Bulldoze that Sandy Bridge

Let's take a look at what we do know about bulldozer

Shared L1 instruction cache for two cores in one module (no size yet Although I could go with 128Kbytes for speculation)
separate L1 data cache 16 Kbytes per core
2Mbs L2 cache shared for two cores in one module.
8Mbs of 2400mhz L3 cache
1866mhz memory
SSSE3 SSE4.1 SSE4.2 AVX I'm sure more I didn't list

I think it would have less missed IPC then phenom II.
don't know if L1/L2 cache in Exclusive or inclusive type yet.

sandybridge doesn't worry me at all, it's like K8<K10/K10.5 really with a few bonuses.

By labelling them FX, I assume that they might be coming out with a dual cpu platform for enthusiasts. Like Quad FX?

I think no. Quad FX was a try to get back the "FX" Performance crown just using 2 processors because one didnt do it. It was a flop. I dont think they will try that again.

In Socket 939 and earlier times, FX was just the highest clocked Athlon, which indeed was the fastest gaming processor on the market those times. I hope they will revive that spirit again (which is very likely, even if its going to be slightly less fast then some intel when it comes out..)

Quote:

---

Originally Posted by demonkevy666

Let's take a look at what we do know about bulldozer

Shared L1 instruction cache for two cores in one module (no size yet Although I could go with 128Kbytes for speculation)
separate L1 data cache 16 Kbytes per core
2Mbs L2 cache shared for two cores in one module.
8Mbs of 2400mhz L3 cache
1866mhz memory
SSSE3 SSE4.1 SSE4.2 AVX I'm sure more I didn't list

I think it would have less missed IPC then phenom II.
don't know if L1/L2 cache in Exclusive or inclusive type yet.

sandybridge doesn't worry me at all, it's like K8<K10/K10.5 really with a few bonuses.

---

Cache is exclusive.

IPC will be higher.

We have not said anything on memory other than 1600MHz on server.

Quote:

---

Originally Posted by JF-AMD

Cache is exclusive.

IPC will be higher.

We have not said anything on memory other than 1600MHz on server.

---

Will there ever be an enthusiast platform, like the Quad FX or would you recommend going to socket C32 or G34 for multiple CPU?

There will not be a dual socket client product. Server products are not designed to be overclocked.

The dual socket client market, at its peak, was .8% of the client market. Then it fell to .4% of the market. I have no idea where it is today, but with 6 and soon, 8 cores available, the need for dual socket is practically non-existent. It was bigger when two dual cores got you to four total cores.

Quote:

---

Originally Posted by JF-AMD

Cache is exclusive.

IPC will be higher.

We have not said anything on memory other than 1600MHz on server.

---

I think you missed read what I wrote.

as I said less missed IPC then phenom II which I meet would be Higher IPC then phenom II.

yes I know you work with Sever not client.

Desktop roadmaps showed 1866mhz a while a ago.

you must be getting annoyed with everyone saying it only has 12.5% over current cores.

I know I am and I don't even work for amd lol

Quote:

---

Originally Posted by JF-AMD

There will not be a dual socket client product. Server products are not designed to be overclocked.

The dual socket client market, at its peak, was .8% of the client market. Then it fell to .4% of the market. I have no idea where it is today, but with 6 and soon, 8 cores available, the need for dual socket is practically non-existent. It was bigger when two dual cores got you to four total cores.

---

( Looks around the house and sees 4 dual socket machines)
I won't argue your numbers but for some of us the dualies present some real world advantages.
1)They are electrically more efficient for the work done than the singles
2) The boards are generally much better built and last longer
I have SM boards that are up to 10 years old and still function as they did on day one
3) Space: I can get X amount of work done in a smaller footprint than if I went to single socket boards.
4) They have more E-PEEN!:rofl:

Number 1+2 are the key to me.
Inital cost is heavier but to me knowing that I'm cutting my electrical costs EVERY day for the 2-3 years between upgrades makes my decisions.
In this day of 16 cents per KW/H it is the smart move.
The added initial cost is recovered over that timeframe and the dependability factor means no downtime form RMA's and such.
Since 2003 when I first went to a dualie ( SM X5DAE) I've never had an issue with a board.

Quote:

---

Originally Posted by Movieman

( Looks around the house and sees 4 dual socket machines)
I won't argue your numbers but for some of us the dualies present some real world advantages.
1)They are electrically more efficient for the work done than the singles
2) The boards are generally much better built and last longer
I have SM boards that are up to 10 years old and still function as they did on day one
3) Space: I can get X amount of work done in a smaller footprint than if I went to single socket boards.
4) They have more E-PEEN!:rofl:

Number 1+2 are the key to me.
Inital cost is heavier but to me knowing that I'm cutting my electrical costs EVERY day for the 2-3 years between upgrades makes my decisions.
In this day of 16 cents per KW/H it is the smart move.
The added initial cost is recovered over that timeframe and the dependability factor means no downtime form RMA's and such.
Since 2003 when I first went to a dualie ( SM X5DAE) I've never had an issue with a board.

---

Lol, JF-AMD already went through the same thing with me on the [H] forums. I think my last shot was at the feasibility of C32, since it seemed like half a G34 to me (so why not get 1 G34, lol), but he mentioned company research already showed there was sufficient interest with large enough orders to allow devoting precious company engineering (and by relation, marketing - and no, these aren't the pretty artwork guys) resources over to this. Same thing for enthusiast dual socket. Research in this direction, however, pointed to a very minority market share and interest overall to even bother. I never got a clear answer whether is was even possible for a 3rd party to build an OCable G34 board or something, but I never directly asked that question, either.

He likes mentioning the Quad FX issue, but IMO, back then, going from dual to quad was already a given for the next gen. But nowadays, going from 12-->24 isn't going to be as quick, nor likely (just IMO). While I don't think AMD should devote resources over to this (Intel doesn't, after all), I do hope it's actually possible to BCLK OC an Opty through a 3rd party (looking at Sapphire's snab of eVGA's mobo team, lol).

Quote:

---

Originally Posted by jeremyshaw

Lol, JF-AMD already went through the same thing with me on the [H] forums. I think my last shot was at the feasibility of C32, since it seemed like half a G34 to me (so why not get 1 G34, lol), but he mentioned company research already showed there was sufficient interest with large enough orders to allow devoting precious company engineering (and by relation, marketing - and no, these aren't the pretty artwork guys) resources over to this. Same thing for enthusiast dual socket. Research in this direction, however, pointed to a very minority market share and interest overall to even bother. I never got a clear answer whether is was even possible for a 3rd party to build an OCable G34 board or something, but I never directly asked that question, either.

He likes mentioning the Quad FX issue, but IMO, back then, going from dual to quad was already a given for the next gen. But nowadays, going from 12-->24 isn't going to be as quick, nor likely (just IMO). While I don't think AMD should devote resources over to this (Intel doesn't, after all), I do hope it's actually possible to BCLK OC an Opty through a 3rd party (looking at Sapphire's snab of eVGA's mobo team, lol).

---

Yea, John's a good guy and well intentioned but we differ on our opinions in this.That's ok, I don't take this stuff personal.
Back with the dual FX boards the timing was wrong and so wasnt the product.
Just too much heat to deal with from what I saw at the time.
We had a guy on WCG build one and to be honest he wasn't happy with the results.
The unfortunate part is that if a company does try something one time and it fails they are predisposed to never try that again and I can't argue with that thinking BUT I think one has to look at the advances that have been made and see it in todays technology.
Now I have a dual socket Asus KGPE-D16 with 2-6168 MC chips( 24 cores total) that runs 100% load 24/7 since last April I think and is as solid a machine as one could ask for. ZERO downtime, loaded temps at app 41C in a 72F room in a open case with the Dynatron A6 HS( best available at the time)
Would I like to see if it could OC?
You better beleive I would and have tried all that I can do find a way to OC it but no luck.
I have gotten to the mental point that it is what it is and will have to stay that way for it's lifespan.

