeDaPlaya]

I remember having to fork-out a couple hundred just a few years ago on a decent quantitiy of RAM when DDR3 came out.

Actually, those of us who built 1st gen i7 systems early were lucky. DDR3 prices really skyrocketed after mid-2009.

 

[sm625]

Interesting that AMD's developer forum activity has increased substantially since the world debut of integrated GPU and CPU with their Ontario fusion part. Seems some have forgotten the prevalant conversation about a 1-2 years ago; GPGPU and OpenCL (and now Microsoft's C++ AMP). There is a big pool of untapped performance there that developers are currently targeting, and really only available to the mainstream on AMD hardware with Fusion.

Yeah but the problem is how do you take advantage of a gpu in your software? Lets say I have an array of 16 bit unsigned integer variables. And lets say I want to multiply each one by 2. Sounds like a great job for a big wide SIMD in a gpu! But how can it possibly be faster to send that entire array out of my L1 or L2 cache all the way to the L2 cache onboard the gpu to be processed, and then returned back to the cpu? How can that be any faster than just running an SSE inside the core? Obviously the only way it can be faster is if it is a pretty big array. Like say an array of 65536 16 bit values. But even then I wonder how much of a gain would I really see? I have no way to quantify it, but my gut tells me it wouldnt buy me much performance. AMD may have a gpu on the same silicon. But until they place their radeon cores right inside the cpu fpu, it doesnt mean much.

What I want to see is a 4 core bulldozer module with a shared GCN fpu consisting of at least four 16-wide vector SIMDs. Yes, an entire GCN compute unit inside each bulldozer module. Then we'll have something.

 

[Puppies04]

I really dont get the 2700k at all, ok it is a cherry picked 2600k so you might get 100mhz more on an overclock for a given voltage but I can't see many OCers buying them because most people who are going to upgrade to SB already have done.

Now if intel had bundled it with a corsair h80 type water cooler and set it at 4.4ghz there might be a few more people who for whatever reason don't/won't OC their CPU themselves who would have bought it.

Just my 2 cents

 

[justinm]

I really wish AMD could release a product that makes you say "wow" instead of damn. It would help competition nicely...

 

[Rifter]

I really dont get the 2700k at all, ok it is a cherry picked 2600k so you might get 100mhz more on an overclock for a given voltage but I can't see many OCers buying them because most people who are going to upgrade to SB already have done.

Now if intel had bundled it with a corsair h80 type water cooler and set it at 4.4ghz there might be a few more people who for whatever reason don't/won't OC their CPU themselves who would have bought it.

Just my 2 cents

Its there to beat BD. Its AMD's fault the 2700k is not much of an improvement, if intel needed to up clocks 500Mhz to beat BD they would have. But since AMD screwed up BD and intel only needed 100Mhz to beat it thats what we ended up with. Obviously its not ment as an upgrade for current SB users.

 

[3DVagabond]

Thing is their OWN road maps they have published only show a max 15% improvement in these parts! That's not going to help them much when they are behind a good 30-50%+ in many many areas. And they aren't dealing with a stationary target either, see Haswell.

I believe that those roadmaps were done with the belief that BD was going to be faster than it is. Maybe there's nothing they'll be able to do about it? Maybe they can get some of the performance though that they were originally shooting for and then have further 10%-15$ improvements on top of that. Bulldozer, as it is, just isn't at all competitive. If they can't get more substantial improvements (across the board, power, clocks/IPC) then they need to just move on from it. A six core Thuban at the same clocks beats an 8 core Bulldozer in both single and multi-threaded work loads. That's really Bulldozers failing. I never expected it to be faster per core/per clock than Intel's current offerings. I did expect it to beat PhII though.

 

[Locut0s]

Its there to beat BD. Its AMD's fault the 2700k is not much of an improvement, if intel needed to up clocks 500Mhz to beat BD they would have. But since AMD screwed up BD and intel only needed 100Mhz to beat it thats what we ended up with. Obviously its not ment as an upgrade for current SB users.

This. But most processor manufacturers also routinely update their top of the line parts once during the product's life cycle. See the release of the 990x to replace the the 880x and everything moving down. But in this case I think the 2700 is still costs more than the 2600 so it's what you said.