Originally Posted by RaV666
so you would rather have amd lose money on a certain market just to please 100 or so person .... and lose how many millions that could have been invested elsewhere that it would make sense ....
so you would rather have amd lose money on a certain market just to please 100 or so person .... and lose how many millions that could have been invested elsewhere that it would make sense ....
WILL CUDDLE FOR FOOD
The 56XX series chipset can already support 1600MHz memory. A BIOS update will allow for support of that speed of memory.
Oh yeah of course. But the issue is that the motherboards which support it wouldn't be able to set the correct clocks even if it supports DDR3 1600mhz unless the SPD has it or if it's adjustable.
And correct memory SPD settings? Those motherboards with that chipset don't allow 'overclocking' or manually changing the settings do they?
Imagine you are the head of the desktop division and the board comes to you and says "here's $5M, go bring us another product for the roadmap."
Do you put that into something that brings in $500K-$1M in revenue or do you put it into something that brings in $10-20M in revenue.
With a full portifolio and all of your R&D dollars accounted for, adding something like this means you have to either A.) hire more people, spend more capital or B.) cancel an existing project.
So, generally speaking, when you ask to add anything to the product roadmap, you are essentially having to make a tradeoff. Nobody is going to ask for more more money in order to lose money nor is anyone going to ask to cancel a profitable program to replace it with an unprofitable program. It just doesn't happen.
Well, how many people/organizations would invest in such endeavour highly rely on the competetivness with intels 2011.
In short, if BD sucks compared to sandybridge.Yea, theres no point.
If it is tho, client base isnt just your 100 peeps at XS.
Plus youre getting brand recognition, which AMD surely could use alot.
Theres a lot of simple folk that buy intel just on the premise that they have "the fastest thing".People arent logical beings, most of them arent at least.And its true for standard joe market and for IT guys too, really little of them are hw oriented.
And i cant really see 5M here, its all ready.Its all there, its not like getting whole new platform, just unlocking the potential in the existing one.
Theyre releasing same chips on same chipsets , with very similar mainboards.
Its just a modification.
Its a discussion which we already had here before with JF, and it ended in impass, but than it was about dual socket/die enthusiast board using phenom X6`s.Phenom x6 isnt that good tho.
But if BD is AMD`s comeback, it changes picture alot.
I believe FX72-74 could lose some money, it was a missfire considering core2quads performance and price.
But i firmly believe AMD made money on FX51-57.And they more importantly made big strides in reputation.
So in the end, viaiability of such product depends on performance figures/scaling/power.Thats at least my point of view.
Theres no way Intel lost money on 1366 and 6 core westmeres.
While there are a reasonable number of people that have dished out for the expensive processors, lets be honest: the people willing to spend $1000 on a processor or $600+ on a motherboard are in the minority. For AMD to develop a platform and get one or more motherboard companies to agree to manufacture them in order to have a few hundred orders doesn't make much sense, nor is it going to magically going to get them mindshare. Intel's mindshare came along with the Q6600 and the Core i7 920, not with the i7 975. It's the combination of performance and affordability. Yes, there are plenty of perceptions, benchmarks, TV/Internet ads and everything else involved, but it's those types of products that get the mindshare. Once people can tangibly see and benefit from the products, they tell others, etc. It's the same reason why there is a reasonably large contingent that have the Phenom II x6 processors, good performance and affordability. Reputation is more than just having the best processor at every level or price-point. Is an SR-2-like system from AMD a cool idea? Sure! would I buy one? Probably not. Like most consumers, I'm a bit stingy in my spending (at least these days) :p. The SR-2 was something Intel/EVGA did as more of a splurge with the success of their product. It wasn't their first thought. I don't think it ought to be AMD's either.
Dual CPU workstation/enthusiast board would be awesome.But even single socket but higher performance platform is not going to happen.
http://store.apple.com/us/browse/hom...co=MTg5MTY5NDQ
You really believe they sell 100 of these ?
Every other graphics designer/musician/photographer seems to own one of them.