Quote:

---

Originally Posted by jeremyshaw

I never got a clear answer whether is was even possible for a 3rd party to build an OCable G34 board or something, but I never directly asked that question, either.

---

Anyone can build anything that they want. However the processor will never have unlocked multipliers.

If the guy who would pick up a grand per proc on a 2P client system says there isn't enough of a market to be financially viable, what would make a motherboard manufacturer want to build a board in order to capture an even smaller share of the revenue per system?

The market just isn't there.

Quote:

---

Originally Posted by Oese

In Socket 939 and earlier times, FX was just the highest clocked Athlon, which indeed was the fastest gaming processor on the market those times. I hope they will revive that spirit again (which is very likely, even if its going to be slightly less fast then some intel when it comes out..)

---

For some AMD trivia, before socket 939 the FX wasn't in a mainstream plattform. IIRC the FX-51 debuted as a socket 940 part along the 754 introduction, requiring buffered RAM like an opteron. For a short time it was 200MHz faster than the 754 flagship.

Quote:

---

Originally Posted by JF-AMD

The dual socket client market, at its peak, was .8% of the client market. Then it fell to .4% of the market.

---

If the parts are not available of course the market will be small. How can you buy something that doesn't exist?

Quote:

---

I have no idea where it is today, but with 6 and soon, 8 cores available, the need for dual socket is practically non-existent. It was bigger when two dual cores got you to four total cores.

---

There are plenty of enthusiasts and professionals that need more power than a 1 socket system can provide. I can already think of gamers and the artists at CGSociety off the top of my head.

The Quad FX platform sucked in performance, value, heat output, power consumption... etc. Of course it didn't take off, was it supposed to?

You're dealing with enthusiasts, people who research their products beforehand. We don't buy poor performing products.



I like the C32 platform and hopefully that will take off. Maybe in the future we can see a C32-like platfrom with more enthusiast features. Crossfire, SLi, overclocking... etc.

If the chips put up good numbers, I see no reason for EVGA not to make an Amd version of the SR-2. ;)

Quote:

---

Originally Posted by 445533

If the parts are not available of course the market will be small. How can you buy something that doesn't exist?



There are plenty of enthusiasts and professionals that need more power than a 1 socket system can provide. I can already think of gamers and the artists at CGSociety off the top of my head.

The Quad FX platform sucked in performance, value, heat output, power consumption... etc. Of course it didn't take off, was it supposed to?

You're dealing with enthusiasts, people who research their products beforehand. We don't buy poor performing products.



I like the C32 platform and hopefully that will take off. Maybe in the future we can see a C32-like platfrom with more enthusiast features. Crossfire, SLi, overclocking... etc.

---

sorry but your only thinking for yourself there ... amd is thinking for the good of the investors ... and the stock holders ... nothing else .. and if amd says the market is too small to even think of investing into client side workstation ... then its for the best .. id want me some dually who oc .. but then again it wont happen soon ...

and the few people that need more then 1 socket they allready have 2 or 4 socket systems ... so ???

JF, convince me to cancel my i5-2500k order :)

Quote:

---

Originally Posted by ChanceCoats123

If the chips put up good numbers, I see no reason for EVGA not to make an Amd version of the SR-2. ;)

---

eVGA's mobo team left for the Sapphire pastures. Root for sapphire to deliver!! :D

Quote:

---

Originally Posted by richierich

JF, convince me to cancel my i5-2500k order :)

---

"...different markets..."

^^what I would say.

No, I want JF to convince me to cancel my 2500k order by showing me something (I'm not part of the dual-socket crowd, I'm talking about mainstream Bulldozer). Throw me a bone willya JF? lol.

Quote:

---

Originally Posted by richierich

No, I want JF to convince me to cancel my 2500k order by showing me something (I'm not part of the dual-socket crowd, I'm talking about mainstream Bulldozer). Throw me a bone willya JF? lol.

---

he clearly said amd wont show benchmark before official launch ... and it wont change ....

I know sn0wm@n, I'm joking. Well I'm most likely going to play with a 2500k until AMD releases BD, and if BD disappoints, well I'll just have to keep waiting until AMD releases something revolutionary.

Quote:

---

Originally Posted by JF-AMD

Anyone can build anything that they want. However the processor will never have unlocked multipliers.

If the guy who would pick up a grand per proc on a 2P client system says there isn't enough of a market to be financially viable, what would make a motherboard manufacturer want to build a board in order to capture an even smaller share of the revenue per system?

The market just isn't there.

---

Ah, thank you very much :)

See, I'm not too bright, and rather dim, so I interpreted one of your other messages as "well not allow our partners to build OCable "entusiast" Opteron mobos"


^^ good example of me o.0

Quote:

---

Originally Posted by richierich

I know sn0wm@n, I'm joking. Well I'm most likely going to play with a 2500k until AMD releases BD, and if BD disappoints, well I'll just have to keep waiting until AMD releases something revolutionary.

---

gotcha!!!


i doubt bulldozer would disapoint .. but then again gotta be prepared for everything i guess

Quote:

---

Originally Posted by Sn0wm@n

sorry but your only thinking for yourself there ... amd is thinking for the good of the investors ... and the stock holders ... nothing else .. and if amd says the market is too small to even think of investing into client side workstation ... then its for the best .. id want me some dually who oc .. but then again it wont happen soon ...

---

And look how good that turned out right? LOL.

Quote:

---

Originally Posted by 445533

If the parts are not available of course the market will be small. How can you buy something that doesn't exist?



There are plenty of enthusiasts and professionals that need more power than a 1 socket system can provide. I can already think of gamers and the artists at CGSociety off the top of my head.

The Quad FX platform sucked in performance, value, heat output, power consumption... etc. Of course it didn't take off, was it supposed to?

You're dealing with enthusiasts, people who research their products beforehand. We don't buy poor performing products.



I like the C32 platform and hopefully that will take off. Maybe in the future we can see a C32-like platfrom with more enthusiast features. Crossfire, SLi, overclocking... etc.

---

So, you are saying it is a good idea for a company to spend $5M in order to bring in a million in revenue?

The problem is not just the small size of the market and the fact that you will lose money. The real problem is that in spending the money to lose money you are bypassing the opportunity to spend the money and MAKE money. The $5M you invest to bring in a million in revenue would be better spent focusing on a product that can bring in $10M or $20M, right?

Given the choice between losing a few million and making a profit, where do you think a company should invest?

Quote:

---

Originally Posted by richierich

No, I want JF to convince me to cancel my 2500k order by showing me something (I'm not part of the dual-socket crowd, I'm talking about mainstream Bulldozer). Throw me a bone willya JF? lol.

---

The best way to convince you would be to let you buy it and then face buyers remorse later, right ;)

Quote:

---

Originally Posted by JF-AMD

So, you are saying it is a good idea for a company to spend $5M in order to bring in a million in revenue?

The problem is not just the small size of the market and the fact that you will lose money. The real problem is that in spending the money to lose money you are bypassing the opportunity to spend the money and MAKE money. The $5M you invest to bring in a million in revenue would be better spent focusing on a product that can bring in $10M or $20M, right?

Given the choice between losing a few million and making a profit, where do you think a company should invest?

