Intel’s 45nm Penryn Working!

[FrgMstr]

Great news for Intel today as it trails earlier reports of its profits dropping 39%. Intel CEO Paul Otellini has stated that Intel now has working samples of its upcoming processors codenamed “Penryn.” Not only do they have the processors working on test equipment but Otellini states that Intel have booted Penryn processors on Windows Vista, Windows XP, Linux, and Apple Mac OS-X operating systems. Intel's Daniel Snyder went on to say:

These are good indicators of how healthy our 45 nm manufacturing and future product designs are looking so far as we plan to begin manufacturing production in the second half of the year ramping to three (!) 300mm factories in 2008.

Obviously this does not bode well for AMD as they have yet to bring forth their new Barcelona architecture processors that will hopefully showcase AMD’s recent move to 65nm process.

 

[jcll2002]

sweet!

 

[jcll2002]

sweet!

since i apparently cant edit my post, ill make another.

These wont be out till Q2 of 08?

 

[JethroXP]

"...second half of the year..." sounds to me like anywhere from July to December of 2007.

 

[CptFalcon]

Anyone smell a 4Ghz OC? How I drool...

 

[Xilikon]

This seems like another nail in AMD's coffin as far as I'm concerned... I bet Intel will take the lead and not getting any competition from AMD.

I heard AMD wished to take another path and do all-in-one systems.

 

[JethroXP]

This seems like another nail in AMD's coffin as far as I'm concerned... I bet Intel will take the lead and not getting any competition from AMD.

I heard AMD wished to take another path and do all-in-one systems.

That's too bad, really. I mean, I've never been a huge AMD fan, only two the *many* PCs I've built in the past 15 years have used AMD CPUs, however I seriously doubt we'd have the huge leap in performance we are seeing now with Core 2 Duo and Core 2 Quad if there hadn't been an AMD nipping at Intel's heels (and actually beating them performance-wise for a couple years too).

 

[pgwalsh]

I'm sure AMD will figure things out, they have in the past.

Being a bit lazy, does anyone know if these will be desktop, server or laptop chips and any other details?

 

[n1ce_hat]

Core 3!

 

[MH Knights]

What socket will these be? I am going to do a full system upgrade in the late fall and will likely go with Intel/nVidia. I see the C2D are LGA775 but will these Penryn CPUs be on something else?

 

[CptFalcon]

I dont think so. At least Intel isn't switching sockets every 12-18months like one company I know!

 

[stonedwaldo420]

Sure am glad I sold my AMD stock almost a year ago! I have to say that it was the best return on any stock I have ever owned.

However, I am still torn by this news. On the one hand, I am happy that Intel is doing a great job of turning out high performance, overclocker friendly parts at a reasonable price. On the other hand, I don’t want to see AMD fail.

I remember when I first got into overclocking, everyone was into the Celeron. After all, most of us enthusiasts couldn’t afford one of the ultra-expensive Pentium chips. This truly is what Intel (or any other company for that matter) will do if they have no competition.

While many other companies and market forces played a part, I think competition offered by AMD is one of the major reasons we have seen such a dramatic price drop in “midrange” CPUs. While I don’t think Intel could ever get away with charging what it once did, I do see the rate of progress slowing down quite a bit if AMD is taken out of the picture.

Unfortunately, I am not sure how AMD stands a chance against Intel. Both companies have a lot of really smart chip engineers, and it seems pretty clear that Intel is a full generation ahead of AMD in terms of manufacturing. The only reason they managed to “win the battle” for a few years was that Intel listened to their marketing guys and not their engineers. They figured that Ghz would sell better than actual performance and therefore wasted YEARS on the netburst architecture.
While the Core 2 duo performance is impressive, imagine what it could be if Intel had not been distracted by netburst? I have a feeling that AMD would not have stood a chance and we would probably still all be paying more for our processors.

Best of luck to AMD… I am rooting for you guys!

 

[[BRO]Alaskan]

Not due till 2Q 2008. That a year and a half from now. AMD might lag for a while but I think as more cores are added AMD's internal memory controller will help even up the odds (for the quad core and even pull ahead for the 8core). I don't want to see AMD go away and leave Intel with no competitor because it will allow them to set any price and slow advancements down to rake in more cash.

 

[rayman2k2]

uh oh, Intel is doomed, they have guy name Dan Snyder working there...



*cough* Redskin's owner *cough*



good for Intel though, can't wait to see how it stacks compared to Conroe

 

[mrbigshot]

Not due till 2Q 2008. That a year and a half from now. .

According to the director of Intel, the company is on-track to produce Penryn in volumes and ship them to customers in the second half of 2007. (summer/fall of this year)

 

[mrbigshot]

does anyone know if these will be desktop, server or laptop chips and any other details?

