backtop


Print 91 comment(s) - last by Targon.. on Aug 24 at 6:58 AM


Intel's "Tick-Tock" long-term roadmap.  (Source: HKEPC)

Intel "Penryn" plans  (Source: HKEPC)
There's more to Intel's next-generation processor family than the 45nm process node

Intel plans to unleash its Penryn family processors next quarter, shortly after AMD releases Barcelona. Penryn is the umbrella for all 45nm Core 2 micro architecture products, including quad-core Xeon Harpertown, quad-core Core 2 Yorkfield and dual-core Xeon, Core 2 Wolfdale processors.

On the surface, Penryn looks like die shrink of last year’s Conroe micro architecture, but Intel sought additional tweaks to the micro architecture to achieve greater performance at the same clock speeds as Conroe processors.

Intel improves existing Wide Dynamic Execution, Advanced Smart Cache, Advanced Digital Media Boost and Intelligent Power Capability, technologies that previously made its debut with Conroe and Merom.

Penryn enhances Wide Dynamic Execution technology with a fast radix-16 divider and improved Virtualization technology. With a fast radix-16 divider, the processor can process 4-bits per cycle instead of the 2-bits per cycle of Conroe – doubling the divide instruction capabilities. Intel VT technology receives enhancements that reduce virtual machine transition latencies by 25-to-75%.

Intel Advanced Smart Cache technology receives additional enhancements, besides the increased L2 cache. Penryn-based quad and dual-core processors will have up to 12MB and 6MB L2 cache, respectively. Intel reduces cache latency in addition to the larger sizes. Penryn features a 24-way associative cache, an upgrade from Conroe’s 16-way associative cache.

New to the Advanced Digital Media Boost technology is the inclusion of a new Intel SSE4 instruction set. SSE4 introduces 47 new instructions to improve performance of video accelerators, graphics building blocks and streaming load. Intel claims a 2x performance gain in video acceleration tasks. There are 14 new instructions for video accelerator performance enhancement. Intel improves compiler auto-vectorization performance with 32 new instructions.

Intel expects SSE4 optimizations to deliver performance improvements in video authoring, imagine, graphics, video search, off-chip accelerators, gaming and physics applications. Also new to Advanced Digital Media Boost is the Super Shuffle Engine. Intel’s Super Shuffle Engine allows for shuffling unpacking, packing, align concatenated sources, wide shifts, insertion and extraction, and setup for horizontal arithmetic functions. Intel claims a “2x faster SSE shuffle instruction execution,” according to briefing documents.

Mobile Penryn processors receive enhanced power saving technologies. New to the mobile Penryn is a deep power down state. In the deep power down state, the processor lowers the core voltage, more so than in the C4 state, and turns off the L1 and L2 caches. Intel claims significant power savings in idle modes for extended battery life with the new power state.

For servers and workstations, Intel has designed Harpertown with additional headroom for the front-side bus. Intel plans to debut Penryn-based Xeon DP, MP and UP processors with 1333 MHz front-side bus, but the architecture has headroom for up to 1600 MHz front-side bus. Penryn-based quad and dual-core Xeons will have three thermal envelopes. Quad-core Harpertown Xeons will have 50, 80 and 120-watt TDP ratings while dual-core Wolfdale Xeons have 40, 65 and 80-watt ratings.

Desktop Core 2 and Core 2 Extreme processors have less thermal bins. Quad-core Yorkfield models have 95 and 130-watt TDPs while dual-core Wolfdale models have a single 65-watt TDP. Intel hasn’t set thermal ratings for its mobile Penryn processors yet.

Penryn follows Intel’s plans of alternating between new fabrication processes and a completely new core within two years. Intel previously released its 65nm fabrication process with the last of its Netburst Pentium D and Xeon DP processors, with Conroe, a new architecture, following months after. Once again, Penryn is a die shrink of Conroe set to debut Intel’s 45nm fabrication process.

Following Penryn is a new Nehalem architecture, based on 45nm with Intel’s new QuickPath technology, formerly known as common-system interface. Nehalem is set for a 2008 introduction, beginning with the Xeon family for servers and workstations. Intel plans to shrink Nehalem to 32nm with the Westmere core. Following Westmere is Sandy Bridge, a new micro architecture based on 32nm.

Expect Intel to debut Penryn later this year with the Xeon family.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Shock!!! Intel Improvements
By DannyH246 on 8/6/2007 6:12:40 AM , Rating: 2
That great innovator Intel has announced improvements to its next chip...

More cache, faster cache, better cache, SSE4.

They have really broken the mould this time!! This is so unlike them to provide such major innovation. We love you Intel!!




RE: Shock!!! Intel Improvements
By DeepThought86 on 8/6/2007 6:28:45 AM , Rating: 1
Too much coffee??


RE: Shock!!! Intel Improvements
By LogicallyGenius on 8/6/07, Rating: 0
RE: Shock!!! Intel Improvements
By 2ManyOptions on 8/6/07, Rating: 0
RE: Shock!!! Intel Improvements
By Targon on 8/24/2007 5:39:11 AM , Rating: 2
Is DDR3 performance really that much faster than DDR2 at this point? How expensive is DDR3 memory at this point? In a server, do you really want to trust a new memory technology?