---

I think you should hire me to oversee your new dual socket workstation division..:rofl:
Here's a thought for you "outside the box" as they say.
Since dualies do exist and will continue at least for the foreseeable future and they have always been regarded as part of the server division and validating server boards is more $$ than desktop,why not turn them over to the desktop division and be rid of all us crazed OC'ers.:D
I can hear you now:
Either your laughing or cursing my name..:rofl:

I'm interested in what the overall IPC increase is, I'd hope it would be a fair amount given how long Bulldozer has been in the works. AMD knows what they are up against, and they need to shine. Server side may be okay with higher core count vs. ipc but the desktop needs a large IPC increase from Phenom 2.

Quote:

---

Originally Posted by JF-AMD

The best way to convince you would be to let you buy it and then face buyers remorse later, right ;)

---

Just watch all the rumor mill sites take this as: JF-AMD states Bulldozer will ROFLPWN Sandy Bridge!! :D :D

Quote:

---

Originally Posted by jeremyshaw

Just watch all the rumor mill sites take this as: JF-AMD states Bulldozer will ROFLPWN Sandy Bridge!! :D :D

---

I think it will with ease! [for server workloads]
8 cores vs 4 cores + HT, in most well threaded apps 4 module BD shouldn't have problem with i2600K. Of course Intel will have their 8 core SB on the market before end of year, but the question was about 4 core variant :).

I wouldn't put much hope in SuperPi scores on AMD to catch SB, but gaming should be much better than on K10 and that means at least not so much behind SB.

I really can't wait to see even Stream benches, if not more! Where is s7e9h3n when you need him most :p:

Quote:

---

Originally Posted by Lightman

I think it will with ease! [for server workloads]
8 cores vs 4 cores + HT, in most well threaded apps 4 module BD shouldn't have problem with i2600K. Of course Intel will have their 8 core SB on the market before end of year, but the question was about 4 core variant :).

I wouldn't put much hope in SuperPi scores on AMD to catch SB, but gaming should be much better than on K10 and that means at least not so much behind SB.

I really can't wait to see even Stream benches, if not more! Where is s7e9h3n when you need him most :p:

---

Intel already have 8 core Nehalem (Beckton), much less 8 core SB (which due to the ring bus, should be easier to scale to many cores :)).

Quote:

---

Originally Posted by zalbard

Same. A 200-300MHz improvement would be meh.

---

If it comes in at 3GHz with an IPC to match a 5GHz Thuban, who really cares what the stock frequency is? It's just an e-peen measure on its own;)

Quote:

---

Originally Posted by [XC] Oj101

If it comes in at 3GHz with an IPC to match a 5GHz Thuban, who really cares what the stock frequency is? It's just an e-peen measure on its own;)

---

I don't think anyone promised insane IPC improvements. What you're suggesting is nowhere near being possible. We've seen what the new arch is about, so we can more or less tell what to expect, IPC wise.

Quote:

---

Originally Posted by Lightman

Where is s7e9h3n when you need him most :p:

---

AMEN! He normally crops up around this time.

Quote:

---

Originally Posted by zalbard

I don't think anyone promised insane IPC improvements. What you're suggesting is nowhere near being possible. We've seen what the new arch is about, so we can more or less tell what to expect, IPC wise.

---

It was an example. A 200MHz increase in clocks might blow us away:up:

Quote:

---

Originally Posted by zalbard

I don't think anyone promised insane IPC improvements. What you're suggesting is nowhere near being possible. We've seen what the new arch is about, so we can more or less tell what to expect, IPC wise.

---

my assumption is that we will see upper 3ghz at launch with 4ghz+ in turbo, that can beat a 5ghz thuban per core, but with 2 extra cores.

Quote:

---

Originally Posted by Manicdan

my assumption is that we will see upper 3ghz at launch with 4ghz+ in turbo, that can beat a 5ghz thuban per core, but with 2 extra cores.

---

I hope so mate! - I placed my bets on Bulldozer late last year and bought a full AM3 setup (due to rumourville saying it was compatible) - to say I was highly dissapointed to learn I have to ditch the most expensive motherboard I've ever bought (CHIV) is an understatement. Most AMD users I've seen have already jumped and bought AM3 boards also - I can see a high number of enthusiast users switching to Intel due to this (me included - And I've been an AMD FAN for like 15 years now!)

AMD had better deliver with bulldozer - thats not a threat, but a realistic attitude (for both shareholders and consumers) otherwise they will be relegated to either the mobile or graphics only markets with anything that can't at least keep up with SB.

The FX branding introduction is slightly re-assurring that it will beat SB - but won't a newer intel platform be due out by then? Ivybridge I think.

Who cares about power useage in the consumer market? My PC is on all the time along with my NAS, two TV's and all the other electricity useage here in the UK and yet still my Eleccy bill is the last of my worries - it's amongst the lowest 3 bills hitting my bank account. I can understand in a server environment about power efficiencies but couldn't care less with my PC at home - IT HAS TO PWN - not be left behind pending 'another motherboard' upgrade - Can you tell I'm dissapointed with the X6 CPU's? lol :down:

Quote:

---

Originally Posted by zoomee

I hope so mate! - I placed my bets on Bulldozer late last year and bought a full AM3 setup (due to rumourville saying it was compatible) - to say I was highly dissapointed to learn I have to ditch the most expensive motherboard I've ever bought (CHIV) is an understatement.

AMD had better deliver with bulldozer - thats not a threat, but a realistic attitude otherwise they will be relegated to either the mobile or graphics only markets with anything that can't at least keep up with SB.

The FX branding introduction is slightly re-assurring that it will beat SB - but won't a newer intel platform be due out by then? Ivybridge I think.

Who cares about power useage in the consumer market? My PC is on all the time along with my NAS, two TV's and all the other electricity useage here in the UK and yet still my Eleccy bill is the last of my worries - it's amongst the lowest 3 bills hitting my bank account. I can understand in a server environment about power efficiencies but couldn't care less with my PC at home - IT HAS TO PWN - not be left behind pending 'another motherboard' upgrade - Can you tell I'm dissapointed with the X6 CPU's? lol :down:

---

i thought the CHIV came out WAY after it was known that BD was only AM3+?
and power consumption of the x6 are incredible. i was worried that people would over volt them just a little and end up pulling 250W because the same thing would have done 170W in a quad. but it turned out to use up much less power per core than most of us thought possible. they really are as good as 45nm can get. so AMD will no doubt be able to do much better with 32nm and advanced turbo/power gating features. server or home, both will see awesome power consumption with BD. i honestly dont get why some people are shocked by the power draw of SB, it wasnt anything unexpected from a node shrink:shrug:

Quote:

---

Originally Posted by JF-AMD

Given the choice between losing a few million and making a profit, where do you think a company should invest?

---

Whichever option gives you the most long term profits.

If investing in the professional and enthusiast side of things increases your reputation, increases the household use of the word AMD, and it at least breaks even then yes. In the grand scheme of things making the enthusiasts happy means making everyone else aware of AMD. You want to go for the larger consumer market of the unwashed masses? You guys need a well executed approach with higher end products for enthusiasts and professionals to talk about and buy, advertising for the rest, and make it easier for developers to produce software for your products. The way nvidia does it. People buy hardware to run software. We don't buy hardware just for the hell of it.

I can see how the move to re-brand ATi to AMD was fairly smart in the grand scheme of things, to make people more aware of the AMD platform. Unfortunately the new AMD Radeon logo looks cheap... like it's a bargain brand.

etc.

I think that a more consumer, professional, and small business oriented dual CPU platform could have a positive impact. The C32 platform looks like its halfway there but its missing a few things. Overclocking is a big one.

Worthwhile to at least send a few interns and get some info from dual-CPU users and who ever else could use it. Graphics designers are a big one since they always need more memory and more computing power and there's more of them now that everyone and their dog has Photoshop and 3DSMax/Maya.