No official details concerning the Penryn chip design were announced this time, however, according to previously published news-stories, the chip code-named Penryn is a 45nm incarnation of the dual-core Intel Core 2 Duo processor for mobile computers (code-named Merom) with SSE4 technology that, Intel promises, is set to benefit a variety of applications, including those involving graphics, video encoding and processing, 3D imaging, gaming, web servers and application servers. It is unclear what kind of advantages besides the SSE4 the new chips will have over the current Intel Core 2 Duo product family. Typically, thinner manufacturing process may allow Intel to crank up clock-speed of the chips, enlarge its cache sizes, reduce power consumption and make them cheaper to make.

 

[nilepez]

Unfortunately, I am not sure how AMD stands a chance against Intel. Both companies have a lot of really smart chip engineers, and it seems pretty clear that Intel is a full generation ahead of AMD in terms of manufacturing. The only reason they managed to “win the battle” for a few years was that Intel listened to their marketing guys and not their engineers. They figured that Ghz would sell better than actual performance and therefore wasted YEARS on the netburst architecture.

The key is we don't know where AMD is. As I recall, we didn't really know what Conroe was going to do until Q2 of last year. I'd also swear that AMD was working on 45nm now and intends to make that transition next year (but I may be mixing up the 2 companies).

At this point, Intel is king....a year or 2 from now, who knows.

I agree 100% with whomever took a jab at AMD for chaning sockets so much. If they hadn't killed 939, I'd probaby just buy another Amd chip...but now I just want to sell my ram and move on (mostly because the ram has a fair amount of value )

 

[CptFalcon]

The key is we don't know where AMD is. As I recall, we didn't really know what Conroe was going to do until Q2 of last year. I'd also swear that AMD was working on 45nm now and intends to make that transition next year (but I may be mixing up the 2 companies).

At this point, Intel is king....a year or 2 from now, who knows.

I agree 100% with whomever took a jab at AMD for chaning sockets so much. If they hadn't killed 939, I'd probaby just buy another Amd chip...but now I just want to sell my ram and move on (mostly because the ram has a fair amount of value )

That would be me dear fellow!

 

[stonedwaldo420]

What makes me suspicious is how few benchmarks we have gotten from AMD. I remember this time last year there were a lot of rumors floating around about amazing Conroe performance. Everyone was skeptical because the benchmarks were carried out on Intel’s own machines, and no one was given much time with them. However, Intel was confident enough to put those number out there long before the processor hit the market.

So far I haven’t seen any secret benchmarks or rumors of amazing performance from AMD’s next chip. I would imagine that if they had an ace up their sleeve, they would “unintentionally leak” some favorable performance numbers. At the very least they would serve to give investors hope of some sort of comeback.

Finally, killing the 939 was a huge mistake. I have one and managed to get a 4400 this Christmas while they were still available on Newegg. I can understand that it is important to Dell that AMD has a platform that can use the same DDR2 memory as the Intel platforms, but AMD should have continued with 939 sales to the enthusiast market.

 

[nobody_here]

but AMD should have continued with 939 sales to the enthusiast market.

why?, no one would be buying them now, at least not enthusiasts......

 

[KingThot]

Official release:
http://www.intel.com/pressroom/archive/releases/20060125comp.htm

 

[KingThot]

Snap, NM that thing I just posted is OLD

 

[Cannydog]

I personally am interested in a 45nm Quad Core. And from at least one rumor, I can remember, it would require a new socket. Intel does like to make money off chipset sales after all. So I'd guess that an "upgrade" is not far down the pipe. And that would follow Intel's normal pattern of "business".

 

[demonx]

Man, I was planning on upgrading to a whole new system here soon. Then I found out that PCI-E 2.0 is coming soon and now this. Looks like I'll have to force myself to wait some more.

 

[Ockie]

Man, I was planning on upgrading to a whole new system here soon. Then I found out that PCI-E 2.0 is coming soon and now this. Looks like I'll have to force myself to wait some more.

If you wait you will never stop waiting.

 

[Ockie]

I personally am interested in a 45nm Quad Core. And from at least one rumor, I can remember, it would require a new socket. Intel does like to make money off chipset sales after all. So I'd guess that an "upgrade" is not far down the pipe. And that would follow Intel's normal pattern of "business".

Please tell me how long LGA775 has been on the market vs it's compeditors....


thank you.

 

[FrgMstr]

Man, I was planning on upgrading to a whole new system here soon. Then I found out that PCI-E 2.0 is coming soon and now this. Looks like I'll have to force myself to wait some more.

Hehe, you will be waiting a long time and then when you get it there will be nothing on the desktop to take advange of the bus' bandwidth...pretty much exactly where we are at with PCIE 1.

 

[ferrisnox]

It’s amazing how quickly the shoes have changed feet. I remember a couple months ago when AMD was killing Intel with dual core processors etc... and now things have changed. I don't think this is the demise of AMD!