AMD will release a DDR3 capable processor when it makes sense(aka, when the market moves to DDR3, not when everyone still uses DDR2).

In addition to this, when AMD makes the move to DDR3 support, the processors will also support DDR2 memory, which means they can work in existing motherboards. With Intel, you will be forced to buy a new motherboard for ANY new CPU core design, even if it uses the same socket.


RE: Shock!!! Intel Improvements
By leexgx on 8/6/2007 5:01:33 PM , Rating: 2
why cant i rate users any more ? (i know i posted now) but i cant rate any one any more as i could before


RE: Shock!!! Intel Improvements
By emboss on 8/6/2007 9:30:26 PM , Rating: 3
Yeah, I can't rate posts any more either. I've also noticed that the "Oops!" errors when posting stopped at about the same time, so I'm thinking a bug introduced in a code update?


RE: Shock!!! Intel Improvements
By bobdelt on 8/6/2007 6:47:12 AM , Rating: 2
Couldn't have said it better myself. Intel is a great innovator; they do have the best processors on the market.


RE: Shock!!! Intel Improvements
By Gherking on 8/6/2007 7:56:01 AM , Rating: 4
I'm a little worried... back in the Pentium 1-3 days, Intel had their cake and were eating it with high prices.

I feel like AMD's increased presence was the reason for the shining development of Core 2.

If AMD stay competitive with Barcelona, it can only be good news for us. We are at a good place in the delicate CPU market/economy.


RE: Shock!!! Intel Improvements
By retrospooty on 8/6/2007 10:37:02 AM , Rating: 1
"I feel like AMD's increased presence was the reason for the shining development of Core 2."

Absolutely... If not for AMD, Intel would still be pushing the P4 netburst on us. We would probably be paying $1000 for a P4 single core P4@ 3.0 ghz at this point.


RE: Shock!!! Intel Improvements
By zsdersw on 8/6/2007 11:17:34 AM , Rating: 5
If not for Intel, AMD would still be pushing 90nm A64's at insane prices, too.


By theapparition on 8/6/2007 11:22:50 AM , Rating: 4
People also forget that without innovation, there is no reason to upgrade, so demand goes down. When demand is down, prices go down. Simple economics.

So as much as everyone would like to think of a single processor company charging 10K per processor, that's not going to happen.


By Oregonian2 on 8/6/2007 1:24:47 PM , Rating: 2
Note also that Intel changed "administrations" a few years ago and this may have had an influence in Intel's stronger motivation as well. New administration started with layoffs I recall...


RE: Shock!!! Intel Improvements
By PaxtonFettel on 8/6/2007 6:48:03 AM , Rating: 2
Erm, this refresh was never touted as a revolution, silly. It is simply a die shrink with a few improvements, so of course it's going to be more of an evolutionary step. The next really interesting thing from Intel, I'm sure in most peoples books, will be Nehalem and QuickPath. It will be good to see the ancient FSB finally ditched for something that makes more sense in terms of modern computing, also seeing how it stands up to hypertransport should be interesting too.


RE: Shock!!! Intel Improvements
By FITCamaro on 8/6/07, Rating: -1
RE: Shock!!! Intel Improvements
By PaxtonFettel on 8/6/2007 7:11:05 AM , Rating: 1
I meant 'from Intel' as in Intels take on an IMC. I am aware that AMD were pretty much the pioneers in terms of IMC use, which is why I said it will be interesting to see how Intels version stands up to AMDs well established system.


RE: Shock!!! Intel Improvements
By Master Kenobi (blog) on 8/6/2007 10:08:05 AM , Rating: 1
I seem to recall a different company coming up with an integrated memory controller before AMD, I keep thinking IBM but I might be wrong. Can't find much information about IMC.


RE: Shock!!! Intel Improvements
By PaxtonFettel on 8/6/2007 10:48:32 AM , Rating: 2
I wouldn't be at all surprised if someone else had come up with it before AMD. Anyone with a little foresight could have seen the impending limitations of FSB even before AMDs IMC came around. However, they were still the first to bring it to the mass market.


RE: Shock!!! Intel Improvements
By theapparition on 8/6/2007 11:20:47 AM , Rating: 5
Intel was the first x86 compatible chip with IMC, introduced in early 90's. It never really caught on, I guess they were too ahead of their time. So for all who think AMD was the pioneer, your dead wrong.

It's time to get off this IMC vs. FSB rant on both fanboy sides. AMD chose that direction to compete with Intel during a major redesign of their chip. Everyone has short memeories, it was only until the last days of prescott did AMD then have a convincing performance advantage.
Intel, however, didn't have the need at the time for a IMC, and thier business strategy was such that they didn't want it. Intel has a much bigger market than AMD, and more diversified. A separate MC, in the northbridge, provided platform flexibility, something AMD has been clamoring for. AMD was also not nearly the chipset manufacturer as intel was. It's not that intel was stupid. It was purely business. Last I checked, Intel was making money, AMD was not.

The FSB allowed intel to get dual cores faster (yes, pieced together rather than native). It also allowed them to have higher yeilds, and price chips cheaper. There is no question that AMD's approach is more future proof and superior, however, even with the superior theoretical advantage, the intel's still managed to be competitive, and over the last year, vastly superior to anything AMD has offered.