Quote:

---

Originally Posted by jeremyshaw

Intel already have 8 core Nehalem (Beckton), much less 8 core SB (which due to the ring bus, should be easier to scale to many cores :)).

---

Well, I knew that. That's the reason of '8 core SB' in my post, specifically pointing to Dusty Bridge architecture variant for LGA2011. That's where all the fun will happen end of this year.

rigt 4Q 2011 for IB. Think, again IB will be relased some months later Komodo. What about speculation about superpi 1M? How much time in second at default clock? I think, Zambezi get 1M about 14s at stock clocks (example 3500MHz chip). And you?:)

Quote:

---

Originally Posted by Manicdan

my assumption is that we will see upper 3ghz at launch with 4ghz+ in turbo, that can beat a 5ghz thuban per core, but with 2 extra cores.

---

More than 25% improvement in IPC :eek:
If so I would be very happy :)

Edit: If it OC like SB too (5ghz+ on water) i will wet my pants....

Quote:

---

Originally Posted by 445533

Whichever option gives you the most long term profits.

If investing in the professional and enthusiast side of things increases your reputation, increases the household use of the word AMD, and it at least breaks even then yes. In the grand scheme of things making the enthusiasts happy means making everyone else aware of AMD. You want to go for the larger consumer market of the unwashed masses? You guys need a well executed approach with higher end products for enthusiasts and professionals to talk about and buy, advertising for the rest, and make it easier for developers to produce software for your products. The way nvidia does it. People buy hardware to run software. We don't buy hardware just for the hell of it.

I can see how the move to re-brand ATi to AMD was fairly smart in the grand scheme of things, to make people more aware of the AMD platform. Unfortunately the new AMD Radeon logo looks cheap... like it's a bargain brand.

etc.

I think that a more consumer, professional, and small business oriented dual CPU platform could have a positive impact. The C32 platform looks like its halfway there but its missing a few things. Overclocking is a big one.

Worthwhile to at least send a few interns and get some info from dual-CPU users and who ever else could use it. Graphics designers are a big one since they always need more memory and more computing power and there's more of them now that everyone and their dog has Photoshop and 3DSMax/Maya.

---

Investing in a platform that used to be .8% of the market then fell to .4% of the market (an nobody knows how low it is today) is not a winning nor profitable strategy.

Also, when you consider the enthusiast piece of the market is only in the single digits (IIRC) and that only a small fraction of those enthusiasts actually buy the $1000 desktop chips, it is pretty clear that having the top performing processor is not how you make a lot of money.

If you think there is a halo effect from the top part, consider that enthusiasts who understand processors would never buy an AMD processor for $300 or $500 that performed worse than the corresponding $300 or $500 intel processor - they would buy the best performance for their budget. That top score only makes them feel good AFTER they buy, if it is the same brand.

And, before anyone jumps on me, I am in the server busines, so this is me looking in at the enthusiast business, I have no impact on what they do.

Quote:

---

Originally Posted by 445533

And look how good that turned out right? LOL.

---

i dont seem to be on the same page as you ... help me understand this post please ...

Quote:

---

Originally Posted by JF-AMD

The best way to convince you would be to let you buy it and then face buyers remorse later, right ;)

---

you've got some confidence their.

Quote:

---

Originally Posted by TheBlueChanell

I'm interested in what the overall IPC increase is, I'd hope it would be a fair amount given how long Bulldozer has been in the works. AMD knows what they are up against, and they need to shine. Server side may be okay with higher core count vs. ipc but the desktop needs a large IPC increase from Phenom 2.

---

this reminds me of song by queens of the stone age
"no one knows."

Quote:

---

Originally Posted by zalbard

I don't think anyone promised insane IPC improvements. What you're suggesting is nowhere near being possible. We've seen what the new arch is about, so we can more or less tell what to expect, IPC wise.

---

AMD in tight spot usually produces something wonderful.

Quote:

---

Originally Posted by Manicdan

my assumption is that we will see upper 3ghz at launch with 4ghz+ in turbo, that can beat a 5ghz thuban per core, but with 2 extra cores.

---

honestly sometimes I wonder why they didn't bother with a Phenom II x8 MCM.
Thuban is kind of a large die surely making less money off them vs x4 is a bit of waste.

Quote:

---

Originally Posted by eXa

More than 25% improvement in IPC :eek:
If so I would be very happy :)

Edit: If it OC like SB too (5ghz+ on water) i will wet my pants....

---

that is what I hope is minim IPC over phenom II.

we actually already had 5-15% with K8 vs K10(phenom I), and another 2-5% with Phenom II.

Quote:

---

Originally Posted by JF-AMD

Also, when you consider the enthusiast piece of the market is only in the single digits (IIRC) and that only a small fraction of those enthusiasts actually buy the $1000 desktop chips, it is pretty clear that having the top performing processor is not how you make a lot of money.

---

I agree, but mostly because $1000 is too much for a CPU, and the cost for ECC memory, the motherboard etc.

The EVGA SR-2 did a smart thing which allowed us to use regular Unbuffered Non-ECC DDR3 memory in a computer. This lowers the cost and as consumers and prosumers we have no need for such features.

Quote:

---

If you think there is a halo effect from the top part, consider that enthusiasts who understand processors would never buy an AMD processor for $300 or $500 that performed worse than the corresponding $300 or $500 intel processor - they would buy the best performance for their budget.

---

Currently a X6 1075T is comparable to about a i7 930 but costs 1/3 less. Same deal with motherboards. An X58 vs an AMD 870 the AMD is about half price. If I had a small budget AMD gives me more performance for my money.

If in the future AMD processor prices are going to meet Intel prices but offer less performance at all price points like in your example... why would anyone bother? Let alone enthusiasts.

Quote:

---

And, before anyone jumps on me, I am in the server busines, so this is me looking in at the enthusiast business, I have no impact on what they do.

---

Thank you for contributing to the discussion anyways.

Quote:

---

Originally Posted by demonkevy666

honestly sometimes I wonder why they didn't bother with a Phenom II x8 MCM.
Thuban is kind of a large die surely making less money off them vs x4 is a bit of waste.

---

It'd have to be really high binned parts to clock to a reasonable level with a voltage that makes the TDP compatible with AM3...if it's even possible. That or really low speed cores, which wouldn't go over well with the general public for a desktop platform. A pair of Propos cores would probably work reasonably well in a setup like that...if they weren't so slow.

Thuban die is definitely large though, nearly 80% larger than Deneb. Thuban seems to require less current per mm2 though.

--Matt

anyone else notice that AMD's CEO suddenly and unexpectadly resigned today? did Bulldozer get sandy-bridged? seemed like the market was rallying behind AMD what a horrible time to sack the CEO unless there is serious trouble ahead?

Quote:

---

Originally Posted by gordy

anyone else notice that AMD's CEO suddenly and unexpectadly resigned today? did Bulldozer get sandy-bridged? seemed like the market was rallying behind AMD what a horrible time to sack the CEO unless there is serious trouble ahead?

---

No input atm, Still on the urge to find out.. very shocking news!:shrug:

informal just posted linkto fud saying that BD should come close to Core i7 parts, but at lower cost. I wonder which Core i7 though, Gulftown or Westmere. Hopefully Westmere. Anyways, since Microcenter just e-mailed me saying that my order for the 2500K is on backorder I'm going to cancel it. I must resist going to the darkside.


Cmon AMD, pull through!

Quote:

---

Originally Posted by 445533

Worthwhile to at least send a few interns and get some info from dual-CPU users and who ever else could use it.