History shows that when AMD has the product, it goes for the high end, so im hoping theyre cooking something for 2012 with BD2.In the meantime maybe some smart fella will write software to overclock theyre server products ;-P
These systems are little different than using a server board and xeon processors. If anything, the motherboard has fewer features. This isn't a "special" system, and is one that you can build with comparable features using a SuperMicro board and opterons for AMD.
Theyre using not only 2P with xeons, but people mostly buy the cheaper nehalem/westmere 1366 based, and soon they will move to s2011 single socket systems.Thats the part where AMD wont have anything.
Well, unless 4 module 2 channel BD is going to be uber awesome and beat that.
There are a lot of people with core i7 920s, who will not move to another 8 threaded cpu. All those sales will go to 6 and 8 core sandybridges on 1356, and amd doesn't even have an answer to those midrange sb chips, let alone s2011. If amd really doesn't want to compete even in the mid-high end, well thats too bad.
Core i7 920 3849B028 4.2ghz cooled by ek hf | 6gb stt ddr3 2100 | MSI HD6950 cf cooled by ek fc | Evga x58 e760 Classified | 120gb G.Skill Phoenix Pro | Modded Rocketfish case + 1200w toughpower | mcp 655 pump + mcr 320 + black ice pro II
It can't be good when the marketing man is playing down the importance of high performance.
id bring in a 10M$ product to the market obviously ... :P not a 500k product ... and my post was to agree with you john
with all of your post you seem to think amd is doomed if it doesnt do as you wish ....
but the reality is amd isnt doing bad these days ... so lets give them credit for that .. and let them do what they know is working for them OK?
WILL CUDDLE FOR FOOD
Where does JF downplay the importance of high performance? He is the server marketing guy btw, not desktop client one.
The whole discussion about a 2P mobo for enthusiasts should be held with desktop marketing guys, not JF, in my opinion. If Bulldozer can beat the Intel offer on either absolute performance, or performance/watt (will be very hard against SB), AMD and/or one of their partners (*cough* Saphirre *cough*) should seriously think about a 2P overclockable mobo and market it as a little personal supercomputer to help save the world from diseases (ie. F@H) or other scientific projects (BOINC etc). Add to that GPGPU computing and you're all set. The x86 market will soon be attacked from below by ARM, so you have to focus on x86's strenghts - performance! Backwards compatibility is pretty meh in the desktop client area... The revenue from such mobo might indeed not cover all costs, but since the marketing costs should easily be split among all line of products, brand recognition should rise and revenue for other product lines, ie. mainstream, would rise too.
Wait, what was this topic about?
I think we should first wait and see how Zambezi performs.With 8 cores that should clock fairly well,2P on desktop is losing its appeal(as opposed to days when Quad Core launched on desktop).
If Zambezi offers good performance ,relative to previous X6 and intel's Westmere, AND if it OCs like we all wish for,I doubt that there would be that many users who would shell out a grand on board and additional grand on a pair of CPUs...All that just to see them do BOINC.You will be able to do distributed computing with C32 board and pair of new Opterons,for roughly the same cost.
Personally, I really don't see the point of pushing overclockable 2P systems at this point. It'll be a vanity project.
This seems like a non issue since a version of what was suggested already exists, this time not some 4x4 'I run hotter and suck at performance' quadfire.
Asus KCMA-D8
http://www.asus.com/product.aspx?P_I...ZS5&templete=3
MSI MS-91F7
http://www.msi.com/product/server/MS-91F7.html
Now, as we all know, the SR5670 will be able to support bulldozer. These 2 boards are Pseudo ATX sized boards which will fit a home enthusiast's case easily. Unbuffered desktop memory is also supported.
To get high performance, simply rely on these sort of boards. They're relatively cheap, the processors are relatively cheap. at ~700USD, (2*Opteron 4180) for the board and processors, they can quite conceivably beat intel's flagship model (i7 980x) at multithreaded benchies. Perhaps take one of these boards, throw on a couple of heatsinks to make it look the part while inflating the price. Ignore processor overclocking (or relabel the chipset to something like 990FX2 to disassociate it before allowing some limited adjustments), maintaining the same SKU as the server parts, allow slight memory frequency adjustment to DDR3 1600mhz which bulldozer is supposed to use, (since most sticks can't do that right now) to make sure bulldozer runs optimally.
