So it's June...

[dr.walrus]

Where is my Bulldozer?

[SXRguyinMA]

You're welcome :D


http://mcclainelectrical.net/images/bulldozer.jpg

[dr.walrus]

That quite genuinely put a big grin on my face

[RogueOpportunist]

Guess that supposed "leak" from gigabyte about the 7th was wrong... quelle surprise. :rolleyes:

[dr.walrus]

Oh. Damn. (http://www.xbitlabs.com/news/cpu/display/20110601180002_AMD_Confirms_Delay_of_FX_Series_Bul ldozer_Microprocessors.html)

[billygoat333]

tsk tsk.

[blaze15301]

it will be here in September to October.

[mDust]

I lol'd at the bulldozer above too.

I think they're going to pull a classic AMD and release too little, too late again. It doesn't look like they're even trying to compete with Intel anymore.
We have the power-users and the money-savers...the consumer markets don't overlap anymore. Now would be a good time for a 3rd chip-maker to step up and pull the proverbial rug out from beneath these two complacent giants.

[x88x]

Now would be a good time for a 3rd chip-maker to step up and pull the proverbial rug out from beneath these two complacent giants.
:whistler:
http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&releasejsp=release_157&xhtml=true&prid=705184

[dr.walrus]

:whistler:
http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&releasejsp=release_157&xhtml=true&prid=705184

shuold have known you'd turn up here with something smart-ass to say :santa:

edit: you totally overuse the devil smiley. I don't like it. I much prefer this guy :santa:

:santa::santa::santa:

[slaveofconvention]

IDN if they're really dead in the water just yet - sure the current state of play isn't good - there's no denying that, but the Classic AMD comment? From my memory, Classic AMD moves were things like the K6 processors outclassing the P3's for a fraction of the price, then the 462 XP processors ourperforming the current P4s and finally the first Athlon 64's making what was left of the P4's go running crying for mummy.....

I'm a partial AMD Fanboy - or I used to be - actually I'm a performance fan lol - I had a K6-200, then a K-6 2 450, then a 1700XP and one of the 754 64 3000's - my current machine is an intel Q6700, and an intel 2600k system is very much in the works....

The only time they've really done the too-little-too-late thing I can recall was the failure of the Phenom to really compete with the i7.... or am I wrong here with my recollections?

[x88x]

shuold have known you'd turn up here with something smart-ass to say :santa:

edit: you totally overuse the devil smiley. I don't like it. I much prefer this guy :santa:

:santa::santa::santa:

lol I use the other a lot because it generally signifies a tongue-in-cheek response...or at least that's the way I read it...also, it's a lot easier to do ": P" than ": santa :" (minus the spaces and "s, obviously). ;)

OT though, I agree; it's annoying that they delayed launch...not that I'm exactly in a position to buy atm, but I was hoping to catch them on post-release price-drop.

As far as AMD 'too little too late', as far as I remember, the only time they've really done that was with the Turion and Phenom chips. Though, I think the Phenom/Athlon II's hit pretty much on the mark for what 90%+ of the market actually needs. Hell, I consider myself a power-user but I have yet to see my 1090T (or my C2Q 9450 before it) tapped out except when I'm folding. I'm perfectly happy with the "80% of the performance for 40% of the cost" strategy, especially when I don't even max out that '80%'. ;)

...oh, there was also the Slot A stuff...the only Thunderbird CPUs I've ever had to deal with were nightmares... :hurt:

I know I've said this before, but I really do think that as time goes on the CPU is going to become less and less important...at least the way things are going now anyway.. Most of the heavy lifting is being off-loaded to GPUs (and to a lesser extent FPGAs), and between the demand for CPU power leveling off and the available CPU power skyrocketing, I think in another few years we're at the very least going to settle firmly into a plateau where just about any CPU on the market will be indistinguishable from any other to 99%+ of users. ...of course, there could be some major change that comes up and creates a massive spike in CPU power demand and proves me wrong...but we'll deal with that if it happens. ;)

[dr.walrus]

I think it's a really true point that out biggest processing demands have always been and will continue to be graphics. I don't think it's true that we can generalise that across all computing, but with stuff like tablets? It's at least 50% of the overhead!