---

I know of one specific man they could go see ... or at least have a phone call with :rofl:

Quote:

---

Originally Posted by richierich

informal just posted linkto fud saying that BD should come close to Core i7 parts, but at lower cost. I wonder which Core i7 though, Gulftown or Westmere. Hopefully Westmere. Anyways, since Microcenter just e-mailed me saying that my order for the 2500K is on backorder I'm going to cancel it. I must resist going to the darkside.


Cmon AMD, pull through!

---

Six cores, so Westmere. That's not enough, though. SB is already faster per core, and s2011 will pull even further ahead.

Why did Dirk leave? :(

Apparently the board didn't agree with the direction Dirk wanted the company to move in.

Quote:

---

Originally Posted by El Gappo

Why did Dirk leave? :(

---

board chairman Bruce Claflin said that Meyer wasn’t the right person to lead AMD through its next phase of “significant growth… market leadership and… superior financial returns.”

Meyer is getting $8.6 million in severance pay including shares in AMD.

superior financial returns.:rofl::ROTF::rofl::ROTF:
------

AMD Seeks New CEO Who Can Challenge Intel, Tackle Tablet-Computer Market
In addition to stepping down as CEO, Meyer resigned from the board. The move capped a 15-year career at AMD, following stints at Intel and Digital Equipment Corp. His departure wasn’t the result of any specific financial or legal matter, said Drew Prairie, an AMD spokesman.

"My blood will always be green," Meyer said yesterday in an e-mailed statement, referring to the color of AMD's logo. "I wish the company well as AMD focuses successfully on the road ahead."

IMO, AMD is run by a bunch of clowns! :yepp:

Quote:

---

Originally Posted by Apokalipse

Apparently the board didn't agree with the direction Dirk wanted the company to move in.

---

Guess who's on the board?

Hector Ruiz. So for all the people who wanted him back, he is back. Like herpes (he never left).

Quote:

---

Originally Posted by 445533

Guess who's on the board?

Hector Ruiz. So for all the people who wanted him back, he is back. Like herpes (he never left).

---

I thought Hector got kicked off the board after he was linked to the Galleon Group LLC insider-trading case?

Quote:

---

Originally Posted by ReKcOlNu

board chairman Bruce Claflin said that Meyer wasn’t the right person to lead AMD through its next phase of “significant growth… market leadership and… superior financial returns.”

Meyer is getting $8.6 million in severance pay including shares in AMD.

superior financial returns.:rofl::ROTF::rofl::ROTF:
------

AMD Seeks New CEO Who Can Challenge Intel, Tackle Tablet-Computer Market
In addition to stepping down as CEO, Meyer resigned from the board. The move capped a 15-year career at AMD, following stints at Intel and Digital Equipment Corp. His departure wasn’t the result of any specific financial or legal matter, said Drew Prairie, an AMD spokesman.

"My blood will always be green," Meyer said yesterday in an e-mailed statement, referring to the color of AMD's logo. "I wish the company well as AMD focuses successfully on the road ahead."

IMO, AMD is run by a bunch of clowns! :yepp:

---

Dirk not just resigned from the CEO post but also quit the board. Seems like this is a very big upset for him and not as simple as just he is having different view with the board. Some power play may involved.
It is hard to imagine that the x86 world lost the viable alternative to Intel.

Quote:

---

Originally Posted by ReKcOlNu

I thought Hector got kicked off the board after he was linked to the Galleon Group LLC insider-trading case?

---

Well he was never formally charged so I don't know what happened. I don't see his name on the board of directors page: http://phx.corporate-ir.net/phoenix....ol-govcommcomp


Looks like you're right, he's gone from AMD.

Quote:

---

Originally Posted by JF-AMD

Investing in a platform that used to be .8% of the market then fell to .4% of the market (an nobody knows how low it is today) is not a winning nor profitable strategy.

Also, when you consider the enthusiast piece of the market is only in the single digits (IIRC) and that only a small fraction of those enthusiasts actually buy the $1000 desktop chips, it is pretty clear that having the top performing processor is not how you make a lot of money.

If you think there is a halo effect from the top part, consider that enthusiasts who understand processors would never buy an AMD processor for $300 or $500 that performed worse than the corresponding $300 or $500 intel processor - they would buy the best performance for their budget. That top score only makes them feel good AFTER they buy, if it is the same brand.

And, before anyone jumps on me, I am in the server busines, so this is me looking in at the enthusiast business, I have no impact on what they do.

---

JF, it's fascinating reading your posts. However I do have something to disagree with on this one.

I think there is a halo effect. I think Intel having the clear performance crown leads to people buying low/mid range Intel CPUs over AMD, because of a feeling they are better. It's never clear for casual shoppers which is the best performing combination of board/cpu/memory at any specific price point. Additionally they're not really fixed to a specific price point. There's movement in either direction as money goes to the graphics card or other parts. People just want an intel because the i7 is the business. I noticed a significant swing to AMD away from Intel when Intel were struggling to match AMD, and now that it's gone full circle and AMD are playing catchup I've noticed the same phenomenon in the other direction.

A good example of this is graphics cards. NVidia and ATI have always competed for the crown. ATI become more interesting at £80 a card if they have the best performing card at £400. Same for Nvidia.

However, I think it's only fair to put all of that in it's proper place. The high end enthusiast self build market is just simply a tiny part of the market and I can't see it driving AMD behaviour.

The only financially viable way is for you to produce chips for the server market, and chips for the big OEMs and then as an afterthought work out how to get a premium from us lot. I think we'd be kidding ourselves on if we thought we had any real influence on AMD at the big spending points.

So overall yes I do agree with you, but I do feel that the halo effect is present. After all, if it wasn't present then why would AMD be the best choice in the value conscious enthusiast market, if there wasn't the Intel halo effect to navigate around.

As far as what's good for enthusiasts.... Intel and AMD alternating the performance crown once a year. When Intel launch a new core AMD price and cut their new core to beat it, and Intel respond. When there is a slight advantage to one side we don't see it, they just make a little more profit, and the other side has to cut a little into their margins. But what we have now is a position where Intel have the undisputed performance crown and AMD even with the new CPUs look unlikely to decidedly take it from them. That's not good for enthusiast consumers.

Quote:

---

Originally Posted by Halk

Intel and AMD alternating the performance crown once a year.

---

That hasn't happened in awhile and AMD has been second place ever since Core 2 Duo came out and wiped the floor.

Quote:

---

Originally Posted by 445533

That hasn't happened in awhile and AMD has been second place ever since Core 2 Duo came out and wiped the floor.

---

That's my recollection too. A64s were where AMD started to beat Intel for quite some time, and at C2D it shifted in Intel's favour.

I'm not in any way qualified to draw a conclusion... but while Intel were playing catchup, and AMD were trying to stay ahead... did we not see the evolution from single to quad core, 64 bit processing go mainstream and clock speeds go up until they hit the wall? Did that pressure cause more advances than normal?

Quote:

---

Originally Posted by Origin_Unknown

I know of one specific man they could go see ... or at least have a phone call with :rofl:

---

There are so few of them in the world we could just call each one individually.

Quote:

---

Originally Posted by Origin_Unknown

I know of one specific man they could go see ... or at least have a phone call with :rofl:

---

Quote:

---

Originally Posted by Halk

JF, it's fascinating reading your posts. However I do have something to disagree with on this one.

I think there is a halo effect. I think Intel having the clear performance crown leads to people buying low/mid range Intel CPUs over AMD, because of a feeling they are better. It's never clear for casual shoppers which is the best performing combination of board/cpu/memory at any specific price point. Additionally they're not really fixed to a specific price point. There's movement in either direction as money goes to the graphics card or other parts. People just want an intel because the i7 is the business. I noticed a significant swing to AMD away from Intel when Intel were struggling to match AMD, and now that it's gone full circle and AMD are playing catchup I've noticed the same phenomenon in the other direction.