Crossfire support with a couple more PCI-E x16 slots to make it look attractive, screw Nvidia's licensing to keep the cost down. Let's face it, most enthusiasts don't need that much processing performance. They like.. play games. Just gotta convince them they need it and should choose it over sandy bridge.
Send it for reviews for mindshare and emphasise on 'zomg we can match/beat the i7 at half the price'. Rinse and repeat it with bulldozer against sandy bridge.
Oh come on, remember the dual GPU strategy pursued by ATi/AMD? Slap 2 relatively lower performance parts and beat the competition by being 'the fastest single card' available.
Btw, I'm getting the MSI MS-91F7
Last edited by Fatfool; 01-17-2011 at 05:09 AM.
NO.with all of your post you seem to think amd is doomed if it doesnt do as you wish ....
but the reality is amd isnt doing bad these days ... so lets give them credit for that .. and let them do what they know is working for them OK?
I did not and dont think amd is doomed if it leaves this segment to the Intel.And i never said so,nor even implied it anywhere.
But i think it would be good for AMD to get back to high performance markets.
And the reason i think AMD should consider getting back to the high end is exactly what you said, they are doing better,and most probably have architecture that they could use to do that.
Im rooting for AMD here.But not blindly following and agreeing with all the things theyre doing.
Corporations underestimate markets, as well as they overestimate them.Corporate policies arent always right, even for them.
And more importantly if we the clients, will always agree with everything they do, we will get inferior products for exorbitant prices.
I dont get this "they must know what theyre doing, theyre big corporation" kind of line of thinking.
If that was the case, AMD wouldnt be in the trouble they were not long ago.
And for the record, i own mostly AMD hardware.
ps.Just look at the ati part of AMD, theyre doing sensible chips,not going crazy, but still theyre competing in the high end markets in smart ways.
Actually after reading this post, I think that's the root I'm going to take with my new build. Thanks for the links.Personally, I really don't see the point of pushing overclockable 2P systems at this point. It'll be a vanity project.
This seems like a non issue since a version of what was suggested already exists, this time not some 4x4 'I run hotter and suck at performance' quadfire.
Asus KCMA-D8
http://www.asus.com/product.aspx?P_I...ZS5&templete=3
MSI MS-91F7
http://www.msi.com/product/server/MS-91F7.html
Now, as we all know, the SR5670 will be able to support bulldozer. These 2 boards are Pseudo ATX sized boards which will fit a home enthusiast's case easily. Unbuffered desktop memory is also supported.
To get high performance, simply rely on these sort of boards. They're relatively cheap, the processors are relatively cheap. at ~700USD, (2*Opteron 4180) for the board and processors, they can quite conceivably beat intel's flagship model (i7 980x) at multithreaded benchies. Perhaps take one of these boards, throw on a couple of heatsinks to make it look the part while inflating the price. Ignore processor overclocking (or relabel the chipset to something like 990FX2 to disassociate it before allowing some limited adjustments), maintaining the same SKU as the server parts, allow slight memory frequency adjustment to DDR3 1600mhz which bulldozer is supposed to use, (since most sticks can't do that right now) to make sure bulldozer runs optimally.
Crossfire support with a couple more PCI-E x16 slots to make it look attractive, screw Nvidia's licensing to keep the cost down. Let's face it, most enthusiasts don't need that much processing performance. They like.. play games. Just gotta convince them they need it and should choose it over sandy bridge.
Send it for reviews for mindshare and emphasise on 'zomg we can match/beat the i7 at half the price'. Rinse and repeat it with bulldozer against sandy bridge.
Oh come on, remember the dual GPU strategy pursued by ATi/AMD? Slap 2 relatively lower performance parts and beat the competition by being 'the fastest single card' available.
