Good non-biased read on Intels new cpus

[JEDIYoda]

Actually if you read whats been published Intel is not looking or aiming at the performance user!!

All these chips are designed for specific segments users..be they photo editing or graphic designers etc.

Intel has even came out and stated they were NOT going after the performancer user!

So all your hype about intel this and intel that is just that meaningless hype!!

Intel has actually gave up on winning back the performace user segment of the market!!

Instead they are going after specific segments of the user end market!!

Intel just has no interest at present in winning back those who switched to AMD!!

I happen to use Intel and AMD on my computers!!

 

[JEDIYoda]

Excellent article Duvie!!

Intel actually also stated that stuff in there press releases from the last few days!!

 

[Sentential]

Originally posted by: JEDIYoda
Excellent article Duvie!!

Intel actually also stated that stuff in there press releases from the last few days!!

I agree it is a very good article. Ed does avery good job with such things. However I kinda thing he's being a little hard on Conroe and Presler. Sure its not an A64 but its still very good. Especially if the clocks I am hearing are correct (which I am quite confident they are)

If Conroe does nearly as good as what I hear Presler and Cedar Mill are doing, it ought to be able to take it down even the FX57/4800+ *quite* easily. Very soon AMD will actually have to take down a 4800+ (not stock of course) Especially with Cedar Mill breaking the 5000+ mark.

The dominance that A64 X2 once had is going to start erroding soon. I hope AMD goes M2 quickly

 

[Avalon]

Originally posted by: Intelia
BUMPITY BUMP

Take it to FS/FT.

 

[fatty4ksu]

Sorry, but that article was filled with intelly envy.

 

[Fox5]

Originally posted by: JEDIYoda
Is the rumor true?

Intelia is really Dothan using another alias???

I'd say no, Intelia is far more bareable than Dothan, and occasionally says something accurate. Dothan is obviously just a flame baiter.

I think that was more of an opinion than anything else.
The Athlon 64, while architecturally similar, is no XP with an IMC.
Here are the improvements AMD made to the Athlon with the K8 off the top of my head:

? Integrated memory controller
? Greatly improved branch prediction with 4X the number of history counter entries
? Larger, lower latency TLBs
? Extended pipeline for higher clock speeds
? Improved Out-of-order engine introducing the 'pack' pipeline stage
? Higher L1 cache bandwidth
? Significantly lower latency L2 with double the bus width (128-bits vs. 64-bit).
? Crossbar, designed with multicore in mind.

It would be interesting to see what kind of benefit the other enhancements make to the Athlon's performance, rather than just the IMC.
I'd imagine the improved L2 is responible for a significant portion of the boost in gaming performance in particular.
Unfortunately, I've tried to find benchmarks of the Athlon 64 using an external memory controller since I read it was able to do this, but haven't found any so far.

Conroe aswell is no Pentium M.
While the Pentium M may have been the basis of Conroe's design, its 4-issue pipeline alone means that it has changed beyond recognition.
Conroe is probably a bigger change from the Pentium M than the Athlon 64 is over the XP.

Athlon 64 has the same core as the Athlon XP, so most of its theoretical abilities are the same. There are also rare cases when an Athlon XP matches an Athlon 64 in mhz per mhz performance.
There are also cases where a P3 matches a P-M in mhz per mhz performance, and those are much more common than XP matching a 64.

 

[BitByBit]

They're not the same at all.
The K7 was undoubtedly the design base for the K8, but AMD made significant changes to the core, some of which I've mentioned.

Here's an in-depth article detailing the Hammer (K8) architecture and mentions the changes made.

 

[clarkey01]

Originally posted by: BitByBit
They're not the same at all.
The K7 was undoubtedly the design base for the K8, but AMD made significant changes to the core, some of which I've mentioned.

Here's an in-depth article detailing the Hammer (K8) architecture and mentions the changes made.

They would be in uproar if you or me said Dothan was PIII based on P6 wouldnt they.

 

[Acanthus]

Intelia is no longer with Anandtech.

So you should change your topic description Duvie

 

[Duvie]

I will leave it as a reminder of darker days we should all strive to avoid....