[Snowman]

i made exactly 2 forays into the intel world and ran back to the luscious pastures of amd both times. For the price of the pentium M and the celeron M chips I had I could have got a much better amd processor at the time but was at the mercy of my parents wanting a prebuilt system. No matter of ram upgrades or gpu upgrades ever made up for that huge bottleneck at the processor.

[dr.walrus]

for me it's just a question of cost and backwards compatibility

[AmEv]

Cyrix?

[x88x]

I think it's a really true point that out biggest processing demands have always been and will continue to be graphics. I don't think it's true that we can generalise that across all computing, but with stuff like tablets? It's at least 50% of the overhead!

I didn't say the heaviest processing loads are all graphics, I said they're mostly being shifted onto GPUs. ;) Ever since GPGPU stuff really took off, most of the things that have historically been the biggest non-gaming CPU consumers (rendering, video processing, etc) have been being moving to GPGPU because they benefit tremendously from the processing style of GPGPUs.


Cyrix?
What about them?

[RogueOpportunist]

Apparently, the currently available B0 and B1 stepping Zambezi/Bulldozer processors can function at around 2.50GHz/3.50GHz (nominal/turbo) clock-speeds and at such frequencies they cannot deliver performance AMD considers competitive, a person with knowledge of the situation said on Monday.

Reading something like that doesn't bode well with me, if the BD chips can't even be "competitive" at 3.5ghz then obviously this new architecture isn't as advanced performance wise as we're being lead to believe, I mean if they can't even "compete" with a 3.5ghz Zambezi at this point in time how do they expect anything to be better in the fall when Intel is going to be punching them in the face with Ivy Bridge?

I dunno, call me skeptical but I think the only bulldozer we're gonna need at the end of this is the bulldozer we'll have to use to clean up this mountain of bulls***.

[AmEv]

I mean, how are they in the competition?

[x88x]

You mean..this Cyrix? :whistler:
http://i428.photobucket.com/albums/qq3/x13931x/cyrix.png

AFAIK, the only thing left of Cyrix is the AMD Geode, that AMD bought from National Semiconductor in 2003 (after NS bought Cyrix back in 1997).

[AmEv]

Ehoh.


What other competitors are there...?

[RogueOpportunist]

Well the Cell chip was supposed to revolutionize computing as we knew it... But it's still a 1 trick console pony... ARM is making progress in the mobile segment with their Cortex chip and I could see them eventually working their way into the desktop market but I can't see them ever going toe to toe with Intel for high performance... I dunno, I had high hopes for AMD but at this point I have to admit I'm a little concerned.

I don't think AMD could ever "die" but with all this money tied up in R&D for BD I could see it crippling the company if it flops... and with all these delays and postponements I have a feeling they are seeing the same thing.

[slaveofconvention]

I actually also had a Cyrix PR133+ system which I bought second hand and lost in a burglary.... I can honestly say, I didn't mind it going lol

[x88x]

nVidia also keeps flirting with making desktop CPUs, building on the success they've been having with Tegra, but I haven't actually seen anything really concrete. I do really hope AMD survives if BD is a flop though....and not just because I'm a bit of an AMD fan. If they go under, Intel would have no competition in the CPU market...which regardless of your opinions on Intel I think we can all agree would be a very bad thing.

[AmEv]

No competition=Price it whatever the heck you want=Computers become ridiculously expensive.

[x88x]

Yep. I mean, Intel's prices are pretty high as it is, imagine how bad it would be if they were the only option.

[AmEv]

Well, soon, people won't buy new processors because they simply aren't able to.
Then the processor market crashes.
Then the whole computer market crashes.
People will be left with 5-year-old machines as the newest ones.

I must say, the Lego thing was chuckle-worthy.