A good example of this is graphics cards. NVidia and ATI have always competed for the crown. ATI become more interesting at £80 a card if they have the best performing card at £400. Same for Nvidia.

However, I think it's only fair to put all of that in it's proper place. The high end enthusiast self build market is just simply a tiny part of the market and I can't see it driving AMD behaviour.

The only financially viable way is for you to produce chips for the server market, and chips for the big OEMs and then as an afterthought work out how to get a premium from us lot. I think we'd be kidding ourselves on if we thought we had any real influence on AMD at the big spending points.

So overall yes I do agree with you, but I do feel that the halo effect is present. After all, if it wasn't present then why would AMD be the best choice in the value conscious enthusiast market, if there wasn't the Intel halo effect to navigate around.

As far as what's good for enthusiasts.... Intel and AMD alternating the performance crown once a year. When Intel launch a new core AMD price and cut their new core to beat it, and Intel respond. When there is a slight advantage to one side we don't see it, they just make a little more profit, and the other side has to cut a little into their margins. But what we have now is a position where Intel have the undisputed performance crown and AMD even with the new CPUs look unlikely to decidedly take it from them. That's not good for enthusiast consumers.

---

There are 2 major groups, the processor aware and the processor unaware.

For the processor aware, they know the different processors, they know performance, and they will buy the best performer that fits in their budget. They represent about 20% (at most) of the market. "Halo" products don't work because these people are smart enough to know the processors and will choose the best one for their system, they will not be fooled into buying a slower product because there is some other benchmark on a part they are not buying.

For the other 80%+ of the market, they are "processor unaware", so they don't know about the halo part. They buy based on price and the features that they get.

I just can't find the place where people are smart enough to know about this halo part, but dumb enough to pay more for a slower part because of that halo part.

Quote:

---

Originally Posted by JF-AMD

For the other 80%+ of the market, they are "processor unaware", so they don't know about the halo part. They buy based on price and the features that they get.

---

So where does marketing fit in? If people are so oblivious, how does marketing work and why does a large company like Intel even need to advertise and promote their products?

And that is what i never do..i never buy a system before i know how fast my application will run on it!
I'am curious when i can see test results from a Bulldozer system with the application i want to use every day.
The Sandy Bridge was not out yet and i know everything i want to see..the day that NDA is lifted (even a little to early) you see everywhere reviews,testings,overclockings
like they are already weeks busy with it.
That makes the big difference between Intel and AMD..how long are the Magny cours on the market..i'am still waiting to see a first test result!
When i know that the Bulldozer can reach the 980x or even do better and the price is right..i will build a Bulldozer system.
If it's more worse or i can't see a testing result then i buy a 980x that i can put in my system or build a second system with Sandy Bridge :)

Hmm..where i belong now..the processor aware group or the processor unaware group if i don't get my results that i want to see ;)

JP.

Ive read this info before
Price is interesting

Quote:

---

Originally Posted by 445533

So where does marketing fit in? If people are so oblivious, how does marketing work and why does a large company like Intel even need to advertise and promote their products?

---

Please don't confuse marketing with advertising, they are 2 different disciplines. Advertising is usually handled by sales functions within most companies or by regional functions.

Marketing is about determining the features, planning the roadmap, positioning the product, developing the brand proposition and engaging the sales/regional go to market organizations.

Quote:

---

Originally Posted by JF-AMD

Please don't confuse marketing with advertising, they are 2 different disciplines. Advertising is usually handled by sales functions within most companies or by regional functions.

Marketing is about determining the features, planning the roadmap, positioning the product, developing the brand proposition and engaging the sales/regional go to market organizations.

---

Hence the word "market" and not "advert" :)

Quote:

---

Originally Posted by JF-AMD

Please don't confuse marketing with advertising, they are 2 different disciplines. Advertising is usually handled by sales functions within most companies or by regional functions.

Marketing is about determining the features, planning the roadmap, positioning the product, developing the brand proposition and engaging the sales/regional go to market organizations.

---

Right, my mistake. Advertising is a subset of marketing.

So why advertise if things are so clear cut? You have the oblivious herd and the tech savvy consumers plugged into the net with the latest benchmark data on your product.

Quote:

---

Originally Posted by JPQY

And that is what i never do..i never buy a system before i know how fast my application will run on it!
I'am curious when i can see test results from a Bulldozer system with the application i want to use every day.
The Sandy Bridge was not out yet and i know everything i want to see..the day that NDA is lifted (even a little to early) you see everywhere reviews,testings,overclockings
like they are already weeks busy with it.
That makes the big difference between Intel and AMD..how long are the Magny cours on the market..i'am still waiting to see a first test result!
When i know that the Bulldozer can reach the 980x or even do better and the price is right..i will build a Bulldozer system.
If it's more worse or i can't see a testing result then i buy a 980x that i can put in my system or build a second system with Sandy Bridge :)

Hmm..where i belong now..the processor aware group or the processor unaware group if i don't get my results that i want to see ;)

JP.

---

What you say is a function of execution prowess and expected performance. Intel knew SB will stomp everything across the board. The problem was that even if you didn't like SB for whatever reason, the second alternative was, guess what, I3/5/7s on the market. So Intel couldn't fail. When you know you'll be much better than the competition, you can relax the NDAs and so on.

What about BD. Fuad tells us it will close to current I7s, a turkish website than says 50% better than current I7 950s as per AMD presentations. Why isn't it compared to 980x ? Or a 2600k ? Had AMD said 50% better than a SB 2600k would have been awesome.

But 50% better than a 950 puts it near Gulftown of comparable clock in the select apps AMD tested it. Improvement compared to K10.5 ? Definately.
Enough to exceed Gulftown and threaten SB ? Don't think so.

Quote:

---

Originally Posted by 445533

Right, my mistake. Advertising is a subset of marketing.

So why advertise if things are so clear cut? You have the oblivious herd and the tech savvy consumers plugged into the net with the latest benchmark data on your product.

---

that is a question for intel

Quote:

---

Originally Posted by JF-AMD

that is a question for intel

---

well played

Hi,

Some info of the size die of Bulldozer???

Quote:

---

Originally Posted by 445533

So where does marketing fit in? If people are so oblivious, how does marketing work and why does a large company like Intel even need to advertise and promote their products?

---

Likely, those people buy the best HP/Dell/Apple/etc system they feel like they can justify. Just in my opinion, though. It's what I used to do before finding out what [H] and XS were.

Quote:

---

Originally Posted by cesariuth

Hi,

Some info of the size die of Bulldozer???

---

"8 cores bulldozer in less die space than 6 cores istanbul/thuban" I think was quoted before.

8 server cores in less die space than current lisbon.

Quote:

---

Originally Posted by JF-AMD

8 server cores in less die space than current lisbon.

---

But, the 8 cores server are the same 8 cores of bulldozer??? :)

Server: Interlagos (2 x 8 core MCM) and Valencia (8 core) support quad channel DDR3
Desktop: Zambezi supports dual channel DDR3

Quote:

---

Originally Posted by Apokalipse

Server: Interlagos (2 x 8 core MCM) and Valencia (8 core) support quad channel DDR3

---

?
How ? Since the dies are the same, you get Quad-Channel in Interlagos by using 2 dual channel dies. Valencia is dual channel.

Quote:

---

Desktop: Zambezi supports dual channel DDR3

---

Which btw is the same die as Valencia.

no, is to same, if people say this "Istanbul die is same as Thuban die". And of course, its not right!