 

[Acanthus]

Nice

 

[DAPUNISHER]

Originally posted by: Acanthus
Intelia is no longer with Anandtech.

Not having been around much, I wasn't aware they finally banned that goober. A late round of :beer: for everyone, on me.

 

[her34]

what's wrong with yonah? you get dual core pentium-m. even if it will be replaced within a year, does it matter since it's a notebook? you don't upgrade notebook components

will merom be vastly superior to yonah?

 

[Sixtyfour]

If you look at the AMD lineup, it's easy to see that their Achilles heel is the lack of a low-powered CPU stable that can compete with the Pentium M.

This is one game Intel is winning, so you can see why they'd like to turn it into the game. AMD has nothing like the Pentium M, so Intel would love to turn the CPU wars into a battle where they're strongest and AMD is weakest.

What a load of crap.

AMD has Turion that runs HIGHER mhz than P-M and uses LESS power than P-M (25w vs 27).
If you pick 20 apps and look at the average performance, you will see that Turion is faster.

P-M does NOT support 64bits, P-M does NOT have NX bit, P-M has pathetic multimedia performance, etc.

Ed is full of s***, he turns "non biased" article about Intel to AMD trashing.
He really needs a GF (or BF )

 

[corpseofworms]

Originally posted by: DAPUNISHER

Originally posted by: Acanthus
Intelia is no longer with Anandtech.

Not having been around much, I wasn't aware they finally banned that goober. A late round of :beer: for everyone, on me.

I didn't notice. Did they get rid of Dothan too?

 

[Pabster]

Originally posted by: corpseofworms
I didn't notice. Did they get rid of Dothan too?

Unfortunately not

 

[coldpower27]

Originally posted by: Sixtyfour

If you look at the AMD lineup, it's easy to see that their Achilles heel is the lack of a low-powered CPU stable that can compete with the Pentium M.

This is one game Intel is winning, so you can see why they'd like to turn it into the game. AMD has nothing like the Pentium M, so Intel would love to turn the CPU wars into a battle where they're strongest and AMD is weakest.

What a load of crap.

AMD has Turion that runs HIGHER mhz than P-M and uses LESS power than P-M (25w vs 27).
If you pick 20 apps and look at the average performance, you will see that Turion is faster.

P-M does NOT support 64bits, P-M does NOT have NX bit, P-M has pathetic multimedia performance, etc.

Ed is full of s***, he turns "non biased" article about Intel to AMD trashing.
He really needs a GF (or BF )

AMD still cannot match the Pentium M platform, in terms of power consumption as they have to rely on third party chipsets which are inferior to the Intel designed one specifically for the mobile market.

The Pentium M with J suffix does exist and has Intel XD Bit Support which is Intel equivalent for NX Bit.

Only fairly recently has AMD released 25W turions that extend to higher GHZ levels, with the highest model being AMD MT-40 2.2GHZ/1MB, how this is higher GHZ then Intel's Pentium M 780 2.26GHZ/533/2MB remains to be seen.

The CPU's TDP alone are not enough, you need to compare the platforms against each other, where AMD does have an Achilles Heel due to poorer chipsets.

Regardless of Intel's present performance in mobile/desktop, their marketshare still remains extremely large in comparison to AMD over 80% in both cases.

On another note, those placeholder products might be interesting, Cedar Mill, Yonah and Presler. At the very least they make Pentium 4 technology cheaper for Intel and bring Dual Core to Pentium M architecture.

 

[Sixtyfour]

Originally posted by: coldpower27


AMD still cannot match the Pentium M platform, in terms of power consumption as they have to rely on third party chipsets which are inferior to the Intel designed one specifically for the mobile market.

It's not the 3rd party chipsets, it's about OEM using turions with crappy other parts.
Remember that A64 has integrated memory controller, so A64 chipsets should be competitive against P-M chipsets.

The Pentium M with J suffix does exist and has Intel XD Bit Support which is Intel equivalent for NX Bit.

Since when, last week ?