[RogueOpportunist]

That's why I don't think AMD would ever "die" like some people speculate, there will always be room for them in the budget market and the fanboi support only grows that much stronger when AMD is the "underdog" so I think the company has at least some stability, plus no matter what happens with the latest battle in the CPU war AMD always has ATI to hold them over in the event something goes really wrong.

[dr.walrus]

I didn't say the heaviest processing loads are all graphics, I said they're mostly being shifted onto GPUs. ;) Ever since GPGPU stuff really took off, most of the things that have historically been the biggest non-gaming CPU consumers (rendering, video processing, etc) have been being moving to GPGPU because they benefit tremendously from the processing style of GPGPUs.


My point was, rendering and video processing ARE graphics. UI is all graphics. Pure number crunching? A relatively small amount of what a desktop actually does.

[mDust]

:whistler:
http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&releasejsp=release_157&xhtml=true&prid=705184

So what's happening with that? It's been pretty low-key for 5+ months...


IDN if they're really dead in the water just yet - sure the current state of play isn't good - there's no denying that, but the Classic AMD comment? From my memory, Classic AMD moves were things like the K6 processors outclassing the P3's for a fraction of the price, then the 462 XP processors ourperforming the current P4s and finally the first Athlon 64's making what was left of the P4's go running crying for mummy.....

I'm a partial AMD Fanboy - or I used to be - actually I'm a performance fan lol - I had a K6-200, then a K-6 2 450, then a 1700XP and one of the 754 64 3000's - my current machine is an intel Q6700, and an intel 2600k system is very much in the works....

The only time they've really done the too-little-too-late thing I can recall was the failure of the Phenom to really compete with the i7.... or am I wrong here with my recollections?
Both companies are doing fine, but they aren't really competing for marketshare lately. Instead, they have split the market into power-users and normal folk who like the higher performance:price ratio albeit with less overall performance.
I read a few reviews on the K6 vs PIII...K6 was faster for office programs and general use, PII and PIII outperformed it in graphical content and gaming by a huge margin.
Likewise, the P4 dominated all due to more and faster cache. The Pentium-Ds were unrivaled unless we include blowtorches...but that is still a close call. Core 2 were faster. i7s are faster. If history repeats itself, the next gen Intels will be faster. Every once in a while an AMD chip would score higher fps in some games, or score higher in some random synthetic benchmark for unknown reasons, but I don't recall AMD ever outperforming the Intel flagship of the time. In fact, a lot of the reviews I read compared mid to high level Intel chips to the latest and greatest AMD chip and the performance was usually about the same. An example of this. (http://www.hardwareheaven.com/reviews/1079/pg1/amd-phenom-ii-x6-1100t-black-edition-review-vs-intel-core-i7-950-introduction.html) That review compares the AMD flagship with the similarly priced i7-950 which have similar performance.
It's also generally been AMD playing catch-up with Intel for the last decade. Intel has been blazing the technology and performance path for too long now. Bulldozer is a good example...AMD is incorporating technology that Intel has been using for years. But it's not coming out for another 6 months apparently...which is also about the time Ivy Bridge releases...that's what I meant about 'too little too late.'

I'm fine with AMD manufacturing lower priced chips but it's my opinion that they should also produce some higher priced, high performance chips again....regardless of whether or not 90% of users need them. When I want to build a high-performance computer, I'd like more options than extreme-priced Intel chips. Intel can get away with charging $1000 or more for their aging high-end chips because there's no competition there. Maybe in 5 years I can look at options from nvidia.:crossed:

[x88x]

the P4 dominated all due to more and faster cache. The Pentium-Ds were unrivaled unless we include blowtorches...but that is still a close call.
That's not the way I remember it. Unless I'm seriously mistaken, the Athlon 64's thoroughly trounced the P4's and PD's in every way.