Under server applications, AMD's have been shown to have better capabilities, but for workstations and desktops/laptops, Cores are king, despite "inferior" architecture. So if you run a server farm, get Opteron, otherwise, your best served by Core. When AMD actally releases official benchmarks of their Stars processors, that may change.

Skip the stupid rants on which is better, we all know its Intel right now. Skip the arguments over innovation, both companies have contributed to the processors we have now.


RE: Shock!!! Intel Improvements
By PaxtonFettel on 8/6/2007 3:15:55 PM , Rating: 2
Thanks for clearing that up, I had no idea that Intel had tried previously with an IMC chip, interesting stuff and it says a lot about the importance of timing in hardware releases.


RE: Shock!!! Intel Improvements
By Chillin1248 on 8/6/2007 5:17:44 PM , Rating: 2
The Intel 386SL microprocessor integrates a fully static processor, memory controller , ISA bus controller, EMS 4.0 hardware and system control circuitry for battery-powered systems. It offers three times the integration of Intel 386SX CPUs.

It was released in the early 1990s.

-------
Chillin


RE: Shock!!! Intel Improvements
By emboss on 8/6/2007 9:38:08 PM , Rating: 2
The 186 (early 80's) was much the same idea - though whether you could say that an 8086-class system had a "memory controller" at all is of course a debatable topic :)


RE: Shock!!! Intel Improvements
By nitin213 on 8/7/2007 5:41:38 AM , Rating: 2
Just remembering from my microarch 101 lessons..... wasnt the bus architecture developed sometime in late-80's for the sun systems (by Andy Bechtolshiem) and that this bus architecture was very successful given its versatility......

Then this would mean that pretty much everything else till then was using IMCs (most likely for SRAM though as DRAM was quite expensive and given the variable latency an awkward design for the DRAM memory controller)


RE: Shock!!! Intel Improvements
By paydirt on 8/7/2007 8:44:03 AM , Rating: 2
The concept of Itanium was way before it's time and part of the reason it flopped is because the developers/programmers did not adopt it.


RE: Shock!!! Intel Improvements
By Targon on 8/24/2007 6:14:55 AM , Rating: 2
No, Itanium flopped because support for x86 applications was horrible and Intel didn't push for UNIX or Linux as the primary operating system. Even now, Itanium holds virtually no advantages, which is why companies continue to avoid it.

The Pentium Pro was also a failure, but Intel managed to salvage some of the work done for their Pentium 2, 3, and even Core. At this point though, Itanium has been an almost complete failure.


By CrimsonFrost on 8/7/2007 1:01:02 PM , Rating: 2
Oh my gosh... ISA boards, that brings back memories. I just recently upgraded an elderly couples computer to connect to the internet, and the computer used ISA interfaces, I had to search all over the internet to find a modem for that thing!


RE: Shock!!! Intel Improvements
By Doormat on 8/6/2007 11:09:50 AM , Rating: 3
DEC Alpha EV7.


RE: Shock!!! Intel Improvements
By Targon on 8/24/2007 6:17:24 AM , Rating: 2
I thought it was the EV6 bus, which was used back in the Athlon days, and I am not sure if it is still being used by current AMD processors.


RE: Shock!!! Intel Improvements
By Sazar on 8/6/2007 5:01:00 PM , Rating: 2
I think by IBM you mean Intel :)

Intel had a couple of different processors configs with a lot of R&D work done and the product essentially complete with an IMC but they were not brought to market due to various reasons.

AMD is hardly a pioneer in that sense, but they were the first company to bring it to market in the mainstream.


RE: Shock!!! Intel Improvements
By Captain Orgazmo on 8/6/2007 6:26:54 PM , Rating: 2
Intel processors prior to Pentium (486, 386, 286, and so on) had an integrated memory controller. I have an old PC magazine from 1993 here outlining the new (at that time) architecture of the Pentium processor. Separating the memory controller allowed higher, asynchronous clock speeds for the processor and double the bits per cycle. Memory bandwidth was not a bottleneck at the time as it has became recently. Essentially, dumping the memory controller off-chip allowed Intel to focus more on CPU development and let others worry about memory end of things. It was a near-sighted strategy, but it worked at the time, because the step from 486 to Pentium was among the largest single steps up in CPU performance in recent computer history.


RE: Shock!!! Intel Improvements
By sdsdv10 on 8/6/2007 8:25:39 PM , Rating: 3
quote:
...but it worked at the time...


Last time I checked, it was still working now. ;-)
C2D > X2


RE: Shock!!! Intel Improvements
By Oregonian2 on 8/6/2007 1:17:36 PM , Rating: 2
Integrated memory controllers have been around forever for embedded processors where the "pulling in" auxiliary everything is emphasized for cost reduction.