Quote:

---

Originally Posted by FlanK3r

no, is to same, if people say this "Istanbul die is same as Thuban die". And of course, its not right!

---

Obviously not :D. Thuban has Turbo Core and DDR3. Maybe Lisbon is closer, though it still lacks Turbo Core.

its a bit diferent die.

Quote:

---

Originally Posted by savantu

?
How ? Since the dies are the same, you get Quad-Channel in Interlagos by using 2 dual channel dies. Valencia is dual channel.


Which btw is the same die as Valencia.

---

There will be 8 core model on G34 with 4 channel support. It will be a "special" part(MCM),with higher clock.It's not Valencia,but still it's effectively 8 core and on G34.

Quote:

---

Originally Posted by cesariuth

But, the 8 cores server are the same 8 cores of bulldozer??? :)

---

Bulldozer is the name of the core. There is an 8-core client zambezi and and 8-core server valencia. They are generally the same, but I cannot commit that the die sizes are exactly the same, so I only make a server comparison of valencia relative to lisbon. They may be the same, but I just don;t know their die size.

Quote:

---

Originally Posted by Apokalipse

Server: Interlagos (2 x 8 core MCM) and Valencia (8 core) support quad channel DDR3
Desktop: Zambezi supports dual channel DDR3

---

Valencia supports dual DDR-3

Quote:

---

Originally Posted by FlanK3r

no, is to same, if people say this "Istanbul die is same as Thuban die". And of course, its not right!

---

What's the difference then ? Size is the same, transistor count is the same, maybe microcode is a bit different, package is different,testing might be more thourough, other than that they are virtually identical.

Lisbon features HT Assist and has 3 HT links for connecting to other processors and I/O devices.

Thuban has Turbo CORE and some other features that Lisbon does not have.

The point that we have been making is that Valencia is smaller than Lisbon. We have not made any statements regarding thuban. The statement MAY be the same for Thuban, but nobody has officially made that statement.

Quote:

---

Originally Posted by FlanK3r

its a bit diferent die.

---

Thuban/Lisbon
http://techreport.com/r.x/phenom-ii-x6/thuban-die.jpg

Istanbul
http://techreport.com/r.x/opteron-2435/die-shot.jpg

Quote:

---

Originally Posted by JF-AMD

Lisbon features HT Assist and has 3 HT links for connecting to other processors and I/O devices.

Thuban has Turbo CORE and some other features that Lisbon does not have.

---

These are features which are ondie, but are not active, as a differentiators, just as Intel enables or disables HT,VT,etc. Such features are controlled by fuses in the die.

Not always, they are both different steppings. You do not understand what you are talking about.

As for the photos BeepBeep 2, keep in mind that you are only seeing the top layer and you are only seeing to a certain level of granularity. Lookingthe same is not the same as being the same. The agrument thrown out here is that they are the same die, and they are not. They may be very similar, but they are not the same.

Quote:

---

Originally Posted by JF-AMD

Not always, they are both different steppings. You do not understand what you are talking about.

...

---

Steppings aside, what is the major physical difference which would make those 2 dies different ?
Feel free to point at least one.

Quote:

---

Originally Posted by savantu

Steppings aside, what is the major physical difference which would make those 2 dies different ?
Feel free to point at least one.

---

Please remember that dies are not just made of one layer. There are several layers in these pics that have either been purposely eroded or are under the visible one...

Also, if you want a hint, look at the die outline. They aren't the same. :up:

Quote:

---

Originally Posted by BeepBeep2

Please remember that dies are not just made of one layer. There are several layers in these pics that have either been purposely eroded or are under the visible one...

---

There are 11 layers. I am fully aware of that.
I'm also aware of the fact the AMD uses the same design for different markets. They might be tuned and have different features active/disabled, but it is the same design, same layout,etc.
It is simply a matter of resources and payback. Having different designs for different markets isn't always worthwhile.

Intel does the same with desktops and 2S Xeons. There is no difference between a I7 980 and a Xeon 5680.
But Xeon MP has its own design with a different die. That's why it took 1 year longer to get it to market. You need to wait for the uarch to be done for the generic parts, than you take it in implement it for the specific part. Not only that, but you validate much more thoroughly and that takes extra time again.

Quote:

---

Also, if you want a hint, look at the die outline. They aren't the same. :up:

---

I have quite good eyesight and I say you look again. They are identical, bar for the the differences in colours and blurring.

Quote:

---

Originally Posted by savantu

Steppings aside, what is the major physical difference which would make those 2 dies different ?
Feel free to point at least one.

---

I pointed out several above and you even quoted that response. Please go back and read it.

Quote:

---

Originally Posted by savantu

I have quite good eyesight and I say you look again. They are identical, bar for the the differences in colours and blurring.

---

I'll tell you for the third time here to look again.

Look at the edge of each picture. What is on the outer edge of each die is NOT the same, not at all.

Quote:

---

Originally Posted by savantu

There are 11 layers. I am fully aware of that.
I'm also aware of the fact the AMD uses the same design for different markets. They might be tuned and have different features active/disabled, but it is the same design, same layout,etc.
It is simply a matter of resources and payback. Having different designs for different markets isn't always worthwhile.

Intel does the same with desktops and 2S Xeons. There is no difference between a I7 980 and a Xeon 5680.
But Xeon MP has its own design with a different die. That's why it took 1 year longer to get it to market. You need to wait for the uarch to be done for the generic parts, than you take it in implement it for the specific part. Not only that, but you validate much more thoroughly and that takes extra time again.



I have quite good eyesight and I say you look again. They are identical, bar for the the differences in colours and blurring.

---

They have different pinouts, even. I heavily doubt 1207 & C32 are somehow, magically pin compatible with AM2+/AM3. Fact is, Istanbul made it to the market long before Thuban. So by your own logic, Thuban =/= Istanbul. So far, only the E0 stepping chips have turbocore (AMD has had per core clock modulation based on workload since Agena, mind you ;)), an "anti-Agena CnQ" setup, in a twisted sense. Yeah, try to argue they had the functionality on every AMD hexcore design. Fact is, they don't.


And *just* incase if you are ignorant to the fact (I'm waiting for a staunch denial...), JF-AMD is John Fruehe, AMD's Server (and a few other things) PM. And if you think Product Marketing is drawing advertisements.... then... "wow" is all I will care to say in response.

AMD always made same dies for desktop/server cpus, and differentiated both by disabling/enabling parts of it,packaging and testing, but most importantly masks on which they were made were the same.
AMD did it because it was most sensible approach for them ,it saves a lot of money and engineering which they dont and didnt have.
It was the case with athlon MP`s, opterons based on athlon 64 and barcelona/shanghai dies, and as far as i know with phenom/thuban dies.
Thats why server parts never had bigger caches for example.Same production line,same mask,same die.Ive never seen a proof of it being different.
I know JF you work for the company, and for some (retarded) clients it may seem as a downside for the product.But we would have to see some proof to believe different.
Server parts dont HAVE more ht links, they dont have them disabled as desktop ones have ;-)
Its like intel is squeezing more money from people over HT, intel wants quite more money to NOT castrate 2600K over 2500K.This extortion type of business is pretty widespread these days.And AMD doesnt do it that much.

@Jeremy,
Do you believe TESLA cards from nvidia use different chip than desktop GFX cards ?
Think about it for a sec.

Quote:

---

Originally Posted by JF-AMD

I pointed out several above and you even quoted that response. Please go back and read it.

---

Yes, I've read it first time. As I've said then, the features you mention are found in all dies, in some they are active in others they are not.
That does not make it different designs.

Quote:

---

Originally Posted by BeepBeep2

I'll tell you for the third time here to look again.