Btw, I'm getting the MSI MS-91F7
As quoted by LowRun......"So, we are one week past AMD's worst case scenario for BD's availability but they don't feel like communicating about the delay, I suppose AMD must be removed from the reliable sources list for AMD's products launch dates"
No problems.This is a cost effective, higher performance solution compared to sandy bridge. AMD has it on their hands, yet (apparently) didn't bother to push for 2P workstation systems/flagship desktop systems, leaving it to their partners to go do whatever they wanted with it. I'll leave JF to refute my allegations if I am mistaken. I mean come on, 700USD for 12 2.6ghz or 900USD for 12 2.8ghz cores plus the board cost? How does an i7 980x match that in price ? (perhaps performance too) Plus it's actually upgradable unlike the X58 platform.Actually after reading this post, I think that's the root I'm going to take with my new build. Thanks for the links.
AMD's used server parts before for flagship models. The Original Athlon FX (socket 940) series, the quad FX (this was a fail though). And at the price the Opteron 4100s are going for (True blue Desktop prices finally! I hope the era of inflated server processors which really were nothing more than repackaged desktop ones is over; not talking bout those 12 core ones which don't have a desktop variant), it seems only logical to keep floggin' as many as possible anyway.
Be careful. You can't really overclock with these boards, and your CPUs aren't going to have that high of clock speeds. I had a dual Istanbul setup on a dual-NB SR5690 board a while back. It did pretty well at threaded work, but I eventually replaced it a year and a half later when Thuban came out. An overclocked Thuban was just about as fast in threaded work as my 12-core machine was and ungodly faster in single threaded stuff, which is important for games.
Particle's First Rule of Online Technical Discussion:
As a thread about any computer related subject has its length approach infinity, the likelihood and inevitability of a poorly constructed AMD vs. Intel fight also exponentially increases.
Likewise, the frequency of a car pseudoanalogy to explain a technical concept increases with thread length. This will make many people chuckle, as computer people are rarely knowledgeable about vehicular mechanics.
Rule 2:
When confronted with a post that is contrary to what a poster likes, believes, or most often wants to be correct, the poster will pick out only minor details that are largely irrelevant in an attempt to shut out the conflicting idea. The core of the post will be left alone since it isn't easy to contradict what the person is actually saying.
Rule 2A:
When a poster cannot properly refute a post they do not like (as described above), the poster will most likely invent fictitious counter-points and/or begin to attack the other's credibility in feeble ways that are dramatic but irrelevant. Do not underestimate this tactic, as in the online world this will sway many observers. Do not forget: Correctness is decided only by what is said last, the most loudly, or with greatest repetition.
Rule 3:
When it comes to computer news, 70% of Internet rumors are outright fabricated, 20% are inaccurate enough to simply be discarded, and about 10% are based in reality. Grains of salt--become familiar with them.
Remember: When debating online, everyone else is ALWAYS wrong if they do not agree with you!
Random Tip o' the Whatever
You just can't win. If your product offers feature A instead of B, people will moan how A is stupid and it didn't offer B. If your product offers B instead of A, they'll likewise complain and rant about how anyone's retarded cousin could figure out A is what the market wants.
And a Bulldozer module being 12% larger than one core (or what one core would be if they didn't use modules) and actually have the ability to process another thread (instead of just using idle resources of the one core)
You are comparing 8 virtual threads with 8 actual threads.
Do you understand how hyperthreading works? It doesn't make each core process two instructions simultaneously.
And that's how Bulldozer modules are superior. An even smaller increase in die area (HT needs +30%, BD module needs +12%) for the ability to actually process two threads simultaneously.
Last edited by Apokalipse; 01-17-2011 at 08:40 AM.
Agreed, but remember BD is going to have Turbo, even in server parts (correct?), so it wil help in more casual PC usage scenarios.
Leopards don't change their spots it seems.
It does. RTFM!
Wrong.An even smaller increase in die area (HT needs +30%, BD module needs +12%)
No it doesn't. It uses the extra thread to reduce the amount of time the core spends idle whenever there's a stalled thread.
The Operating system sees two cores. That's just to allow the extra thread to be issued. It can't process them simultaneously.
Elaborate.....
Last edited by Apokalipse; 01-17-2011 at 10:26 AM.
Posting Permissions
Reply With Quote
Bookmarks