RE: Shock!!! Intel Improvements
By twajetmech on 8/7/2007 12:31:12 AM , Rating: 1
Lets also not forget that until the late 80's AMD was nothing more than a fab and memory maker Who then reversed engineered Intel's then current cpu and then sued to have the right to produce it without license, no inovation there. They then were sued (successfully) by Intel for copyiing their MMX technology (again more non inovation) They also began to copy Cyrix cpu designs for RISC processors until that market went nowhere (some inovation huh !) So, with all that inovation, and the lack for any real R&D (at the time) AMD was able to underprice Intel and produce copies of Intel cpu's for less money. I must find it humerous that AMD now wants to sue Intel more making deals with large mfgs on bulk price rates for cpu's. How short people's memories are. AMD did not invent the IMC


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/7/2007 4:49:45 AM , Rating: 2
Lets also not forget that until the late 80's AMD was nothing more than a fab and memory maker Who then reversed engineered Intel's then current cpu and then sued to have the right to produce it without license, no inovation there.

Wrong, they made a clone of the 8080 with reverse engineering but they were licensed by Intel in 1982 to produce chips for them. This lasted until 1986 when Intel broke their agreement with AMD. Amd won and was awarded damages for breach of contract. Intel then sued AMD for copyright infringement in 1990, which they again lost.

They then were sued (successfully) by Intel for copyiing their MMX technology

Wrong again, Intel sued both AMD and Cyrix for misuse of the term MMX. Intel and AMD settled with AMD acknowledging MMX as a trademark of Intel and Intel granting AMD the rights to produce their K6 MMX processor.

They also began to copy Cyrix cpu designs for RISC processors until that market went nowhere

Wrong yet again. RISC processor was IBM's design not Cyrix's. Both Cyrix and AMD produced RISC chippery based on IBM's design. Oh and if 'that market went nowhere' you can put down your PSP, Gameboy, and DS as they all incorporate RISC designs. And while your at it, box up your Playstations, Xboxs, and Nintendo consoles as they are also platforms derived from RISC designs. Oh, and I'd hate to miss mentioning that fastest SPECint and SPECpf processor is a RISC - IBM's Power6.

So, with all that inovation, and the lack for any real R&D (at the time) AMD was able to underprice Intel and produce copies of Intel cpu's for less money.

Until 1994's court decision concerning the breach of contract, AMD was a fab for Intel. Immediately following the decision in 1996 AMD released its K5 processor which was a completely in-house design. From that point on they were an independent producer of cpus.

I must find it humerous that AMD now wants to sue Intel more making deals with large mfgs on bulk price rates for cpu's. How short people's memories are.

Including your memory, it seems like you've totally forgotten this article has nothing to do with the lawsuit.

AMD did not invent the IMC

Well at least you got one thing right out of that whole paragraph of BS.


RE: Shock!!! Intel Improvements
By Makaveli on 8/7/2007 12:30:13 PM , Rating: 1
Wow Cyborg you totally owned that guys post. Thanks for putting up the correction information. Made him look like a 8 year old girl!

=P


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/7/2007 6:56:34 PM , Rating: 2
While I was a bit young to mess with the 8008/8080, by the time the 8088 came out I was messing with computers ( can't call them 'PCs' back then ) off and on. So I was following all this tech news back when 70% of the people reading this were still learning their ABCs.

.... man, I'm getting old :(


RE: Shock!!! Intel Improvements
By odiHnaD on 8/7/2007 2:18:16 PM , Rating: 2
Cyborg, I agree with you except for one point:

The Xbox (original) was actually x86 based (watered down PIII)

The 360 however is RISC (PowerPC) based.


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/7/2007 6:44:58 PM , Rating: 2
Correct, a 733 MHz Pent 3 - should have been more specific with that.


RE: Shock!!! Intel Improvements
By Ren01 on 8/8/2007 6:13:23 PM , Rating: 2
Actually it was a mobile Celeron in a Micro-PGA2 package. You couldn't really call it a pIII with only 128k cache because it's cache was only 8-way associative. I have a tualtin based Celeron with 256k cache but again it is only 8 way associative.


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/9/2007 2:24:24 AM , Rating: 2
It can go either way, it was based on the Coppermine core of both the Pent 3 and the mobile Celeron. The reason I refer to it as a Pent 3 is it has a 133 HHz FSB while the Celeron was capped at 100 MHz. But it did also have the lowered cache like the Celeron. Everything else is identical between the two chips.


RE: Shock!!! Intel Improvements
By Ren01 on 8/9/2007 4:53:59 AM , Rating: 2
Actually the mobile celeron series was offered at 133 FSB. It is by no way a pIII. The 128 cache and 8 way associativity is really what defined the difference between a mobile celeron and a mobile pIII of the era. The xbox CPU was a mobile Celeron.


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/9/2007 6:38:56 AM , Rating: 2
No Celeron was created at 133 MHz during the P3 days. You'd be hard pressed to find a mobile at even 100 MHz during the Coppermine range.

http://www.cpu-world.com/info/Intel/Celeron-vs-Pen...

But were are cutting hairs here, the Xbox core was a mix of both cpus based on the same initial design.


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/9/2007 6:46:00 AM , Rating: 2
Also the Xbox cpu has the 8-way associative cache that the Pent 3 has.


RE: Shock!!! Intel Improvements
By CyborgTMT on 8/9/2007 6:50:45 AM , Rating: 2
Man 3 posts in a row... really wish you could edit.