Only fairly recently has AMD released 25W turions that extend to higher GHZ levels, with the highest model being AMD MT-40 2.2GHZ/1MB, how this is higher GHZ then Intel's Pentium M 780 2.26GHZ/533/2MB remains to be seen.

I forgot that 2.26ghz P-M.. my bad.

The CPU's TDP alone are not enough, you need to compare the platforms against each other, where AMD does have an Achilles Heel due to poorer chipsets.

Intel chipset are not any better, some just claim those are, but it's not the case.
Good OEM can pick better chipset, or worsest chipset.
So it's about large OEM's, and how much they invest to designing of a good AMD laptop.

ps. don't you just love my sig ?

 

[the splat in the hat]

Originally posted by: Dothan
DADDY STRIKES BACK !!!

Intel has chips coming that are so fast it make AMD look like a duck !!!

YONAH, MEROM, SASSAMAN, what awesome chips we have in our sights !!!

AMD user had better start selling !!!

hahahahahahah coff, burp farrrrt , HAHAHAHAHAHAHAHAHA......

 

[corpseofworms]

Originally posted by: Pabster

Originally posted by: corpseofworms
I didn't notice. Did they get rid of Dothan too?

Unfortunately not

Why the hell not? I'd much rather have Intelia around who occasionally tries to back her statements up than the guy who tried to ban half this board. . .

 

[Dadofamunky]

I generally agree with the article about Intel's direction. To my eyes, it's also the case that Intel does offer a better overall mobile solution than AMD at the present tme. Though the Turion does seem to be viable, AMD doesn't have anything equivalent to the Centrino/Sonoma/Napa chip set that accompanies the Pentium M. This does not mean that Turion is a bad chip. far from it. But it is never going to penetrate to the mass market to the Dells, HPs, and Gateways. That's just a fact.

***EDIT*** Yep, Turion does look really good - and cheaper overall. Intel charges more for the integration of the Centrino/Sonoma chipset, which seems logical to me. If I knew I could trust a specific Taiwanese vendor to stand behind their product, I would have considered it, all else being equal. But it's likely you won't see more than a third-tier provider (or one or two second-tiers) providing Turion laptops. So there you take more of a chance in quality. ***END EDIT***

Also consider the fact that portables are outselling desktops (by a thin margin, to be sure). The article also mentions Apple in a sidelong fashion. It's no accident that the product line that Intel announced is also the same product line that Apple is actually adopting, based on the apparent chronological sequence for Apple's switchover.

It really looks like Intel is ceding the performance desktop market to AMD. Given that it's a relatively small slice of the market, I think Intel will probably be comfortable with that as long as they continue to improve their mobile and mainstream platforms.

If AMD ever really nails a truly integrated mobile solution, more fur could fly down the road. I think Turion is a first baby step for AMD. Now that they effectively have an almost unassailable base, they are in a better position than they were several years ago when they were just Intel copiers. They have succeeded in forcing Intel to respond to competition. Now if someone could just do the same thing to Microsoft.

 

[pm]

So Intel is going to start their 65nm fabs with one kind of processor, then start changing over to another six months later. Sounds like a mess to me.

Give me a break. It's Gordon Moore's 2nd law of semiconductors - always keep your fabs full. When you intro a new process, you should always put on a product that minimizes changes from the previous generation - because you'll have your hands full debugging the new process, and also you want to make sure that the new design is ready and the more features you pack into it, the more risk there is that the schedule will slip. This is the way it's been since the early 80's. I can remember 386's processors that did this... and this author calls it "a mess"? It's supposed to be standard operating procedure - both at Intel and pretty much everywhere else. If you have spent $4 billion on a new 65nm fab, the last thing in the world you want is for it to sit idle while you debug some new different design - stick with the basics for process debug and then bring in the optimized designs later... not a new thing at all.

 

[DarkKnight69]

The thing that impresses me about the dothan I have (and I am planning on upgrading soon, is that when I am dont with it, I can pull it out, throw in in a HTPC and run it stock on passive cooling. I had it on passive cooling before I o/ced it!

 

[Duvie]

Hey go run that oc'd dothan in the autogk test thread!!! let see how it does on a desktop environment with OC'd ram and bus likely!!!