What I think we are seeing is going to be a new cyclic leap-frog of microprocessor technology. Initially AMD was ahead with the research and first to tap out its new design. With superior technology they go into full scale production. That just gave Intel time to think and design the next wave, which it did and it went into full scale production and is currently superior. Well with the money and research involved in succeeding generations, you don't just stop production in the middle of the run; you keep running while your engineers come up with the next design. So it’s just Intel’s turn in the spotlight.

Now if they stay in that spotlight for more than six months... then its time to talk words of demise. But it is definitely premature at this point.
My .02 cents

 

[Cannydog]

Please tell me how long LGA775 has been on the market vs it's compeditors....


thank you.

LGA775 has been available for roughly 18 months vs socket 939 (now discontinued) having been available for 30 months. However, that comparison is irrelevant. I wasn't trying to compare Intel to AMD. I was saying that in the last 10 years I've been building systems they've changed their CPU interface 7 times (or more if you count all the server variations). Plus they have two more variations on the horizon in the Socket B (1366 contacts) and the Socket H (a 715 contact replacement for the Socket T).

Now I'm not certain as to when these new variants will be released. I was commenting on the fact that Intel produces mainstream chipsets, as well as processors. They can change things up whenever they want. This is part of what Intel might refer to as "Progress". The affect that has on some consumers continues to amuse me.

 

[nilepez]

LGA775 has been available for roughly 18 months vs socket 939 (now discontinued) having been available for 30 months.

I think you're overstating the 939. As I recall it came out late 2004, and with the possible exception of one or 2 boards, wasn't really out until January of 2005.

They announced it was dead last spring (or early summer) and quit making 939 CPUs, as I recall, last summer. At best, I'd say it had a 24 month run and given that they announced it's death last spring or early summer, I could argue that if the first MB hit the streets in december of 04 (if it was november, add a month) its run was maybe 18-20 months (I'm not sure exactly when the last new processor for the 939 came out nor exactly when they quit makingA64 CPUs, but I thought it was this past summer).

 

[Cannydog]

I think you're overstating the 939. As I recall it came out late 2004, and with the possible exception of one or 2 boards, wasn't really out until January of 2005.

They announced it was dead last spring (or early summer) and quit making 939 CPUs, as I recall, last summer. At best, I'd say it had a 24 month run and given that they announced it's death last spring or early summer, I could argue that if the first MB hit the streets in december of 04 (if it was november, add a month) its run was maybe 18-20 months (I'm not sure exactly when the last new processor for the 939 came out nor exactly when they quit makingA64 CPUs, but I thought it was this past summer).

First, the 939 pin CPU was still in production until December of 2006. I'm certain of this fact. Otherwise I did not rely just on my memory. I looked up "CPU Socket" on wikipedia. It gave me separate articles on the sockets in question. So I'm NOT overstating anything. In fact, if you check out the socket 478 you'll find it lasted longer than the 939. (Perhaps because it followed the socket 423 which didn't even last a full year.)

But the point I was trying to make is that Intel has a history of changing CPU interfaces frequently. And they have indicated that by the time 45nm production starts they will switch to a new socket. So don't automatically expect to be able to upgrade to a 45nm just because you currently have a socket T motherboard that you really like.

 

[nilepez]

First, the 939 pin CPU was still in production until December of 2006. I'm certain of this fact. Otherwise I did not rely just on my memory. I looked up "CPU Socket" on wikipedia. It gave me separate articles on the sockets in question. So I'm NOT overstating anything. In fact, if you check out the socket 478 you'll find it lasted longer than the 939. (Perhaps because it followed the socket 423 which didn't even last a full year.)

But the point I was trying to make is that Intel has a history of changing CPU interfaces frequently. And they have indicated that by the time 45nm production starts they will switch to a new socket. So don't automatically expect to be able to upgrade to a 45nm just because you currently have a socket T motherboard that you really like.

It took a lot of searching, but you're right and I was wrong. I was confusing the last time to order CPUs with last ship date (and there's at least one A64 that's still shipping, which could be ordered trhough Q4).

I still think the socket was basically dead once they announced they weren't going to make new CPUs for it after this year (with the exception of teh 4800+)....and I still think it was premature.....but that doesn't make me any more right about when they quit making processors.

 

[Duke3d87]

since i apparently cant edit my post, ill make another.

These wont be out till Q2 of 08?

Q2 '08 sounds closer to the Nehalem launch

http://www.xbitlabs.com/news/cpu/display/20060428162855.html

If you look, Nehalem is supposed to appear sometime in 2008 and Penryn in 2007.

Or if you follow Intel's tick-tock way of releasing products, it's 1 year for ever new product (every year a new Intel core comes out and every two years, a CPU is made with a new process).

So a year from the Core 2 launch would place it anywhere between June and August. But I expect XS to have benchmarks far before then being that A0 boots several oses.