Look at the edge of each picture. What is on the outer edge of each die is NOT the same, not at all.

---

Maybe I can't see the picture because of the trees, but frankly, except the blurring and different color tones, to me they seem identical.
Feel free to point with an arrow where you see the differences please.

Quote:

---

Originally Posted by jeremyshaw

They have different pinouts, even.

---

What has pinout to do with the design ? Do you think a K8 on socket 939 is different from a K8 on socket 940 ?
Or Prescott on socket 478 was different from a Prescott on socket 775 ?

Quote:

---

I heavily doubt 1207 & C32 are somehow, magically pin compatible with AM2+/AM3.

---

What I'm arguing has nothing to do with compatibility. I thought it was obvious.

Quote:

---

Fact is, Istanbul made it to the market long before Thuban. So by your own logic, Thuban =/= Istanbul.

---

Which reinforces the point. If server made it out before desktops, it means it has no extra features which needed more time to be implemented and validated. It was just a matter of marketing; supplying the highest margin market segment until production ramped up and desktops could be supplied.

Quote:

---

So far, only the E0 stepping chips have turbocore (AMD has had per core clock modulation based on workload since Agena, mind you ;)), an "anti-Agena CnQ" setup, in a twisted sense. Yeah, try to argue they had the functionality on every AMD hexcore design. Fact is, they don't.

---

You don't add turbocore with a stepping. Either that functionality is there from the uarch phase, when all the control logic, power delivery and measure points were thought out, or you're in for a full respin ( the designed is changes, layouts, number of transistors, etc ).
So the point is that all hexacore had turbocore in, but wasn't enabled in all products. Either because of time-to-market restrictions or complexity, it could be validation of the turbocore feature wasn't ready when the design was thought fully functional w/o turbocore. Thus they decided to ship it with the feature disabled and when the validation is done, they enable it.

Quote:

---

And *just* incase if you are ignorant to the fact (I'm waiting for a staunch denial...), JF-AMD is John Fruehe, AMD's Server (and a few other things) PM. And if you think Product Marketing is drawing advertisements.... then... "wow" is all I will care to say in response.

---

I'm fully aware of who he is. That doesn't mean he knows everything about server cpus or be completely open about all issues. Do you suppose he knows the instruction latencies in current CPUs or other really technical stuff ? I bet he doesn't. And frankly, he doesn't need to. He's not selling circuit design techniques, but functionality. He's selling performance per watt or per $, not L1 nanosecond access time.

I've met people on different forums which can eat JF for breakfast when it comes to the insides of AMD CPUs. They are SW designers, HW engineers, you name it. They know those quirks either because they need to in their daily activities or as a hobby.

Anyway, I prefer to leave the topic where it is and agree to disagree.

Quote:

---

Originally Posted by RaV666

AMD always made same dies for desktop/server cpus, and differentiated both by disabling/enabling parts of it,packaging and testing, but most importantly masks on which they were made were the same.

---

The word "always" is very dangerous word to use because it paints you into a corner.

I happen to know the differences between the parts. If I run out of Lisbon parts I can't put a Thuban die in a C32 package and sell it as a Lisbon. And if I run out of Istanbul I REALLY cannot put a Thuban die in a socket F1207 package.

Physically the dies may fit, but there are logic level changes between each of those dies. Each of the three is a different stepping, Istanbul is D0, Lisbon is D1 and Thuban is E0.

Quote:

---

Originally Posted by savantu

What has pinout to do with the design ? Do you think a K8 on socket 939 is different from a K8 on socket 940 ?
Or Prescott on socket 478 was different from a Prescott on socket 775 ?

---

Thought: how is 1207(F) different from.... 1207(C32)? Must be the same CPU/Socket? Interestingly enough, they aren't. For that matter, there are different revisions of 1207F... (or TYAN is really as bad as they seem) with and without split power planes.

Quote:

---

Originally Posted by savantu

Which reinforces the point. If server made it out before desktops, it means it has no extra features which needed more time to be implemented and validated. It was just a matter of marketing; supplying the highest margin market segment until production ramped up and desktops could be supplied.

---

So one can grab any two Thuban dies (hey, they have later stepping ;) - *must* be later and more advanced with even MORE "unlocked" functionality!!) and slap them onto a MCM package?

Quote:

---

Originally Posted by savantu

You don't add turbocore with a stepping. Either that functionality is there from the uarch phase, when all the control logic, power delivery and measure points were thought out, or you're in for a full respin ( the designed is changes, layouts, number of transistors, etc ).
So the point is that all hexacore had turbocore in, but wasn't enabled in all products. Either because of time-to-market restrictions or complexity, it could be validation of the turbocore feature wasn't ready when the design was thought fully functional w/o turbocore. Thus they decided to ship it with the feature disabled and when the validation is done, they enable it.

---

So basically, what AMD is doing doesn't fit your predefined idea of a stepping. That's fine. AMD is run on your opinion, anyways. It's heavily unlikely AMD was already exploring the idea with Agena's CnQ, nor is it likely that stepping D1 is different from E0. It's just sheer impossible, according to you. Just because you think some nonexistant functionality is there, doesn't mean it's even remotely similar to another idea in the next stepping.

Quote:

---

Originally Posted by savantu

Anyway, I prefer to leave the topic where it is and agree to disagree.

---

Not surprising. People with nothing to offer, love pretending to have the high horse to ride upon (and twisting words, but who knew?).

EDIT: I don't like dragging other people into my own junkpile, but post above me also adds in a lot of value.

Quote:

---

Originally Posted by JF-AMD

The word "always" is very dangerous word to use because it paints you into a corner.

I happen to know the differences between the parts. If I run out of Lisbon parts I can't put a Thuban die in a C32 package and sell it as a Lisbon. And if I run out of Istanbul I REALLY cannot put a Thuban die in a socket F1207 package.

Physically the dies may fit, but there are logic level changes between each of those dies. Each of the three is a different stepping, Istanbul is D0, Lisbon is D1 and Thuban is E0.

---

In this instance i would say its pretty possible, they are different masks and istanbul never made it to desktop.Different revisions.
And just a question, could you exchange barcelona and shanghai "parts" ?
And if not, isnt it because after initial production, they got fused/programmed with corresponding features and got their different code names ?.Same revision.
My logic is, yea, they cant be changed one for another, but its because they already got fused differently.Which doesnt change the fact that theyre the same physically.

Its like, phenom 920/940.They were the same as 910, BUT, 920/940 didnt work with am3.910 did.They were same chip with same revision, but different fusing/microcode.

And yes, i shouldnt use the word "always" :-) .But "most of the time" ;-P

Quote:

---

Originally Posted by RaV666

In this instance i would say its pretty possible, they are different masks and istanbul never made it to desktop.Different revisions.
And just a question, could you exchange barcelona and shanghai "parts" ?
And if not, isnt it because after initial production, they got fused/programmed with corresponding features and got their different code names ?.Same revision.
My logic is, yea, they cant be changed one for another, but its because they already got fused differently.Which doesnt change the fact that theyre the same physically.

Its like, phenom 920/940.They were the same as 910, BUT, 920/940 didnt work with am3.910 did.They were same chip with same revision, but different fusing/microcode.

And yes, i shouldnt use the word "always" :-) .But "most of the time" ;-P

---

This started with the discussion that these were the SAME die, not interchangeable.

They are not the same die. I know this to be a fact.

I am not sure where all of these people get the ideas that every chip has the exact same features and it is just a fusing option. There are several things that are fuseable, but some of the things that I have mentioned are not fuseable. Many of the layers may be the same, but not all of the layers are the same.

Can we all please drop this discussion?