It's also generally been AMD playing catch-up with Intel for the last decade. Intel has been blazing the technology and performance path for too long now.
Again, the Athlon 64 was a very notable exception to this. With it, AMD was the first to introduce a 64-bit architecture (well, unless you count the failboat that was Itanium) and the first to integrate the memory controller into the CPU (something that it took Intel until the first-gen i-series to do).

Interesting that the last time AMD trounced Intel on the performance field was also the last time that they did a complete architecture redesign and the last time they introduced major industry innovations...and also the last major chip release before they acquired ATI (a move that I don't think they have yet fully recovered from financially). Coincidence? I think not. It does stand to reason that Intel more often pioneers major innovations though...their net income was more than 24 times AMD's last year...hell, their net income was more than twice AMD's total assets that year... o_O

[mDust]

That's not the way I remember it. Unless I'm seriously mistaken, the Athlon 64's thoroughly trounced the P4's and PD's in every way.


Again, the Athlon 64 was a very notable exception to this. With it, AMD was the first to introduce a 64-bit architecture (well, unless you count the failboat that was Itanium) and the first to integrate the memory controller into the CPU (something that it took Intel until the first-gen i-series to do).

Interesting that the last time AMD trounced Intel on the performance field was also the last time that they did a complete architecture redesign and the last time they introduced major industry innovations...and also the last major chip release before they acquired ATI (a move that I don't think they have yet fully recovered from financially). Coincidence? I think not. It does stand to reason that Intel more often pioneers major innovations though...their net income was more than 24 times AMD's last year...hell, their net income was more than twice AMD's total assets that year... o_O

I read some old reviews and it looks like I didn't give AMD enough credit. The FX chips were king for a while. I honestly don't remember that being the case...probably because they were all over $1000, which was not the section I was paying close attention to.:) I still wouldn't call it trouncing though. The reviews put the FX on top by a rather small margin in most benchmarks and the P4s and PDs were on top for several tests in different reviews...especially gaming which is probably what I looked at the most. I remember that the Pentium Ds were unrivaled overclockers with the right cooling...but at stock speeds they weren't as good as I thought.
I also didn't know that Intels original 64-bit chips were a failure and then when they saw AMD64 become successful, they cloned it and rebranded it Intel64...clever...:facepalm:
I apologize AMD.
If it helps get more high end AMD chips on the market again, I'll take one for the team and accept ATI as a gift so they can concentrate on Bulldozer and do it right.

Intel is so big because they have their fingers in other pies such as chipsets, mobos, etc. and they own much of the global enterprise market. AMD is no slouch when it comes to earning though.
AMDs bank account:Intels bank account::Michael Jordans bank account:Bill Gates bank account
...as we've all seen that comparison floating around the net. I guess the reason Intel is richer is because they are nerdier than AMD.

[AmEv]

Anyone know, how's the lesser-known competetion doing?

Transmeta?

*Dead 2009. Argh, why?????

[x88x]

Transmeta?

*Dead 2009. Argh, why?????
*cough*becausetheysucked*cough*

[AmEv]

*cough*productorcompany*cough*

[x88x]

lol, idk, tbh. (wow, that was a bad sentence)... I never actually used any of their products myself, but I remember reading lots of reviews of laptops that used their chips and the reviewers constantly talking about how slow they were...

[dr.walrus]

I had to write about the Crusoe as part of a paper last year, it was a fascinating bit of kit actually. Used an onboard emulator to accept native x86 instructions and convert them into native code. Pretty low power too..

Problem is, the power savings were never totally realised, speed requirements were definitely not met, they tried to get off the ground at the start of the Japanese recession, and, let's face it, what's the point in trying to emulate x86 with a general-purpose architecture - why not just use x86?

If there had been healthy competition in terms of different architectures out there, it would have had a chance. But we've got to face facts, there wasn't and there still isn't - heck, even Macs are rocking x86 now.

The resulting losses were big enough to bury Transmeta for good. The company actually launched the Efficeon, which fixed all the problems with the Crusoe and was a great little low power processor, but by then, the damage was done and noone considered Transmeta a serious alternative.