Anyway wanted to post documentation for my findings, and I'll use Anand himself for this:

The CPU that powers the Xbox is a Coppermine based Pentium III with only 128KB L2 cache. While this would make many think that the processor is indeed a Celeron, one of the key performance factors of the Pentium III that is lost in the Celeron core was left intact for this core. The Coppermine core was left with an 8-way set associative L2 cache instead of the 4-way set associative cache of the Celeron. Based on what we've seen with the Coppermine and Coppermine128 (Celeron) cores we estimate that the 8-way set associative L2 cache gives this particular core a 10% performance advantage over the Coppermine128 core of the Celeron.



http://www.anandtech.com/showdoc.aspx?i=1561&p=1


RE: Shock!!! Intel Improvements
By Ren01 on 8/9/2007 3:53:52 PM , Rating: 2
http://en.wikipedia.org/wiki/Xbox

I agree though, we are splitting hairs. In the end it was better than a desktop Celeron, but worse off than a desktop pIII.


RE: Shock!!! Intel Improvements
By Oregonian2 on 8/13/2007 5:17:25 PM , Rating: 2
quote:
Wrong, they made a clone of the 8080 with reverse engineering but they were licensed by Intel in 1982 to produce chips for them. This lasted until 1986 when Intel broke their agreement with AMD. Amd won and was awarded damages for breach of contract. Intel then sued AMD for copyright infringement in 1990, which they again lost.


Yes, AMD successfully defended the point that they had a license to Intel's patents as part of their manufacturing stuff for Intel.

quote:
Wrong yet again. RISC processor was IBM's design not Cyrix's. Both Cyrix and AMD produced RISC chippery based on IBM's design. Oh and if 'that market went nowhere' you can put down your PSP, Gameboy, and DS as they all incorporate RISC designs.


I don't recall the details of Cyrix and AMD RISCs (don't even recall them making any to tell the truth), but just thought I'd point out to those who don't know that "RISC" wasn't an alternative to x86, it was an alternative to "CISC" which in addition to the x86 also included Motorola's 68K, and most all 8-bit processors. RISC was the brainchild of a professor from Stanford if I remember right (his name is on the tip of my tongue....). It just means the architecture is simpler with a Reduced Instruction Set (RIS part of RISC). Idea was that a simple architecture could be made to run at higher clock rates due to the simplicity and be done quicker (didn't take into account that a CISC company could have "infinite" resources, as it did). Funny thing is that RISC processors that were actually made had huge instruction sets anyway. :-) The Power Processor from IBM was just one in the RISC category of processors (as is ARM, MIPS, and others) all of which are different.

One of things not mentioned is a company that AMD picked up that had been designing a super-fast new architecture x86, but had been doing it f-o-r-e-v-e-r, but never came out with anything. AMD bought them and eventually put some of it into their parts at some point. Driving me crazy that I can't think of their name... sigh.

P.S. - AMD did have some very innovative stuff, they just didn't catch on. They once made SRAM memory chips that had a "valid" bit that came out. This allowed for asynchronous designs that were very fast and could run as fast as "typical" or however fast chips one had rather than running as slow as the "worst case" slow ones ran. I liked it (I liked to design things on the bleeding edge) but I must have been the only one. :-) Problem was that second sources left that pin out, and this made it hard to put into a design.


RE: Shock!!! Intel Improvements
By Targon on 8/24/2007 6:39:09 AM , Rating: 2
Deals based on volume was never what the lawsuit was about. The lawsuit is based on Intel using extortion tactics to force companies not to sell products for AMD processors.

No company is allowed to pay money to reduce or limit the sale of a competitor's products. Intel could legally provide discounts based on volume, but not for limiting sales that support a competitor. If Intel were to pay a company(either rebates or cash or in some other way) to keep the number or percentage of sales of AMD products down, that is a violation of the rules of business in almost every civilized nation out there.

The only time when a company in general may legally provide incentives to not sell products from other companies is when there is an exclusivity agreement, and even then, there are still laws about how you do it. Without monopoly laws in effect, Intel could come up with an agreement with a company where that company would only sell Intel processors or systems.

The problem then is that Intel should not be able to say that that company could not sell motherboards that SUPPORT AMD processors(meaning sales of Asus motherboards that support AMD processors should be fine). If Intel pushes via deals to keep motherboards that support AMD processors off the market, then that may be considered an unfair business practice.

All of this also ignores the rules for when a company is considered a monopoly. Now, AMD was not seen as a threat to Intel in any way until the AMD K6 was released. At that point, any Intel dealings to keep AMD out of the market would need to be looked at to make sure the rules for fair business were being followed. It wasn't until the original Athlon was released that AMD had enough of a reason to clearly see the unfair business practices that Intel was using. Asus releasing the K7M motherboard for example was sold in a plain white box because Asus was worried that Intel might take action against Asus for their support of the Athlon. Fear of having supplies of Intel chipsets reduced for "retaliation" is an indication that Intel business practices were illegal.


RE: Shock!!! Intel Improvements
By jazkat on 8/6/07, Rating: -1
RE: Shock!!! Intel Improvements
By Misty Dingos on 8/6/2007 9:52:39 AM , Rating: 2
Just a suggestion. Anger manangment class.

And a question. Why must you hate so?

Too many negative vibes for a Monday morning.

A little advice. Go hug a puppy.


RE: Shock!!! Intel Improvements
By Ringold on 8/6/2007 8:30:40 PM , Rating: 2
No! Might crush the puppy.

Hug a kitten, man. (I hate cats)


RE: Shock!!! Intel Improvements
By Oregonian2 on 8/13/2007 6:49:26 PM , Rating: 2
Hey man, no kittens!

Let him hug a full grown cat.

It'll take care of him. :-)


RE: Shock!!! Intel Improvements
By Moishe on 8/6/2007 9:58:36 AM , Rating: 6
Most people don't care about brands. We care about performance. When I built my last computer CD2 faster than the Athlon X2 equivalent and could over clock higher.

I think I paid around $45 extra for the CPU and maybe $20 for the mobo... So I made a decision to pay a little extra and get a better performing system.

It's not about Intel or AMD. It's about performance. Choice of manufacturers is not a moral question.


RE: Shock!!! Intel Improvements
By Moishe on 8/6/2007 10:47:55 AM , Rating: 2
Why the rate down? Is my statement not fact?

Most computer enthusiasts are not fanbois, they're performance junkies, or like me they're thrifty and want the best bang for the buck.


By lumbergeek on 8/6/2007 4:20:31 PM , Rating: 4
Don't know why the rate-down. You're exactly right. I build what makes sense for my customers. I currently, personally run a C2D. Previous to that I ran an Athlon. It's about the best performance per dollar. Maybe my next personal box will be a Phenom, maybe it won't. Time will tell.


RE: Shock!!! Intel Improvements
By P4blo on 8/7/07, Rating: 0
RE: Shock!!! Intel Improvements
By flipsu5 on 8/6/2007 10:45:18 AM , Rating: 3
tone down, skippy, save your excitement for Nehalem, where the change would be something more revolutionary than evolutionary.

It's a little surprising to add to Penryn so early. I suppose they want to separate from 65 nm C2D more cleanly.


RE: Shock!!! Intel Improvements
By vignyan on 8/9/2007 6:56:13 AM , Rating: 2
exactly.... Thats the moto of the tick-tock model... you do one big thing every year.. That way, you are safe about both the things... If they did both at a time, it will be like the prescott disaster...

Also its good for the fabs... they dont have to lie idle when there is a week or two push in the tape-out while moving to a new architecture and new process at the same time...


RE: Shock!!! Intel Improvements
By 3kliksphilip on 8/6/2007 12:44:49 PM , Rating: 2
At the moment they don't have to do much. AMD hasn't done much since the core 2 duos came along. If Intel did do something radical (ie, 16 cores) then they'd be pummeled into the ground by various companies suing them for unfair monopolistic stuff. AMD have enough on their plate as it is. Perhaps the CPUs are going to become more like GPUs, in that both companies introduce newer, slightly better models every year and a half or so. Evolution rather than revolution.

Recently Intel have moved into Quad core territory, I'd like to see them make them more efficient. That seems to be AMD's ACE at the moment, even incorporating the power required by CPUs into their new ranges.


RE: Shock!!! Intel Improvements
By aurareturn on 8/8/2007 6:02:17 AM , Rating: 2
quote:
Perhaps the CPUs are going to become more like GPUs, in that both companies introduce newer, slightly better models every year and a half or so. Evolution rather than revolution.

Until recently, the video market absolutely owns the CPU market.

New GPUs were coming out each year that were nearly twice as fast as the previous year. And the old GPUs dropped prices fast(sub-$200 for the last generation's highest end).

The CPU market moves far slower than the GPU market and teh GPU market will(hopefully) get back on track when AMD releases a new GPU architecture.


RE: Shock!!! Intel Improvements
By Gneisenau on 8/8/2007 8:57:40 AM , Rating: 2
I wouldn't mind that too much as long as CPU's don't start following GPU's thermal requirements. Two effecient heaters in one enclosure is more than I really want to deal with.


RE: Shock!!! Intel Improvements
By Targon on 8/24/2007 6:53:22 AM , Rating: 2
A big difference between CPU and GPU technologies is that a CPU is generally running in serial, while GPUs can run in parallel.

Rendering an image on the GPU means you will see improved performance by adding more pixel pipelines and allowing calculations on more and more objects at the same time.

In the CPU world though, applications tend to run things in order, and the ability to design an application where things can run in parallel seems to be limited, at least in the MS Windows world. Multi-threaded programming is starting to be a little more common due to the multi-core processors, but in general, the benefits of adding more processor cores on individual applications are fairly low.

If/when AMD or Intel release a good way to allow multiple processor cores to boost application performance, we will see this change. Until then, just throwing MORE cores into a CPU will have less of an effect than adding more pixel pipelines to a GPU.


By IntelUser2000 on 8/6/2007 10:04:39 PM , Rating: 2
quote:
That great innovator Intel has announced improvements to its next chip...

More cache, faster cache, better cache, SSE4.

They have really broken the mould this time!! This is so unlike them to provide such major innovation. We love you Intel!! quote>

FYI, this is in a long time that Intel made performance tweaks that would have otherwise been a simple process shrink. All the shrinks of Pentium, Pentium II, Pentium III, Pentium 4 were simple shrinks+cache. Prescott was very anticipated because of the architectural enhancements that would normally have been a process shrink, but that came out bad. In contrast, Penryn will offer performance increase what cache size increase won't be able to do it alone.


RE: Shock!!! Intel Improvements
By slickr on 8/7/2007 1:39:28 AM , Rating: 1
Yes Intel are the man. If it werent for Intel we would still be playing on those 120nm sockedA processors, unabble to extract a zipped file beacose of their limitations.

And even so what was wrong with Intel holding 90% desktop market share with P2 and P3? Those processors were ultra good. My P3 lasted me 4 years before statring to feel it's age.


Wolfdale @ 2.33 vs E6550 is looking real good
By EastCoast on 8/6/2007 8:27:37 AM , Rating: 5
It appears that HKPEC is getting hammered at the moment but if you are able to get the page to show up the benchmarks are conclusive vs Conroe.
http://www.hkepc.com/bbs/hwdb.php?tid=837360&tp=In...
^^benchmark results




By EastCoast on 8/6/2007 8:30:13 AM , Rating: 2
actually that's HKEPC


By togaman5000 on 8/6/2007 11:26:15 AM , Rating: 2
thanks for the link man, its a good conroe vs penryn comparison.

from the numbers i saw, the average improvement clock for clock for penryn is in the ballpark of 5%. Even more exciting, the gaming performance increase is more like 10%, with one benchmark showing a 30% increase! thats simply insane for a die shrink and a few minor improvements, but hey, i'm not complaining


By Oregonian2 on 8/6/2007 1:22:38 PM , Rating: 2
Yeah, one of them was over 100% improvement. I wonder if any of those applications are using the new added instruction set. If not, the improvements would be even larger when they do (mostly in the specialized media/video manipulation areas I think). HD editing needs all the help it can get.


By Continuation on 8/6/2007 1:34:05 PM , Rating: 2
According to the article only DivX 6.6 Alaph has SSE4 support, hence its 116% performance improvement. When other apps get around to support SSE4, hopefully we'll see similar improvement there as well.


By IntelUser2000 on 8/6/2007 10:16:19 PM , Rating: 2
quote:
thanks for the link man, its a good conroe vs penryn comparison.

from the numbers i saw, the average improvement clock for clock for penryn is in the ballpark of 5%. Even more exciting, the gaming performance increase is more like 10%, with one benchmark showing a 30% increase! thats simply insane for a die shrink and a few minor improvements, but hey, i'm not complaining.


I read that Half Life 2 uses shuffle instructions in the code and that's one of the major advancements in Penryn.

The overall performance increase is actually greater than 5%. IPC improvements are generally shown in office applications like Word. I'd say its probably around 7-8% there, and 9-10% when including games. Things like 3dmark is very synthetic so I don't know if that's a good comparison :P.


By MrTeal on 8/6/2007 12:21:31 PM , Rating: 2
I'd like to see those power consumption numbers confirmed somewhere with an article in English, but if they are accurate, that's an amazing drop in consumption. Hopefully they are as overclockable as everyone hope.


By SexyK on 8/6/2007 8:14:32 PM , Rating: 2
Dang! 59W at full load. Impressive. Can't wait to grab one of these babies.


Graph design
By v1001 on 8/6/2007 12:48:35 PM , Rating: 5
Intel is really rubbing it in with it's 3 giant phallus graphs lol.

You're winning we get it already ;)




RE: Graph design
By johnsonx on 8/6/2007 3:38:05 PM , Rating: 1
are you trying to say that instead of 'Tick' and 'Tock', it looks more like 'Dick' and 'Cock'?


RE: Graph design
By deeznuts on 8/6/07, Rating: -1
RE: Graph design
By johnsonx on 8/7/2007 5:32:37 PM , Rating: 2
by the way, are you the same deeznuts I see sometimes on the UT2k4 Onslaught servers?


Nehalem samples available October
By Continuation on 8/6/2007 2:09:42 PM , Rating: 2
According to this HKEPC article

http://www.hkepc.com/bbs/hwdb.php?tid=837360&tp=In...

Nehalem samples will be available in October.

The 2nd to last paragraph:

"Remarkably, Intel has already told motherboard suppliers that samples of next generation "Nehalem" processors will be available in October. Nehalem will be launched in the 2nd half of 2008. Penryn will only need to last 6 - 9 months before being re-positioned as a low-to-mid tier product."




RE: Nehalem samples available October
By deeznuts on 8/6/2007 3:17:04 PM , Rating: 3
No F'in Way dude. A year early that is crazy.


By redpriest_ on 8/8/2007 4:44:59 PM , Rating: 2
Sampling != close to production. Sampling generally means you're going to be waiting a year minimum to see it, sometimes even more.

Intel has to qualify IMC, QuickPath, and a handful of other technologies, which they're all trying to do at the same time for the first time. It isn't easy.


Agena
By Treckin on 8/6/2007 9:07:51 AM , Rating: 4
Man, Agena better really fucking rock, for AMD's sake. I have a nagging suspicion that these improvements may be just enough, coupled with the 45nm die-shrink, to close any performance gap left by Agena/Barcelona.
In the Quad Core segment though, I feel optimistic that the native-quad core design will still leave something wanting for Intel's 2+2 design. Even Intel stated in a release that they were astonished that AMD would attempt such an innovation in the same revision as an archeteture release, and were not envious of their task...

Again, Agena/Barcelona better kick ass, otherwise there is no pretty end in sight for DAAMIT (sorry for the lame Inq reference 8-0 <==3 )




RE: Agena
By PaxtonFettel on 8/6/2007 9:48:31 AM , Rating: 2
More to the point, if DAAMIT don't remain competitive it'll be us that gets screwed.


off the tick-tock
By flipsu5 on 8/6/2007 10:20:18 AM , Rating: 2
It looks to me the Penryn 45 nm tick wasn't enough by itself, so they snuck in a few semi-tocks?




RE: off the tick-tock
By coldpower27 on 8/6/2007 11:36:50 AM , Rating: 2
It's more geared toward saving costs then a architecture overhaul, but still does bring a range of improvements.


RE: off the tick-tock
By johnsonx on 8/6/2007 11:37:06 AM , Rating: 2
maybe those were clicks?


So...
By Engine of End on 8/6/2007 3:55:05 PM , Rating: 3
Westmere = Nehalem-C?
Sandy Bridge = Gesher?

Didn't know they changed the names already.




Impressive
By crystal clear on 8/6/2007 6:28:46 AM , Rating: 2
Appears Q4 07 will be very interesting indeed-

08/01/2007

According to documents seen by X-bit labs, Intel will release its new “Intel Core Extreme” processor with four processing engines made using 45nm process technology in Q4 2007. The document states that “launch of the Intel Core Extreme processor pulls into Q4’07,” emphasizing that the actual brand-name of the product is not yet determined. What is highly likely is that the top-of-the range extreme processor will be clocked at 3.33GHz, will use 1333MHz processor system bus and will have 12MB of level-two cache in total.

Contemporary mainboards based on Intel 3-series chipsets already support Intel’s 45nm central processing units, whereas some mainboard makers, namely Asustek Computer, have also enabled support for the new chips on previous-generation motherboards.

Initially there will be only about 2% - 3% of Intel’s desktop chips code-named Yorkfield made using 45nm process technology in the chipmaker’s product mix, however, the share of quad-core 45nm offerings will increase to 5% - 6% in Q2 2008, whereas the volume of 45nm products will get to little less than 30% of the desktop product mix in the second quarter next year.



http://www.xbitlabs.com/news/cpu/display/200708012...




All looks great but 2.33ghz?
By daveG on 8/9/2007 12:01:04 AM , Rating: 2
The new Penryn shows an undeniable advantage over an E6550 c2d at 2.33ghz. Not that this isn't a great accomplishment but does anyone know how it compares to the E6850? In my wish list, the E6550 wasn't in the running.

Even if it is a lot more efficient, the E6850 has a faster clock. Anyone got an idea what the difference is going to be there?




Lets just wait and see
By rhog on 8/6/07, Rating: -1
RE: Lets just wait and see
By h82bl8 on 8/7/2007 7:03:20 PM , Rating: 1
Only 1 problem with that. AMD is bleeding money like an 80 year old on blood thinners. If they keep this up they have around 3 (Q3,Q4,Q1) quarters until they go bankrupt. I wouldn't think that loosing ~500 Million bucks a quarter and selling your chips for dirt to retain MSS would be considered good. If Barcelona was going to be their silver bullet - we would be seeing benchmarks by now.


RE: Lets just wait and see
By vignyan on 8/9/2007 7:12:18 AM , Rating: 2
Its $600+mn actually... unless you consider $100mn as a negligible amount! ;)

And i dont think not releasing the benchmarks has got anything to do with the bad performance of barcelona... i guess its just about keeping the market curios! :)


RE: Lets just wait and see
By 3kliksphilip on 8/9/2007 12:39:22 PM , Rating: 1
I definitely hope so. I got myself a core 2 duo, but I desperately want the Barcelona to be worth the wait. It's been said a million times before, but if AMD were to go, Intel could raise their prices and lower their specs. It's really unfair- AMD pummeled Intel for years, but because of its smaller size it didn't really bother Intel that much. Most of the people that I know who want to upgrade have Intel 2.8's and stuff. Even the Intel's naming scheme was rubbish- I couldn't remember what a 640 or 760 was, for instance. Now the tables have turned and AMD are struggling. I almost want to know what position the companies are in in a year's time, just to see how desperate the situation is at the moment.


RE: Lets just wait and see
By Targon on 8/24/2007 6:58:51 AM , Rating: 2
There have been benchmarks of Barcelona, but until the B2 stepping, AMD was having some problems ramping up the speeds of the new design.

AMD's K10 design is very good, but putting four cores onto the same die(instead of two dual-cores in one package) is a bit complicated, and it has taken AMD an extra six months to get the kinks out of the process.

R&D in any industry can sometimes have great results(the original Athlon and Athlon 64 chips), but it can also be a disaster(Pentium Pro, Itanium, AMD K5). It doesn't help that until recently, Intel was paying companies to exclude AMD from the market or limit the presence of AMD products.


"The whole principle [of censorship] is wrong. It's like demanding that grown men live on skim milk because the baby can't have steak." -- Robert Heinlein














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki