AMD Launches Turion Mobile Processor

justforaday writes "Earlier today, AMD launched their Turion mobile processor, which is based on the AMD64 architecture. This is set to compete directly with Intel's Centrino (Pentium-M) line of processors. Chips will initially be clocked between 1.6 and 2.0 GHz. Looks like we should be seeing some nice low-powered 64-bit notebooks in the near future."

Comments

Availability

[fembots]

In my town of 1.5M people, I haven't managed to find a store that sells AMD-powered laptop. We must be the forgotten ones.

Re:Availability

[spywarearcata.com]

I've been using a Fujitsu AMD laptop for some time. The following URL should work from your town just fine:
http://www.computers.us.fujitsu.com/

Re:Availability

[Anonymous Coward]

They have this new thing called the internet. It sells you stuff and things.

Re:Availability

[Anonymous Coward]

Before this came out (and we haven't seen the benchmarks yet), why the hell would you want one?

Re:Availability

[lanc]

Come on. The diversity in HW-business has lost lots. Have a look at NetBSD's supported architectures. The most of them are sadly dead. HP gave up alpha, Sun is panicing and running to x86_64/Linux, the only two players that managed to bring something new are IBM, and the Power-architecture, and Apple with the G5-way. But Intel's Itanium is quite dead, though AMD is coming up, the Opteron is worth a look. Give them a little time, let diversity have a comeback.

Re:Availability

[Anonymous Coward]

You realise that the G5 is Power architecture don't you?

And, only Sun's low-end systems are x86_64. The high-end Sun Fire beasts are still SPARC-based.

Re:Availability

[dsginter]

You live in a "town of 1.5M people" and you don't have a Best Buy?

Not that I like Best Buy.

You *could* order one from that thar intarweb, too.

Re:Availability

[UserChrisCanter4]

HP/Compaq sell AMD-based notebooks. Gateway/eMachines sell AMD-based notebooks. Fujitsu sells AMD-based notebooks. Sony has sold AMD-based notebooks, although I can't speak about their current line-up. All four of these comapnies have their products available in large consumer electronics stores such as Best Buy, CompUSA, Circuit City, etc.

In fact, the only big-name companies who don't seem to offer AMD notebooks are Dell (obviously), Toshiba, and IBM. Toshiba's the only one of that group that even sells through retail channels anymore.

Where do you live?

Re:Availability

[sonoluminescence]

I read that as "one-point-five MegaPeople".

I need to get out more.

Re:Availability

[Matthaeus]

But he already said he doesn't have a laptop! How's he going to get on the internet, doofus?

Yes, this is sarcasm.

Re:Availability

[lanc]

You realise that the G5 is Power architecture don't you?

erm, yes, you could be right there I guess.

And, only Sun's low-end systems are x86_64. The high-end Sun Fire beasts are still SPARC-based.

Yes, yet. But I am not anymore sure how long, I mean have a look at how the old/great/noble Sun with his lordly Sparcs, and high-ends is rushing to make business out of the x86/linux market. I mean I hope I am wrong with this, and do really wish a sparc/solaris(10) comeback.

Re:Availability

[UranusReallyHertz]

MegaPeople. I see these all the time, especially at all-you-can-eat restaraunts. I think there number is increasing rapidly in America.

Re:Availability

[Anonymous Coward]

They're running rampant while not stomping around slashdot.

Re:Availability

[jsight]

I know I've seen them in stores, but try Gateway and eMachines to see for yourself.

Re:Availability

[niko9]

They have this new thing called the internet. It sells you stuff and things.

Thank you Al Gore!

Love Slashdot

Re:Availability

[corngrower]

MegaPeople. I see these all the time, especially at all-you-can-eat restaraunts.

Yes, and his town is so small it only has 1 1/2 of them. That's probably why the store in town doesn't sell AMD notebook computers.

Re:Availability

[NanoGator]

" We must be the forgotten ones."

That and Intel pretty much owns that market. From what I understand, Intel is generally preferred as the laptop processor mainly for its cost savings.

If I'm in error, please correct me. Keep in mind, though, that I have a small light-weight TabletPC running on a Centrino processor that runs between 3 and 4 hours on a single charge. I'd really be interested to know if AMD's laptop processors to date can even compare.

Re:Availability

[nvrrobx]

Where do you live? Circuit City, BestBuy and CompUSA all sell Athlon powered notebooks.

HP makes a few Pavillion models with Athlon processors.

Re:Availability

[bprime]

I work at CompuSmart (compusmart.com), where I will gladly pimp^H^H^H^Hintroduce you to lots of Acer and Compaqs that are powered by Athlon XPs and Mobile A64's. Probably only helps you if you're in Canada, though.

Re:Availability

[Anonymous Coward]

ROFL OMG PWNT N00B /lick

Re:Availability

[RzUpAnmsCwrds]

"In my town of 1.5M people, I haven't managed to find a store that sells AMD-powered laptop. We must be the forgotten ones."

Do you have a Sams Club, a Best Buy, an Office Depot, or a Circuit City?

You're not looking hard enough.

Re:Availability

[Anonymous Coward]

TAHT D3S3RV3S A R3AL ROFL

Re:Availability

[Anonymous Coward]

I'm pretty sure it's call the internets.

Re:Availability

[utlemming]

I just bought myself an HP with a 3200+ AMD64, and all I can say is wow! I love the thing. It is a little bit on the heavy side, and the fan runs (although not a loud fan) all the time. But the performance is mind blowing. I am so happy with the purchase. At school we have dual P4 2.8's that we use for programming, and this thing runs the programs on campus just as fast if not faster. I am a little bummed about not waiting for the Turion, but frankly, I am extremely happy, and would recommend an AMD64 notebook to anyone. I looked for a while and when the price was right, I jumped on it. Anyhow, like I said, if your thinking about an 64 laptop go for it. You won't be disappointed with the performance (well, except for the hard disk lag).

Re:Availability

[gl4ss]

you managed to actually go to a store that sells laptops and managed to actually ask for one or a list of laptops that they sell?

redundant, yes i know, but seriously.. if you look at any place that lists/has even 4 different models there's pretty good chances of there being amd based one in there.

that or you live in zimbabwe or something.

you could buy one in any western town with 100 000 inhabitants...

Re:Availability

[shutdown -p now]

It enlarges your penis as well!

Re:Availability

[Anonymous Coward]

He may not have a laptop, but he managed to post to this group. So, he should be able to order a laptop on the internet. Right?

Re:Availability

[Anonymous Coward]

thanks for the sarcasm but no, he didn't tell us that he doesn't have a laptop. reread his statement...

Re:Availability

[ShieldW0lf]

Ewh.. HP?

Yuck!

Re:Availability

[Anonymous Coward]

the intraweb? whats that? is it new? can i eat it? does it taste good?

Re:Availability

[jericho4.0]

You're blind. I know of two places in my town of 10,000 w/AMD64.

Re:Availability

[chthon]

I have an AMD64 laptop, and with cpufreqd installed I can get 2 hours of run time.

Mind you, my screen is always at its brightest, and maybe the power management could be better, and I also installed 512 Mb RAM extra (up to 1 Gb).

Re:Availability

[chthon]

Even with less habitants. When I was looking for a new laptop last year (September), there was already much choice in laptops with AMD64 processors.

Re:Availability

[Anonymous Coward]

If only they would drop that Broadcomm wireless crap.

Re:Availability

[hhlost]

http://www.linuxcertified.com/linux-laptop-lc2464. html
< drools >

Re:Availability

[TeknoHog]

I read that as "one-point-five MegaPeople".

You read it quite correctly. This is how scientific prefixes work.

A megasecond, for example, means one million seconds. It's not some special kind of second.

Question

[Anonymous Coward]

This may seem like a silly question, but won't that only be useful if the laptop is going to have more than 4GB of memory? How often does that happen?

Re:Question

[mabinogi]

mine currently has 2GB, and it was purchased last year.....

so > 4G won't be all that far away.

But apart from that, there's more advantages to Athlon64 than the AMD64 instructions (which you're not going to be able to take advantage of right now anyway unless you're running Linux).

Re:Question

[bersl2]

There are other advantages, namely the additions to the x86 ISA that fix some of its shortcomings. There are 8 more general-purpose registers and 8 more SIMD registers.

Re:Question

[reality-bytes]

But apart from that, there's more advantages to Athlon64 than the AMD64 instructions (which you're not going to be able to take advantage of right now anyway unless you're running Linux).

And I thought the only people buying AMD64 CPUs *were* Linux users ;-)

Re:Question

[delirium of disorder]

Might not be that common now, but the ammount of RAM necessary for the latest applications is always increesing. I've looked at SPARC based laptops with dual processors and 16GB of RAM. 64-bit is definitly necessary if you are looking at the high end.

Check it out.
http://www.partnersdata.com/cgi-bin/searchit em?id= 8434

If you don't want to bother to get a quote, one of these configured with just about all the extras will run you $46,933 according to a quote I got.

Re:Question

[Anonymous Coward]

Somebody can correct me if I'm wrong, but I'm pretty sure that current x86 machines can address over 4GB of ram as it is. Linux supports 64GB already, if I'm not mistaken.

Re:Question

[dsginter]

This may seem like a silly question, but won't that only be useful if the laptop is going to have more than 4GB of memory? How often does that happen?

As often as it does in game consoles.

But, let us imagine for a moment, that some breakthrough in memory technology allows for a high-density, non-volatile, high-speed, random access memory. At this point, we can just stuff 10 or 20GB of this new NVRAM in the machine and not worry about a hard drive and its power consumption.

Re:Question

[powerlord]

I agree ... and let me be the first to say that 64bit CPUs should be enough for everyone ;)

Re:Question

[Bun]

32-bit processors are limited to addressing 4GB in TOTAL, including paged RAM (swap files or partitions). With a 64-bit operating system, an AMD64 laptop will be able to fully address 2GB of physical RAM and another 20TB or so of paged RAM (if your hard drive is big enough). Also, it gets you by the 2BG per-process limit of 32-bit Windows applications. Not a bad thing, IMHO. I don't know who is likely to need that kind memory in a mobile platform today, but in the not-so-far future it could be a necessity. See this article for an example of the advantages of 64-bit architectures in a photographic manipulation application.

Re:Question

[Bloater]

As bersl points out, there are other advantages. But I'd like to correct a common misconception - the 32 bit PC chips from (I think) Pentium pro onward from intel, and some time later from the other pc chip manufacturers, support much more than 4GB even though they only support 32 bit addresses in the CPU registers. 32 bits limit any given process on an OS such as Linux or Windows to 4GB of addresses at any one time. Although how much of that can be mapped to real storage is OS and OS version dependent. The AMD 64 initialy provides a virtual memory address space of 2^18 GB, but that is not directly related to the amount of physical memory that can be operated by an OS. apparently that 2^18 (262144) GB may be extended to 2^34 (17179869184) GB in later models.

Re:Question

[Jeremy Erwin]

Let's see here:
log(64 gb)= log(1Gb)+log(64)=36 bits, so you're probably thinking of Intel's "Physical Address Extension" scheme, which allows programmers to mess around with 16 memory segments, each of 4 gb. 64 bit chips (such as the Athlon64, POWER, Sparc, and Alpha) can address large amounts of unsegmented memory-- the PowerPC 970, for instance, can address up to 4096 gigs of memory, as it has a 42 bit address bus.

On a PAE system, processes still can only address a certain amount of memory (3 GB, default), though you can configure your system to push the kernel out of the way and onto a different memory segment. This slows down syscalls, though.

Re:Question

[Anonymous Coward]

nah, 64-bit registers are excellent for floating point operations, and they're way more fun to write assembly for...

Re:Question

[Anonymous Coward]

It's not just about the memory silly. You can run software that can use 64 bit integers.

Re:Question

[andreyw]

Incorrect. Please read the Intel docs, specifically the sections on PSE/PAE, which both have been avaiable since the Pentium Pro. The virtual addressable space will still be 4GB, however the 36-bit address bus will allow for a theoretical limit of 64GB.

Re:Question

[Celestial Avenger]

Just like how Gates said 640K should be enough for everyone, right? We've heard it before. ;)

Re:Question

[Hack Jandy]

Athlon 64 processors are not really useful because of their 64-bit addressing capability. The thing that makes Athlon64 (and Turion, and the newer Semprons) speed demons is the on board memory controller integrated into the CPU. No other x86 processors do this.

Re:Question

[Weirsbaski]

This may seem like a silly question, but won't that only be useful if the laptop is going to have more than 4GB of memory? How often does that happen?

The existing 4GB limit, which a 64-bit chip+OS gets around, is a virtual address space limit. Running a 64-bit OS will let more than 4GB of code/data exist in the address space at the same time.

Existing 32-bit x86 chips can already use PAE extensions to access more than 4GB of DRAM plugged into the motherboard, although only 4GB will be accessable (ie- mapped in the virtual space) at any one time (and there's lots of overhead in changing the 4GB currently mapped).

Re:Question

[bluGill]

The 80386 supported more than 4GB of memory. I don't think anyone actually built a motherboard that could take that much memory, but it could in theory deal with it. You just had to use the same ugly segmented memory model that everyone hated about the 286 while in 32 bit mode. Everyone just set their segment to 0, since that still allowed 4GB of ram, and nobody with a 80386 had that much memory.

Re:Question

[bluGill]

As the other guy said, not true. The 32 bit 80386 supported more than 4GB of RAM, if you could find that much back when it was a used (you couldn't). You just had to use segmented memory which everyone hated. (Back then the 286 was still fresh in everyone's mind, and the 286 couldn't do anything non-trivial without segmented memory. Today we are spoiled because few people care about such things.

Re:Question

[OAB_X]

Not to metion 3D Now! and Ext. 3D Now!, both of which Direct X supports (in varying degrees of implementation).

There is no legal reason for Intel not to use them, they just have a hard time admitting that AMD beat them at something I guess.

Re:Question

[mp3phish]

your are correct. There are tons of compiler optimizations that can be done witht he athlon64 that cannot be done with any previous version of the x86 ISA... That being said, compiler technology isn't really taking advantage of this yet...

The real advantage to having an athlon64 in a laptop is power savings and raw memory bandwidth and ultra low latency...

For the first time ever (aside from crusoe but their mem bandwidth is not that much) we are seeing ultra low powered CPU's put into 12in sub-notebooks with on die memory controllers.. This is the single most advantageous feature of the athlon64. This is also packed into the Turion. This integrated memory controller means there is no exchange with an off die chip soldered onto the mobo (which adds several fold latency to every memory request and write). Instead, the CPU talks directly to main memory. This reduces memory latency to a fraction of intel's and older athlonXP laptops. and since latency is reduced significantly (im talking several times lower than the fastest centrino's memory controller) and memory efficiency is increased, you get much more memory bandwidth on top of all the above advantages.

It has already been shown in every reviewer's benchmarks that memory intensive applications run significantly faster on athlon64's than any other platform (including Intel P4 EE) This is due to raw bandwidth and raw quickness (latency)

Now, we can all sit here and say that the chip is capable of calculating 64 bit numbers without having to break it down, and that it is capable of running more than 4GB ram, and that it has more general purpose registers (which aren't taken advantage of when running winXP 32 btw) and that it has shed a lot of legacy logic which is implemented in microcode instead (like older stuff DOS uses)

But in the end it is the memory controller which is the real improvement with athlon64. The 64bit-ness of the chip, and all those other features will have their time to shine.. Just not in the near future (the lifetime of a new laptop you might buy this year)

Re:Question

[Anonymous Coward]

This post brought to you by Automated Joke Destroyer 5000.

Re:Question

[Ohreally_factor]

2 GB should be enough for anyone. =)

Re:Question

[Tarqwak]

To be fair, Transmeta Efficeon range has integrated DDR memory controller too. It's not a true x86 CPU but it can morph x86 instructions into VLIW ones.

It's a shame that it hasn't been too successful :/

Re:Question

[Bloater]

386 segmented memory maps directly to addresses in the 32bit logical address space. That means even the large granularity segments don't help (granularity flag in segment descriptor set). You need to be able to map the 4GB logical address space to places in the larger physical address space. That means using the paging extensions from the pentium-pro (if somebody can correct me on which processor introduced it, I'll accept that since I'm not bothered enough to go through my intel architecture manuals).

Re:Question

[bluGill]

Its been too long since I've seen the 386 specs....

However the short of it is: adding these changes to the 386 is fairly trivial in that no changes to the instruction set are required, and only changes to a small part of the hardware. (Of course this isn't to imply any idiot could do it, only that it isn't a major effort)

Re:Question

[renoX]

> That being said, compiler technology isn't really taking advantage of this yet...

Are you sure? While programming a register allocator which works efficiently with 8 registers must be really hard, going from 8 to 16 is pretty easy, no?

Especially for multi-target compilers which generates also code for RISC and their big number of registers..

Now I agree that fine-tuning can be hard: if memory serves, the first 8 "legacy" register in x86-64 can be used with shorter instruction (1 byte smaller) so you can gain a little performance by ensuring that the most used registers are legacy registers..

Re:Question

[mp3phish]

Yes and no...

In general yes it will be easy to comple to the new registers... But it will take a much longer time to tweak it to those registers... let me explain

Yea, a compiler can immediately take advantage of the new registers because they are essentially general purpose just like the other 8... but they are currently tweaked to the smaller 8 registers and will have a hard time being tweaked to the 16 regs because they must maintain backwards compatability (which implies they will keep their compiler tuned to the 8 rather than 16) and I just assume that they are still focusing on 8 registers in the general case...

Eventually -- years down the road -- the mindset will be more registers and they will have a new set of assumptions to code to.. Not to mention that current 32bit processors have register re-naming which kinda-sorta eliminates the dire need of more registers (kinda-sorta because it is still better for compilers to have more, but just not as detremential with register renaming)

Now, that being said. I'm not a compiler coder so you can't take everything I said above to be 100% accurate but it is reasonably justified to myself considering what I know now.

price points

[bwthomas]

AMD Turion 64 mobile technology models ML-37, ML-34, ML-32, ML-30, MT-34, MT-32, and MT-30 are priced at $354, $263, $220, $184, $268, $225 and $189 respectively, in 1,000-unit quantities.

Doesn't this seem like a high price for bulk chips?

Re:price points

[ackthpt]

Doesn't this seem like a high price for bulk chips?

No. How long have you been following CPU pricing? It's always high for the first offerings. By this time next year you'll see half that line discontinued and the remainder heavily discounted.

Rule #1 regarding technology: As soon as it hits store shelves, it's already obsolete.

Re:price points

[Anonymous Coward]

Have you looked at Pentium M prices?

http://www.intel.com/intel/finance/pricelist/

Re:price points

[Erwos]

I was under the impression that high-end Pentium M's ran in the $600 range. Those prices actually struck me as quite low.

-Erwos

Re:price points

[Upphew]

Yup. Your impression seems right: http://www.pcworld.com/news/article/0,aid,118258,0 0.asp

I'm Waiting For Benchmarks

[periol]

It's not like AMD has been stellar in the mobile processor world before this. Centrino has been a no-brainer for a while in the laptop world.

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Yeah, a no-brainer. Anyone who buys anything other than AMD has no brains.

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Does that mean the fact that I bought an Acer Aspire 1513LMi with 64bit 3400+, 15" TFT, GeforceFX Go5700 64MB, 60GB hard drive, DVD+-RW, 802.11g, gigabit ethernet etc. etc. etc. mean that I have no brains? Ha ha ha ha!!! Keep your piddly centrino laptop, I have real work to do.

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Acer?! LOL, riiiight. I think we know who does not have any brains here.

Acer is the Hyundai of the computer world.

Re:I'm Waiting For Benchmarks

[ShnowDoggie]

Here is a link to Tom's Hardware:
http://www.tomshardware.com/hardnews/20 050310_1036 52.html

Right now AMD seems to be on top of the game. INTEL is a bit behind. From all that I have read, AMD has had good luck with the 90 nm process. That means less heat, better battery life. And more power when you need it. Of course complete overall performance is still to be seen....

Re:I'm Waiting For Benchmarks

[periol]

it's /. so i don't expect you to read, but i'm writing from my Athlon desktop machine, and I've had AMD chips on my last three desktops.

Notice that's DESKTOPS.

In my experience, the majority (but just to help your remedial reading skills, that's NOT ALL) of the people buying the AMD 64-bit laptops up until now probably really needed a desktop instead. The AMD mobile chips are terrible and have been for a while and consume an awful lot of power. Combine that with better design for pentium m and centrino laptops, and yeah, it has been a no brainer. I wouldn't mind that changing though.

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Acer is the Hyundai of the computer world.

Nope. Acer is the Ferrari of the computer world. It even goes VROOOOMM when it boots up. I wish I was kidding, but I'm not.

Re:I'm Waiting For Benchmarks

[Anarchitect_in_oz]

i would have said...
"anyone whos buy anything other than PowerPC for a laptop has no brains."

Re:I'm Waiting For Benchmarks

[Entropius]

Centrino is a no-brainer in only two segments of the mobile market: the thin-and-light long-battery-life market and the lots-of-cash market. The top-end Pent M's may equal the top-end mobile Ath64's at some tasks, but not for anything near the same price.

However, the mobile Ath64 has its own niche carved out: relatively inexpensive mainstream/DTR notebooks. The extra battery life provided by the Pentium-M isn't that big of a deal on machines with big screens and beefy video cards, where the processor isn't the primary power consumer.

I'm typing this on an eMachines M6811, an A64 3400+ machine with plenty of goodies, a big screen, and a video card beefy enough to play anything I throw at it. I got this machine for $1200; you'd be hard pressed to find a comparable Pentium-M machine for the same price, even a year later. It gets 2.5-3 hours on battery.

If you're looking for all-around laptops that aren't that expensive, have decent battery life, and offer beefy performance, the A64 beats the Pent-M hands down.

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Jesus Fucking Christ. That was a troll. And you retarded, slashbot moderators put it at 'Funny'. It was a blatently retarded fanboi post. I'm going to go away and cry now.

As for the rest of you who replied to it, YHBT. YHL. HAND.

Re:I'm Waiting For Benchmarks

[ciroknight]

Looked into the Celeron-M much? Pretty much all of the same offering as a Pentium-M, minus some of the super nice speedstep features, and some L2. It's for the midrange laptops, and I'd say AMD will have a hard time competing with it, except for the fact that Intel has pushed "Celeron" as meaning "cheap" instead of "better option for most people who don't need bells and whistles". I believe if Intel pushed Celeron as being their mainline, they could use the Pentium line as being in the same position as the "Pentium EXTREEEEM edition", and probably charge a bit more for them. Of course, that'd be even more reason for Intel bashing.

They've both got something coming if IBM decides to start making PPC-based PCs and Laptops. If Freescale would just hurry up with those damned Dual core G4's, and IBM would go ahead and trim the G5 down...

BTW, I'm a known Intel enthusiast. But in my defense, I own at least as much AMD/PPC equipment.. so I'd rather consider myself a "microprocessing enthusiast".

Re:I'm Waiting For Benchmarks

[Anonymous Coward]

Looked into the Celeron-M much?

Not the OP, but I know for a fact that A64 3400's spank the celeron-M up and down the block. Not saying it sucks, it's a fine chip in its own right, but performance wise, the OP's processor is a better choice in terms of performance and the price is pretty good.

AMD launches 64-bit Turion processor for notebooks

[Anonymous Coward]

Article without the annoying green advertisement links.

Expected to Launch Notebook PCs Worldwide in Coming Months

AMD today introduced AMD Turion 64 mobile technology, the latest in a succession of computing innovations based on the industry-leading AMD64 architecture. AMD Turion 64 mobile technology is uniquely optimized to bring award-winning AMD64 performance to thinner and lighter notebook PCs with longer battery life, enhanced security, and compatibility with the latest graphics and wireless solutions, both today and tomorrow.

Together with industry-leading partners, AMD enables a best-in-class ecosystem of open industry-standard technologies, helping PC manufacturers to deliver feature-rich systems that satisfy the variety of ways in which people use their notebooks.

"By embracing a broad partner community and industry standards, AMD is both providing choice and stimulating innovation," said Roger Kay, vice president of Client Computing at IDC.

"We have unleashed 64-bit mobile performance, allowing business professionals and on-the-go consumers to explore the freedom of mobile productivity with AMD Turion 64 mobile technology," said Marty Seyer, corporate vice president and general manager, Microprocessor Business Unit, Computation Products Group, AMD. "This is just the first of many innovations that we are planning to pioneer with this new product family made for mobility, choice and best-in-class notebook designs."

Leading global manufacturers, hardware vendors and software suppliers have expressed support for AMD Turion 64 mobile technology. "Fujitsu welcomes AMD Turion 64 mobile technology innovation, and we congratulate AMD on their success," said Mr. Kazuhiro Igarashi, general manager, Mobile PC Division, Personal Business Group for Fujitsu Limited.

Business professionals and consumers worldwide can expect to see notebooks based on AMD Turion 64 mobile technology starting later this month in retail stores and through commercial distribution channels. The first systems are expected from Acer worldwide and Fujitsu Siemens Computers throughout Europe.

"We have had a great deal of success with our Ferrari branded notebooks based on AMD64 technology," said Walter Deppeler, senior corporate vice president, Acer Inc. "We look forward to continued success with our forthcoming notebook designs based on AMD Turion 64 mobile technology."

"As the leading European IT provider, we offer our customers world-class computing technology," said Herbert Schonebeck, Vice President Business Unit Consumer, Fujitsu Siemens Computers. "We are introducing innovative consumer notebook PCs based on AMD Turion 64 mobile technology to provide our customers efficient platforms to connect from virtually anywhere."

In addition, ASUS, Averatec, BenQ, MSI and Packard Bell are among the leading, global computer manufacturers who have indicated they will support AMD Turion 64 mobile technology.

Availability

AMD Turion 64 mobile technology models ML-37, ML-34, ML-32, ML-30, MT-34, MT-32, and MT-30 are available immediately worldwide.

In order to help consumers and business professionals simplify their notebook purchasing decisions, AMD Turion 64 mobile technology uses a new series of model numbers designed to provide a simple designation of both relative performance and degree of mobility within the processor family. The two letters of this model number indicate processor class, with the second letter designating increasing degree of mobility, as measured by power consumption.(a) As the second letter approaches the end of the alphabet, "higher" letters indicate greater mobility. The numbers indicate relative performance within the processor class. Higher numbers indicate higher relative performance among the AMD Turion 64 mobile technology family.

Pricing

AMD Turion 64 mobile technology models ML-37, ML-34, ML-32, ML-30, MT-34, MT-32, and MT-30 are priced at $354, $263, $220, $184, $268, $225 and $189 respectively, in 1,000-unit quantities.

What I'm interested in...

[PornMaster]

Are any of the major motherboard manufacturers going to put out a micro-ATX or smaller desktop motherboard for this to build a media PC upon?

Re:What I'm interested in...

[HoneyBunchesOfGoats]

DFI and AOpen both make boards. See http://www.techreport.com/reviews/2005q1/dfi-855gm e-mgf/index.x?pg=1 and http://www.extremetech.com/article2/0,1558,1728546 ,00.asp for reviews of each.

Re:What I'm interested in...

[Lisandro]

You can find them, for these and the Pentium-M. They're not cheap though - neither the motherboards nor the processors.

If you want to keep it affordable, you'll have to check Transmeta or VIA, though they lack a bit in performance.

Re:What I'm interested in...

[Lisandro]

Bad phrasing - you will be able to find them :) The CPU is brand new.

Re:What I'm interested in...

[drinkypoo]

How is this informative? Those are Pentium-M boards. We're talking about Turions here. They're not even out yet.

Re:What I'm interested in...

[RzUpAnmsCwrds]

"Are any of the major motherboard manufacturers going to put out a micro-ATX or smaller desktop motherboard for this to build a media PC upon?"

If it's socket-compatible (as are most AMD mobile CPUs), you'll be able to use it in desktop motherboards.

Regardless, you can already undervolt a A64 Winchester significantly. Mine runs at 1800MHz / 1.1V @ ~ 23W.

I have a Zalman CNPS7000AlCU thermally controlled by my BIOS, and the fan doesn't even turn on unless I fully load the CPU for several minutes. The CPU idles at 42C with the fan off (of course, this is at 800MHz / 0.8V).

Re:What I'm interested in...

[Anonymous Coward]

AFAIK, it uses a stock Socket-754 mobile socket that's already being used by the mobile Sempron, among other CPUs. All that's needed is BIOS support for the Turion, and it will probably work on desktop Socket-754 boards as well.

Re:What I'm interested in...

[dago]

All recent AMD mobile processor had the same socket and run in the same motherboard as they desktop version (some may better support the mobile features, some not)

Logically, it'll be the same for the turion, so just take a normal mATX motherboard and that should be ok.

This is contrary to you can do with intel's p-M, which has a nice price premium and requires special expensive motherboards.

Re:What I'm interested in...

[Nuffsaid]

Since the socket appears to be the same 754 used for Semprons and Athlon64, you should be able to put one of these in many desktop motherboards with, at most, a simple BIOS update. The same applies to upcoming dual-core CPUs from AMD: they should slip into socket 940/939 with no problem, which is amazing when you think that you replace one CPU with what are essentially two in the same hole. Nice thing is that AMD, in contrast to Intel, doesn't make chipsets, as it hasn't any interest in making you buy a new motherboard.

Low Power? Better Link

[ackthpt]

Looks like we should be seeing some nice low-powered 64-bit notebooks in the near future."

First, why such a lame link for this post? Here's a more informative one Editors, sheesh!

Ok, let's look at this without the rose coloured glasses:

64 bit CPUs 25 - 35 watts (peak)

Better get used to lugging a massive battery if you plan to really get the most out of that 64 bit CPU. Honestly, why else would you have a laptop with such a race horse in it if you weren't going to be using it to its capacity? Ok, maybe you're one of the few who run a fully 64 bit OS (probly Linux) with fully 64 bit apps (did yer own builds targeted at AMD64 instructions) and you really need the full addressing, but won't be using all the ponies. Odds are, you'll be an early adopter for somewhat nebulous reasons. Cool, but not essential

Re:Low Power? Better Link

[bani]

that's not too bad, compared to the 25w of intel's pentium-m.

also remember than intel understates their peak power while amd overstates theirs. dont recall who did the test, i think it was the german c't mag who found the discrepancies between claimed and actual power consumption.

Re:Low Power? Better Link

[Anonymous Coward]

laptops make nice replacements for a desktopm where space is at a premium, where you can use the power for intensive tasks, the mobility feature in being a laptop comes in as a nice bonus.

Re:Low Power? Better Link

[ackthpt]

laptops make nice replacements for a desktopm where space is at a premium, where you can use the power for intensive tasks, the mobility feature in being a laptop comes in as a nice bonus.

I use a laptop as my primary workstation. The downside is they usually have much slower bus speed, which you tend to notice when loading windows apps or large files.

Re:Low Power? Better Link

[doormat]

Better get used to lugging a massive battery if you plan to really get the most out of that 64 bit CPU.

What? Peak use is 25W-35W. Pentium Ms use 22W, but then you also have a northbridge and a southbridge. Whereas most Athlon-64 based (and therefore Turion 64) have only one other chip (other than the CPU). At the least, they're about the same in terms of power usage.

Re:Low Power? Better Link

[Surt]

What they commonly have is a much much slower hard drive, which you notice when loading apps or large files. 4200 rpm is most common, 5400 less so, and 7200 rpm quite rare. Any typical desktop has at least a 7200 rpm drive, and quite commonly 10k or even 15k rpm drives. You can get a surprising amount of perceived speed boost by ordering a laptop with a 7200 rpm drive.

Re:Low Power? Better Link

[Fweeky]

These are peaks too; you'll only be hitting them with your laptop plugged into the wall and running cpuburn, otherwise they'll be ticking over in low power mode.

Re:Low Power? Better Link

[toddestan]

That's not that bad power usage at all. Remember, a lot of current laptops (as in just about any low to lower midrange model, as well as many "desktop replacement" models) have a full desktop chip crammed in there. The exact same chips that put out 75W+ in your desktop computer. Just like this Toshiba with a full blown Northgate 2.0Ghz P4. Oh wait...ow...my leg...!

It's exciting to finally see AMD come out with some chips that can compete with the Pentium M in power usage though - the XP M just wasn't quite there.

Re:Low Power? Better Link

[10Ghz]

Usually Intel publishes the "typical" power-consumption of their CPU, whereas AMD publishes the maximum power-consumption of their CPU.

Hmm just now released

[OverlordQ]

I swear they started back in January

Pretty smart processor naming...

[Infinityis]

If you tell someone about your computer processor, and you cough slightly while you say the processor name, you end up with a Turing machine. If only it were true...

Re:Pretty smart processor naming...

[Jerf]

you end up with a Turing machine. If only it were true...

You want a Turing Machine in your laptop?

Me, I find lugging that infinite tape around a real bitch.

Re:Pretty smart processor naming...

[beelsebob]

The Turing Machine does not specify an infinite tape - it specifies that if more tape is required it must be available - so your laptop must instead be able to produce tape at any time it needs it.

Re:Pretty smart processor naming...

[FRiC]

... but if you don't get the cough quite right, you could end up with a Durian machine.

Intel / AMD = Microsoft / ???

[Anonymous Coward]

I really wish we could complete that ratio.

No, I'm not saying that Intel is as bad as Microsoft (although those recent anti-trust actions might make one wonder). I'm just saying that look how well things have turned out for the consumer on the processor front because there's a viable alternative.

Wouldn't it be cool to have a competitor for Microsoft the way AMD competes with Intel?

Re:Intel / AMD = Microsoft / ???

[SelectionShort]

What about Apple? They about to be a lot bigger competitor. And then there's linux too. There is a lot of competition but only we (Slashdot readers) pay attention to them.

Willy G

Re:Intel / AMD = Microsoft / ???

[Elranzer]

IMO the ration seems to be... Intel/AMD/IBM = Windows/Linux/MaxOSX

Re:Intel / AMD = Microsoft / ???

[Anonymous Coward]

Wouldn't it be cool to have a competitor for Microsoft the way AMD competes with Intel?

It's coming.. but would YOU support such a company they way you support AMD? and no, we are not talking Linux as it's a rather "niche" market. Some Linux Zealot will now mark this post as flamebait. Linux (It is only a Kernel) doesen't have the Edge or the backing to take it where it needs to go to beat Microsoft -- when I say this, I mean the likes of Novell/SuSE, Red Hat, IBM, etc.. (a small hint would be IBM selling off it's PC division)

Yes, I know they have done "some" good things with/for Linux. For them it is just $$$, they are not giving the 110% they could.

Apple you say? no, I can't seem them doing any better then they already are.. give it a few more years and they will be back to where they was. Apple are the Yo-Yo are the computer industry, they start of with something good, then somehow slip up? hmm, plus they always seem to be quite happy with there "5%" of the industry. They certainly lack the minerals. Steve Jobs is nothing more then a business man that knows how to use the little people that make up the 99.9% of Apple Computer, Inc.

Microsoft really do need a REAL competitor, they are quite stale compared to the once vibrant company it once was.

Infact.. the whole computer industry could do with a radical shake-up.. at best it seems "Same shit, different day".

Give it 5 years, and there will be some light at the end of the tunnel..

Some Like it Hot, not Me Though

[slipnslidemaster]

Is there any information on it's thermal output? Is there a reference design for how laptops should be designed to handle the heat?

I'm in the market for a new tablet and while I love it, the Pentium M that I'm currently using turns my lap into a puddle of skin and rayon within a few minutes. I'm due for an upgrade and since AMD is always a leap or two ahead of Intel, I'm wondering about the heat.

Any thoughts?

Re:Some Like it Hot, not Me Though

[VoidWraith]

Thermal output is about equal to power consumption. It matters how big the die is, but only in terms of how much can be dissipated at a time. So look at it pumping out a little more than the Intel chips.

And don't use any laptop (of recent construction at least) on your lap unless if you value your prospective progeny.

Re:Some Like it Hot, not Me Though

[jlehtira]

The power consumption seems similiar to the Pentium M, so the heat must also be. How hot the laptop-bottoms get depend on their design. What I do with my laptop is put a paper folder between it and my lap. It's sturdy, has lots of paper for insulation and it's a bit soft, so it's really more comfortable in every way ;). Plus, I always have the folder with me.

Forbidden

[Anonymous Coward]

That's because AMD would get into big trouble if they sold these in Pyongyang.

Re:Forbidden

[Anonymous Coward]

TAHTS T3H FUNN3H ---> IT MAKES ME lol AND rofl

Re:Laptop

[alienspanke]

"Now I can finally get a new laptop"

---Why now, the AMD will be no better or more capable than Intels Centrino chips. And the Centrino clocks at a higher 2.2.

Also there's no on-board WLAN adaptor dispite AMD's attempt to market the Turions "wireless compatibilty".

I'm gonna stick with Intel.

Now then Apple ...

[Anonymous Coward]

... get your shit together and give me my g'damn G5 Powerbook!

Re:Now then Apple ...

[Anonymous Coward]

Amen!

Re:Now then Apple ...

[Enrique1218]

My thoughts exactly. This is embarrassing. AMD beat IBM to the punch and developed a low power 64bit processor. Comon, its a PowerPC!!! It's suppose be low powered!!! Remember the G3!!! Get it together!!!

So that means...

[game kid]

...my new laptop won't lower sperm count? Seriously, I can't stand putting a hot laptop on (oddly enough) my lap, so they become tabletops instead for me.

These Turions better generate less heat or I'd have lost all hope. And feeling in my groin.

Re:So that means...

[Infinityis]

You know, your right. I think maybe the women of the world must have conceived laptop computers--they are infinitely more suited to such devices. They don't have to worry about sperm count, fried leg hair, and plus, anyone whose witnessed a mom/wife/etc. cooking knows that women are more resistant to heat in general.

Of course, on the flip side, it could be a ploy by men to ensure the continuing presence of female secretaries/administrative assistants...

Re:So that means...

[klevin]

The vast majority of sperm killed by using a laptop on your lap are because the the heat buildup caused by sitting with your legs pressed together (something like 75% of the increase in temp). The laptop contributes relativley little to the heat buildup.

Any guy who routinely sits with his legs pressed together (as opposed to the common, more relaxed posture) runs a higher risk of infertility/low sperm count.

Re:So that means...

[Animaether]

It's modded Funny for obvious reasons, but I think it should be noted that this is in fact not a laptop; it is a notebook.

Then again, I don't know when I last saw an actual laptop. The terms laptop and notebook have been used as if they're synonyms for quite a while now.

The difference between the two should be obvious. One is designed for use truly 'on the go' - on your lap when riding the bus if need be. That'd be the laptop.
A notebook, on the other hand, may be portable and easy to carry around - but is primarily intended to be used in a proper workspace - and thus on a table/desk/whathaveyou.. just not your lap.

Other differing factors are:
- overall weight (you don't typically want 6lbs+ on your lap, or carry it around a lot)
- overall size (notebook with giant display, too bulky)
- balance (don't want something on your lap that tends to tip over to the left due to all the major electronics being on that side)
- heat (heat should be dissipated and fanned out to the side, airflow intake should be coming from the other side, the back, or in the center/towards the back (so as not to get covered by one's thigh)
- access ports, auxiliary button should be on the side (otherwise you have to set the computer aside when, for example, loading a DVD and the DVD bay inserts at the front. oops..

And probably a few more points I've forgotten since I took that ergonomics class.

Re:So that means...

[Anonymous Coward]

it could be a ploy by men to ensure the continuing presence of female secretaries/administrative assistants...

I second that plan. Any plan by women to give us a wireless pr0n route is fine by me.

Re:So that means...

[themysteryman73]

That's why they call them notebooks now.

Re:So that means...

[draziw]

For home or office use - check out the iLap
http://www.raindesigninc.com/ilap.html
(I wouldn't travel with it in my bag)

Silent computing applications

[apharov]

This should be a very interesting product for the silent computing community (check SilentPCRReview) if some manufacturer decides to make desktop motherboards for it. People are already using Mobile Athlon XP's and Pentium-M's in desktop computers because of their relatively low heat output.

My own wish would be to some day have a passively air cooled computer. Running an Athlon64 passively with only one big slow (=nearly silent) case fan is already feasible with carefully managed airflow inside the computer case. Turion could be theoretically cool enough to cool with convection airflow if the potential is realized with wellmade motherboards.

Re:Silent computing applications

[Anonymous Coward]

I have one of those "passively cooled" computers. It's called a 486.

Re:Silent computing applications

[oringo]

There are already many passively-cooled ones. Check out mini-itx.com.

But I like my AMD as it is...

[nilbog]

I like being able to make toast with my old processor...

Re:But I like my AMD as it is...

[Lorkki]

Buy a new P4 and you can fry your steaks as well.

Heat?

[Cannedbread]

my p3 1ghz laptop has problems with overheating, luckily its out of warranty. With AMD chips running at the temperatures they do, it will be interesting to see how these laptops manage to stay cool. my balls cant handle anything that hot so im staying away

Re:Heat?

[sonoluminescence]

I hear that with the new 90nm process that AMD chips are running much cooler these days.

Re:Heat?

[Anonymous Coward]

"AMD chips running at the temperature they do"? What temperature is that? Do you have any idea? Or are you just parroting the stupid crap people have been spouting ever since some idiot showed us that yes, if you remove a processor's heatsink, bad things happen?

In case you're wondering, the hottest processors available now are from Intel, clocking in at a toasty 100 watts. Intel's planned dual core processors are predicted to achieve a full 130 watts. AMD's hottest processors don't break 90. This chip runs at 35 under maximum load. This is very comparable to a G4, so laptops built on it will probably resemble powerbooks in their heat output.

Digital AlphaBook - 64 bit notebook in 1998

[BiggRanger]

I have a Digital AlphaBook, the first 64 bit notebook computer! 266 MHz Alpha Processor (EV4 I think) 128 Meg RAM, and runs OpenVMS! Best of all I got it back in 1998. Nobody was even thinking 64 bit desktops back then. And you know what, it still rocks. I've rebuilt the battery pack and still use it today. You should see poeples face when VMS is booting up and it goes into the DEC windows manager.

Re:Digital AlphaBook - 64 bit notebook in 1998

[msbsod]

Too bad they don't make them anymore! The future for a new OpenVMS/Itanium notebook looks extremely bad. HP does not want to kick Microsoft's a$$ with a real good operating system for the desktop and notebook market. And Intel dropped all plans to produce an Itanium for desktop machines (Itanium was suppose to be The 64-bit replacement for the Pentium!), not to mention any plans for a notebook version of the Itanium. AMD understands that a processor has to cover a broad market. Well done, AMD!

Re:Digital AlphaBook - 64 bit notebook in 1998

[Anonymous Coward]

Nobody was even thinking 64 bit desktops back then.

Not unless you count Silicon Graphics. Sure, they are considered workstations but people used them as desktops.

Re:Digital AlphaBook - 64 bit notebook in 1998

[turgid]

Not unless you count Silicon Graphics

...or Sun (UltraSPARC) or DEC (ie Alpha) or HP (PA-RISC) or IBM (POWER).

Re:Digital AlphaBook - 64 bit notebook in 1998

[Anonymous Coward]

The Sun laptops might have you beat on the date for 64-bit laptops, or they might not. Can't remember right now.

Re:Digital AlphaBook - 64 bit notebook in 1998

[TheGratefulNet]

I'd love to see a photo of it booting. (I used to work at DEC and have fond memories of vax/vms...)

could you snap a photo while still in the text-mode bootup phase?

TIA,

Turion

[kihjin]

Vegetables... what's next, candy? I think I see a marketing idea here... anyone looking for a Chocolate Chip?

Re:Turion

[Anonymous Coward]

Actually they dropped the D in turdion...

Yeah

[Anonymous Coward]

Yeah I am looking forward to those HOT Lenovo laptops running that chip and Gentoo!

Hype64

[itedo]

I just bought a notebook (IBM Thinkpad R51) and I am pretty happy with that (FreeBSD runs on it perfectly).

But why the heck are they putting 64bit in a mobile computer? Well, we are living in a consuming world of almost two types: buy-and-throw-away and buy-although-you-don't-know-what's-it-good-for

How ridiculous..

Re:Hype64

[captain_craptacular]

Maybe because you gain more than you lose by going to the 90nm 64bit architecture? It's faster and cooler than the chip it replaces. It just happens to be 64bit at well.

It's not like AMD just said "hey, lets blow some smoke up the consumers arse and put a 64 bit processor in a laptop!". If you hadn't noticed they've been moving in this direction for say, a couple years now.

Re:Hype64

[itedo]

Of course, the 64bit Notebook was inevitable. But I tend to a common sense, not to a marketing strategy.

The priority of a good notebook is of course performance and mobility - not being bloated with less or even unsupported technologies.

Re:Hype64

[drinkypoo]

Have you noticed that the Hammer core runs 32 bit applications faster than either the corresponding[-speed] Athlon XP part, or the competing Pentium 4 part?

You don't even have to use it as 64 bit to get benefits from using a laptop with this new core. The power management in the Opteron and Athlon64 is also a bit better than the power management in the mobile Athlon, from what I understand, and I'm not even talking about Turion here, just the normal processors.

Re:Hype64

[Anonymous Coward]

The sooner 64-bit processors become common in hardware, the quicker they'll be supported well in software.

So I consider this a good thing. I'm already happily running FreeBSD/amd64 on my laptop, but it would be nice if my next laptop could be 64-bit with longer battery life...

they should

[Anonymous Coward]

... get your shit together and give me my g'damn G5 Powerbook!

They should think seriously about it. Apple has always had great laptops, but in the last couple of years the x86 has progressed a lot. I don't really know if buying a powerbook with a G4 CPU really pays off, when compared with latest's encarnations of pentium-m or the one from this announcement. It's not 64 bit, it's just the whole thing - the pentium-m "II" looks very very nice and G4 is...well, a bit old, and may I say "slower"?

1MB max of L2 cache - suckage

[melted]

The reason why Pentium-M is as performant as it is is in part because it has 2MB of on-die cache. Don't expect miracles from a chip that has a half or even one fourth of that.

Re:1MB max of L2 cache - suckage

[Tumbleweed]

L2 size is _far_ the only measurement of CPU performance. Also, the first generation of Pentium M had only 1MB of L2.

Re:1MB max of L2 cache - suckage

[Anonymous Coward]

Insightful?! This joker doesn't have a clue what he's talking about.

Re:1MB max of L2 cache - suckage

[rob_squared]

Just remember the thing about cache, more doesn't always equal better. Remember that the more cache you have, the more addressing you have to do to properly manage it, and that takes prescious time.

Re:1MB max of L2 cache - suckage

[ChronosJB]

Umm...the performance benefit of that 2MB L2 cashe is not that much better than 1MB. It is maybe a 3-7% increase in performance. Granted that is Intel vs. Intel not Intel vs. AMD since AMD just released this CPU. AMD CPUs have usually shown better efficiency when it has come down to computation power with 512kb and 1MB L2 cashe and Intel has had troubles even with 2MB L2 cashe due to the memory controler being on the motherboard and not on the CPU.

If AMD were to up their L2 cashe to 2MB or higher, the performance gain would probably be about the same as Intel's, 3-7% increase per 1MB increase.

We will see who the real mobile CPU winner is once we get some benchmarks to compare, but until then we are all speculating as to which CPU is better.

Re:1MB max of L2 cache - suckage

[Tarqwak]

2 MB L2 cache doesn't do much by itself, give that Dothan core a 533 MHz FSB instead of the 400 MHz Banias was on and it starts to shine (and of course the power consumption goes up too). Speaking of which, AMD Turion 64 (and Athlon 64) have memory controller in the core not in north bridge like Intel processors, so that 25/35W CPU power envelope includes MC overhead too.

BTW, most 512 kB L2 Athlon 64 CPUs run circles over P4 3.6+ GHz with 2 MB cache in gaming benches, so cache alone means squat ;)

And it looks like it's going to be two years until Pentium M range gets 64-bit core. AMD Turion 64 already has 16/16 INT/XMM registers, SSE/SSE2/SSE3 support etc.

Re:1MB max of L2 cache - suckage

[drinkypoo]

L1 cache is more important than L2. FAR more important. Pentium-M has 32+32kB. Turion has 64+64kB. In case you didn't notice, that's twice as much. (Your math was ok on the L2 cache, but still...) Don't expect miracles from P-M, a chip that has half as much L1 as the Turion.

Re:1MB max of L2 cache - suckage

[merreborn]

Now, is the Pentium-M a 64 bit processor that runs two 32-bit instructions per clock cycle?

All I know is Pentium-Ms benchmark like a desktop Pentium 4 of twice the clockrate (that is, a 1.5 ghz pentium-M performs like a 3.0 ghz desktop). I have no clue *why*.

Re:1MB max of L2 cache - suckage

[oringo]

P-M's 2MB L2 cache also has much less latency than AMD64's L2 cache (~10 cycles vs ~20 cycles). So applications whose data fill well into the 2MB cache run really fast, but those who can't will suffer through a 533MHz FSB.

Re:1MB max of L2 cache - suckage

[Anonymous Coward]

Performant isn't an english word. What does it mean?

Re:1MB max of L2 cache - suckage

[toddestan]

The reason why Pentium-M is as performant as it is is in part because it has 2MB of on-die cache. Don't expect miracles from a chip that has a half or even one fourth of that.

My AMD Athlon XP only has 256KB of L2 cache, you insensitive clod!

Re:1MB max of L2 cache - suckage

[m00j]

Has to do with things like the longer pipeline of the P4 being less efficient. The Pentium M is (loosely) based on the Pentium 3. If you remember when the Pentium 4 came out clock for clock it was worse than the Pentium 3. It is just overall more efficient - the P4 is very very inefficient simply because they can clock it higher that way, and Joe Dumba$$ comes to the store and thinks that a higher number means more performance.

Try explaining to someone why a 1.3Ghz Celeron M (based on the Pentium M - the new Celeron M is basically a Banias without the advanced power management) is higher performance than a 1.3Ghz Pentium 4! That is hard :p

Re:1MB max of L2 cache - suckage

[TeknoHog]

Pentium-M has 32+32kB. Turion has 64+64kB [of L1 cache].

Fine, but if you're running 64-bit code on the Turion and 32-bit on the P-M, then the effective size (number of items stored) is the same.

Re:1MB max of L2 cache - suckage

[ponos]

Fine, but if you're running 64-bit code on the Turion and 32-bit on the P-M, then the effective size (number of items stored) is the same.

The code size in 64-bit mode is NOT double. All 32-bit instructions work unchanges in 64-bit mode. The jump/memory addressing instructions need 64-bit operands and a special prefix, but unless you are really working with 64-bit ints most of your instructions are unchanged (although the ones that use the new 'r' registers will need a prefix). I'd estimate the increase in code size to be ~30-50%.

As an example (not very accurate) gcc-3.3.4 (32-bit) cc1 is 2964192 compiled for i686 with moderate optimization (slackware package) while gcc-3.4.2 (64-bit gentoo) compiled with -O3 -march=athlon64 cc1 is 4222760 (42% larger). Note that optimizations are for speed, not for size.

P.

Intel fanboys will love it

[tepples]

AMD shot itself in the foot. Look at how the Intel fanboys will murder the name: "Hmm... Centrino or Turdion?"

Re:Intel fanboys will love it

[Anonymous Coward]

Cuntrino.

Anyone that lowers themselves to pointless name calling of opponents products in their fanboi world automatically forfeits any argument.

Re:Intel fanboys will love it

[Anonymous Coward]

Cuntrino

FIGURE 1.1: An example of Anti-AMD ownage.

Re:Intel fanboys will love it

[ChronosJB]

LMAO! That is so true...I have been in local computer stores and listened to uneducated people being scammed into believing that P4 desktops and laptops are better than anything AMD has to offer. He also went as far as calling AMD a worthless company that sells worthless CPUs (odds are he was being paid to say those things because what real computer person would make such a blatant lie. Intel may be better in somethings that AMD is not and vice versa, but to call AMD garbage is just not very computer geeky).

I just get strange looks when I start to gag, wheeze, and cough *bull sh!7!* then walk away.

I have also talked several people into not purchasing a computer that says "Intel Inside" because the place that sells these computers they have internet connections (stupid them) and I go to sites and show these customers the benchmarks that compare the two CPU types. About six or seven times the people I showed the benchmarks to walked out with an AMD64.

Yeah me!

Re:Intel fanboys will love it

[Freaky Spook]

I walked into franchised computer store in my area that deals with HP/Compaq units & a few other major brands.

I was bored & pretending I was in the market for a new PC when a salesman was trying to flog me off a 533mhz P4 2.8 with a 80gig hdd & 256mb of RAM for what I could have paid for a new Prescott running a 120gig SATA & 512mb of Ram.

When I enquired about AMD products he laughed at me and said they were unreliabale, prone to overheating, weren't compatible with as many produtcs as intel & that Windows even ran better on a P4 then an AMD processor.

I just walked off, I can't believe the amount of people that would actually walk into that store & get suckered out of their money.

Question:

[Neuk]

Stupid little boy:"Papa Beaver?Why does AMD always do things that are of no worthy advantage?" Papa Beaver:"Well son,it's because of an interesting process called 'within ten years I can safley promise that our company will be the most wealthy and industrious of all in our class'." Stupid little boy:"Huh?" Papa Beaver:"No more questions.Let's have a story."

Cen-Turion

[Anonymous Coward]

...Obviously wanting to be the strongest. Or to have the last word. Or both ;-)

--JAB

Pretty corny processor naming...

[Anonymous Coward]

What's next... the Hardion?

Blast Off...

[sinfree]

To where did the launch it? The Moon?

kind of late?

[bogaboga]

Aren't AMD kind-of late? When one talks about mobile computing, "centrino" comes to mind. Right?

Re:kind of late?

[Rosco P. Coltrane]

Aren't AMD kind-of late? When one talks about mobile computing, "centrino" comes to mind. Right?

No, "heavy suitcase" and "outlet in airport to recharge goddamned battery that failed in a middle of a save" is what comes to mind when one talks about mobile computing.

Re:kind of late?

[Ford Prefect]

Turion ... Centurion ... Centrino?

Admittedly the average laptop buyer probably doesn't care about Latin or the Roman army, but perhaps that's how the name was derived?

People keep asking me if my Latop has 'Centrino'. Given that it has a big Apple logo on it, I feel obliged to explain what Centrino is - it's not WiFi in general, just an Intel chipset. Maybe AMD were hoping for a bit of that brand-name to rub off?

Re:kind of late?

[Ziviyr]

Centrino hasn't sold me. you need more than one word and alot of colory moths to sell me a product.

Re:kind of late?

[Anonymous Coward]

That's why they also gave you the blue men flying around on centrino hoverboards. Crap man, what more do you want?

Re:kind of late?

[Ziviyr]

Ya know, my mother was half blue. Now we have big corporations paying people to act like idiots in blueface. Just not cool...

Re:kind of late?

[tomstdenis]

I dunno about peeps like you but I use the battery in a pinch only anyways...The real benefit of a laptop [imho] is just the ability to carry a computer around with you that's less bulky than a desktop.

I don't see the value of cracking a laptop open while walking around or something [even a small one] cuz the likelyhood of dropping it [at least for me] is too high.

So I'm likely to have a plug available [even when I lug my american laptop to france for business for example...].

As for the issue of "amd not being portable" that's just lame. I've taken my AMD XP-M [barton like] based laptop pretty much all over. It's never "too much to carry" and comes in very handy.

I'm not saying pentium-m based laptops aren't cool they're just not that impressive. My brothers Pentium-M Compaq 2200 series laptop gets four hours on battery... so does my AMD XP-M Compaq 2100 series laptop...whoopy

What people still fail to understand that is at idle your motherboard, ram, screen and hard drive still take quite a bit of power.

Tom

Alan Turing laptop

[Anonymous Coward]

I'm not quite sure I want something named after Alan Turing sitting on my lap.

Re:Too much embedded crap

[Dashing Leech]

"Whatever happened to just adding a better battery and a pcmia card."

I'd say the exact opposite. To me, progress is doing something more efficiently. Adding a better battery is like saying progress in automobiles is best achieved by leaving them as gas-guzzlers and giving them bigger tanks. If a computer can do just as much with less power consumption, and produce less heat, I'm all for it. I hate wasting energy, especially since I pay for a lot of it. Other benefits of lower power and/or heat generation are less fan time (annoying noise) and a smaller and lighter power supply.

The replies are all true.

[Anonymous Coward]

The parent poster has no clue.

Now what would it take to

[antikristian]

put that cpu in a well-built reasonable (portable) sized laptop with a nice screen (high dpi), good nvidia graphics and a "normal" pricetag.

I fear something has to freeze over before I get my dream laptop, I might as well start building it out of lego...

Re:Now what would it take to

[Chandon Seldon]

You, like many other people, seem to be confused about what a good computer should cost.

Let me clarify:
The minimum price for a decent 3d-capible desktop computer is about $800. For a laptop, it's twice that.

Re:Now what would it take to

[Anonymous Coward]

I'm viewing this on my AMD64 notebook (a Compaq Presario R3140CA. Hey shut up! I like it! ), and I get as much life out of my battery as I did with my old Centrino notebook and my older Athlon-XP notebook (lotsa heat off that sucker), and it seems well-built, I carry it everywhere, the 16x9 widescreen is nice, it's using a 32 meg Geforce4 (I'm not happy about the 32 meg part tho), and it cost me 1500.00 CDN, which was not bad at all.

Mind you, I have heard that lego solved the heat dispersion issue, so you might indeed go with that, I s'pose.

So what you are saying is...

[evolutionaryLawyer]

using my powerbook for photoshop in my lap can reduce my fertility?

WooHoo! I am gonna photoshop the hell out of some shit and then go have freaky irresponsible sex!

Nvidia

[Anonymous Coward]

Try finding a decent laptop with a geforce chipset :/

Face it - Intel/ATI own the mobile segment.

Pentium M's aren't cheap either

[IoN_PuLse]

It seems to be a trend with the good low-power high-performance laptop chips, because the Pentium M's sure aren't cheap either...

AMD and Laptops??

[hawkeye_82]

My heart weeps and my thighs burn

Re:I used to name my machines with Dragonball Z

[Anonymous Coward]

what

Re:Too much embedded crap

[jedimark]

Hmm.. Integration is definitly a key to making things more energy efficient. Couple that with killing off some legacy stuff, and I see a good future in low power consumption for computing in general.

While my notebook doesn't have serial ports, parallel ports, floppy drives, etc, I know the (cheap sis) chipset still has some wasted circuitry inside (while although probably doesn't consume much power in it's unused state, the space could be reused to bring things closer together). (And how long until they get rid of 16bit mode in the cpu.. yuck.)

In the meantime, maybe it's time to regain some of that currently wasted heat back into electricity. Help run the fans off CPU heat?

And maybe a plugin solar panel to charge the battery?

But I guess it kind of hampers portability.. :-(

Thats my 2 cents..

Pronunciation?

[sabNetwork]

I'm not really into carnivorous plants, so I hadn't heard this word before.

Here's a WAV pronunciation.

tur-e-on

Not very phonetic. Shouldn't product names be easy to read without hearing them?

I imagine that I'll hear many tur-yun's, tur-eye-un's, and too-rye-un's.

Re:Pronunciation?

[Anonymous Coward]

What will happen next?

People pronouncing Aluminium Aluminum?

Re:Pronunciation?

[Anonymous Coward]

Could be worse. Could be called "Muine"...

Re:Pronunciation?

[Ewasx]

Shouldn't product names be easy to read without hearing them?

You mean like the rest of the english language?

Why is it _far_ more important?

[melted]

I thought the bandwidth between RAM and the processor was the main limitation of the current processors, not bandwidth within the processor itself.

Besides, P-M has a maximum power dissipation of 24W (or 12W for ULV version) whereas Turion is 35W. Those P-M laptops already run pretty darn hot. I'm not sure I'm willing to buy a laptop that's 40% "warmer".

Re:Why is it _far_ more important?

[Panaflex]

Here's what I'm thinking.

Pentium M: L1 cache 32+32 (64Kb total), L2 cache off chip 2Mb, Northbridge + Southbridge chipset

AMD Turion: L1 cache 64+64 (128Kb total), L2 cache on chip 1Mb, Single-chip chipset.

So, yeah, Pentium-M by *itself* will consume 22 Watts, but you've got to add in the North+South chipset... guessing it's about 10 watts or more.

Turion is basically 25 or 35 Watts, plus a single chip chipset I'd guess would only have 5 watts usage since the memory controller is on the Turion itself.

So, Total watt usage would be roughly:
Pentium M: 24W + 10W = 34watts
Turion: 25W + 5W = 30 watts.

Am I guessing right here?

-Pan

Pentium M boards and Turion Chipsets

[mparaz]

Those are boards with chipsets for Pentium M "Dothan."

Here's a review for the DFI 855GME-MGF

VIA announced a Turion chipset - for laptops. I'm curious as to what Socket these use (754? 939?) - no mention on the sites I've read.

Re:Pentium M boards and Turion Chipsets

[sagekoala06]

they use 754 http://www.amd.com/us-en/Processors/ProductInforma tion/0,,30_118_12651_12658,00.html

Re:AMD launches 64-bit Turion processor for notebo

[toddestan]

In addition, ASUS, Averatec, BenQ, MSI and Packard Bell are among the leading, global computer manufacturers who have indicated they will support AMD Turion 64 mobile technology.

Packard Bell?!? Is this some kind of joke? I thought they died about the time the PII came out.

Turion= 2 chips in 1

[beesquee]

Keep in mind these chips have integrated memory controllers and northbridges unlike Pentium-m's. Northbridges alone use about 5W of power so this could be subtracted from their consumption for a more accurate comparision of the two.

Hyundai

[zogger]

think again

Compaq/HP has a nice Athlon64 notebook...

[Anonymous Coward]

and it's cheap. It's the R3000Z and for a little over US$1000, you get an Athlon64 3400+ + 512mb RAM + 60gig disk + 54mbit WIFI + DVD burner + the usual assortment of ports and 1680x1050 15.4" display. Just go to Compaq's website and shop away.

I haven't seen many A64 notebooks or PCs at retail locations though. I bought one of the above notebooks a few months ago and friends of mine have been gradually adopting them since it seems to be a good mix of power and price.

YMMV

Turing?

[ic3p1ck]

Ah, but the question is does it pass the Turing Test?

Turion only has memory controller on chip

[melted]

So it still needs the full chipset. Therefore you're not guessing right.

Re:Turion only has memory controller on chip

[Panaflex]

No, the memory controller is on the CPU. SO you have one less chip. (The northbridge usually has the MCU.)

To sum it up: a full motherboard comparison will perhaps consume less wattage.

-Pan

Re:Turion only has memory controller on chip

[drinkypoo]

It might need a north and south bridge still - AGP is normally in the north bridge, along with the memory controller, and the interface to the south bridge. I don't know much about chipsets for hammer but it would make sense that you would only need one chip to handle agp and the normal south bridge functions if you didn't need an interface between bridges, and you didn't need a memory controller. Even if it is two chips, it's still going to be simpler than the intel chipset, which contains a memory controller.

speed boost... but detremental power savings...

[mp3phish]

Yes.. you can take the advice of the parent poster... but let me explain a couple things which dictate that 7200RPM laptop drives are not always better....

1) Drive density: a laptop drive at 80GB is roughly twice as dense (bytes per square inch) than an 80GB desktop hard drive

2) Head travel time: In a 2.5in laptop drive, the hard drive head does not have to traverse more than ~1in (from center to outter track) of the drive to do any given seek. A desktop drive must traverse roughly 50% more distance (and time) on random seeks

3) Heat: 5400RPM drives produce significantly less heat than 7200RPM drives.. on top of this, a 5400RPM drive built with the same quality as a 7200RPM drive has significantly longer average lifetime.

4) Noise: Equivilently built drives one running at 5400RPM and one at 7200RPM. The 7200RPM is significantly louder and produces high pitched noise (nowadays, all laptop drives are hydro bearings so you can't cop out and say that a cheaper 5400RPM drive will use cheaper ball bearings... like you used to be able to say)

5) battery life: less heat == less watts == less power consumption. Remember that the hard drive is the second largest drain on your battery when talking about centrino/turion systems (LCD is the first). In a P4 laptop then the CPU uses more power than the drive. a 5400RPM or 4200RPM drive has faster spin up times. It has lower sustained power consumption, and will generally give you a longer battery life on the order of an half to full hour or so in a midsized notebook using default battery.

6) Data integrity/ruggedness: a slower spinning hard drive will not have as detremental of an affect on your data if the drive is bumped during reads/writes. Think of it as hitting a large speed bump going 54MPH vs going 72MPH.

Now some of this needs explanation: (1) tells about drive density. What this means is that a 7200RPM 3.5in drive is about as fast as a 5400RPM 2.5in drive in sustained reads/writes. (2) tells about seek time (latency). A laptop drive at 5400RPM has a faster seek than a desktop at 5400RPM. I will admit that a 7200RPM desktop drive has faster seeks, but not significantly so. A faster spinning drive can seek faster only when the head is in place and it must wait for the drive to rotate to the corect angle in order to read the requested data. It does NOT make the head travel faster. For this reason high RPM speeds are good and well for seek time... but using smaller platters is also a very good way to reduce seek time. Not to mention that loading programs and loading large video files or photoshop files, etc... are not highly seek dependant. They are sequential read dependant. Database accesses, or accessing a badly fragmented hard drive are cases where faster seek will help you out. But in a laptop system where you are loading programs and files and keep your disk defragmented it will do you very little good.

Now.. I'm not saying that 10K and 15K rpm drives are bad... they are great for seek time and they are great for high transaction/sec databases... What they are overkill for is desktop systems which the user would typically be loading programs or transferring files from one disk to another or loading large files... Because most 10K drives are around 36GB or 74GB... and they are actually marginally _SLOWER_ at sequential read/write than cheaper and larger 7200RPM drives on the desktop....

now.. Apply all of the above to a laptop 5400RPM drive vs the standard desktop drive of 7200RPM.. I hope you are able to see what I'm getting at. a 7200RPM drive in a laptop is significantly faster than an equivilently sized desktop drive at 7200RPMs... Anybody who says they feel a slowdown on 5400RPM laptop drives vs. a 7200RPM desktop drive is either using very low GB laptop drive, or very large GB desktop drive, or is just fooling themselves... In general, a 5400RPM laptop drive performs approximately the same as a 7200RPM desktop drive in most end user desktop applications...

Re:speed boost... but detremental power savings...

[Anonymous Coward]

Your points make interesting reading... but practical experience says otherwise. A 5400 rpm Laptop drive seems significantly less fast than one on a desktop. In fact yours is the first argument that I have heard to the contrary. Can you point us to some benchmarks that support your arguments?

Re:speed boost... but detremental power savings...

[eelke_klein]

These benchmarks do not support his claims: Notebook
Desktop

Re:speed boost... but detremental power savings...

[Anonymous Coward]

You cannot compare these two benchmarks because the drive sizes are different, that was the whole point of the original poster.

Re:speed boost... but detremental power savings...

[mp3phish]

Those benchmarks don't make sence.... My origional argument talks about 1) Sequential reads and 2) Seek times...

I never claimed that the 5400RPM drive is faster in ALL respects... But in sequential reading a 5400RPM drive is as fast or faster than the same size desktop 7200RPM drive simply because of raw data density

This tomshardware review doesn't touch on these points..

Re:speed boost... but detremental power savings...

[mp3phish]

I'm sorry I don't have hard independant data to back up my claims... I do have a lot of experience with them and have run my own benchmarks.. (I'm mostly talking about sequential reads here, as seek is not that important on an average user's computer)

I'm just a hardware enthusiast trying to bust out a myth that i have always had a prob with..

Re:speed boost... but detremental power savings...

[moonbender]

Many things you say are true, but with respects to speed you are simply wrong. You say in another comment that you don't have any hard data - well, I'm looking at hard data, namely the current issue of the German computing mag c't. The average sustained transfer rate (read) of the 5400 rpm 2.5in drives are within 20 to 30 MB/s, while the 7200 rpm 3.5in drives routinely transfer at 45 to 55 MB/s, with 10krpm and 15krpm drives scoring even higher, obviously. And in a recent test on StorageReview a 10krpm (!) 2.5in HD targeting servers doesn't fare very well, either.

Re:speed boost... but detremental power savings...

[eelke_klein]

1) The read performance graph gives you the min, avg and max sequential transfer for the disk tested. Min is at the inside, max is at the outside. 2) What do you define as seek time? The access time given on tom's hardware page is the time of positioning the head and waiting for the data to pass. Which is AFAIK the most common measurement used.

Maybe you do not think its fair to compaire notebook disks from dec 2004 with 3.5" disks from feb 2004, how about these from march 2003 3.5 even the Maxtor D740X with 40 GB capacity is faster. Lets try some older models march 2001. Lets see they still have a slight advantage even with only 20 GB the platter. But who compaires disks from march 2001 with disks from december 2004?

Re:speed boost... but detremental power savings...

[mp3phish]

Your numbers are wrong.. I have a 160GB 8mb cache WD1600JD SATA drive here and on the outter tracks of the disk it only gets 40MB/s sustained...

On an 80GB 7200rpm drive it is much lower than that.. compared to an 80GB laptop drive it will be faster...

You only quote german's magazine on 2.5in vs 3.5in.. but you dont' quote any drive storage sizes... The storage size is relevant to the drive density... Of course a desktop drive at high density (such as 250GB) will be faster than a common 40GB laptop drive at 5400RPM.. it has more density even though the 40gig is smaller...

I am talking about a direct comparison. 80GB laptop vs 80GB desktop drive.. If you measure those you will get a faster rate on the laptop drive even though it runs at a slower RPM.

Re:speed boost... but detremental power savings...

[mp3phish]

Just wanted to make another followup... most 10K drives go slower than 7200RPM drives at sustained sequential reads. The reason is because most 10K drives are at most 74GB.. (you can get some larger scsi ones but they are through the roof in price)

a cheaper 250GB 7200rpm drive will beat out a more expensive 10K 74Gig drive in sustained reads/writes.. (im not talking about seek here)

And that is where I think that German magazine gets it wrong.. They do a weighted average of seek and sustained reads and writes with buffering to try to determine an overall "drive performance" rather than individual benchmarks for each type of access...

Re:speed boost... but detremental power savings...

[mp3phish]

You are comparing a 3.5in disk with 100GB per platter times 2 platters to a 2.5in drive with a single 80gb platter (or a single 100GB platter) so obviously the desktop drive will beat it.. it has a higher density AND a higher spindle speed.

Re:speed boost... but detremental power savings...

[moonbender]

I referred to "average sustained transfer rate (read)", I thought that was pretty explicit. This is not some overall drive performance number. They got the number by reading the whole drive from first to last sector, a standard testing method to get this kind of number as far as I am aware.

As I said, the standard laptop drives were about half as fast as the standard desktop drives throughout the test. WRT comparing drives of the same size, that's hard when there are so few desktop hard drives available in 60 or 80 GB. For what it's worth, a SCSI 10kprm 3.5in HD (Maxtor Atlas 10K V) with 70 GB was nearly THREE times as fast as some of the 2.5in HDDs. 71.8 MB/s - nice! And the fastest drive in the test is a 35GB SCSI 15krpm drive (Fujitsu MAU3036NC) with a mean sustained read of 77 MB/s - 3.5in, obviously.

Re:speed boost... but detremental power savings...

[eelke_klein]

A two platter disk does not have a higher transfer rate then a single platter disk! Only one head is used for reading at a time. I can know I have had a 20 GB version of the Quantum Fireball Plus AS 60GB tested in the march 2001 article it has the same transfer rate and seeks slightly faster because it has less heads to move. If you actually has taken the time to read the articles you would have noticed that the seagate was the only 100 GB/platter drive in the test the other drives had lower densities. You would also have noticed that the fastest disk in the notebook test was actually a 60 GB 7200 rpm disk and even that one can hardly keep up with the old 20 GB 7200 RPM fireball I have had.

Re:speed boost... but detremental power savings...

[juhaz]

I don't know if that WD drive you got the numbers from is just old (relative to the laptop drive your putting it against), or if your system has some issues, but I've got to agree with the others here, you're wrong.

And yes, the age does make _huge_ difference. I've got 1½ years old Hitachi that does about 60MB/s, few years older Maxtor gets about the same as you do, ~40'ish.

Fastest 7200RPM IDE/SATA 3.5" drives do about 65MB/s sustained outer zone transfer rates, 30-40 on inner zones. Fastest 2.5" drives reach bit under 40 on outer zones and the very best barely reach 25 on inner.

I believe you're wrong on the very first point, and since all the rest rest upon it... Laptop drives don't have higher platter densities than desktop drives, I wouldn't be surprised if they're actually lower in most cases, but at the very best it's same. How did you think they can stuff 3x more data on desktop drives even though physical size difference isn't all that large?

There are also several other significant factors that make desktop drives faster - they can add more platters and heads since there's more space, and read them in parallel, larger platters also lead into faster outer zone spinning speed with same RPM, and last but not least, notebook drives just haven't been optimized for performance.

Re:speed boost... but detremental power savings...

[Crazy Eight]

Your observations make me pine for drive shaped more like a skyscraper than an Altoid tin with 50 1" platters on one long spindle.

Re:speed boost... but detremental power savings...

[mp3phish]

I think that would be a great idea.. cylyndrical drives rather than wider ones... But there are cost issues (mainly that they have to make a seperate head for each platter) and the larger size head arm would mean it is slower to seek... it is all tradeoffs no matter what your drive geometry might be...

Re:speed boost... but detremental power savings...

[mp3phish]

"WRT comparing drives of the same size, that's hard when there are so few desktop hard drives available in 60 or 80 GB."

Well, I don't know about how the drive market is in Europe.. But in the USA most OEM desktops come with 80-120GB hard drives standard... Most laptops come with 60-100GB laptops standard...

Low end laptops(and desktops) come wiht 40GB drives.. higher end laptops come with 80-100GB drives... Higher end desktops come with 120-200GB drives... (this is all generalizations.... but you get my idea)

all that being said, I don't know how you can say that "so few desktop hard drives available in 60-80GB" because 40 and 80GB drives are among the most popular in consumer desktop systems... And you won't find lower than 40GB in a laptop unless it is ultra low end (sub 700$ range). 40GB is as small as a laptop -- and desktop-- drive get in the general case...

My whole origional post is a direct relation of equivilently sized hard drives.. As soon as you start throwing in 250 gig desktop drives and compare them to 80gb laptop drives, you are throwing out my initial assumptions and are making new ones... with new initial assumptions, yes, my argument doesn't hold out...

I will agree with you 100% that a higher density desktop drive wil definately outpace a lower density laptop drive in sustained sequential reads... This is not my argument...

My argument specifically compares 2 equivilently sized drives (such as 80gb v 80gb) one in a laptop and one in a desktop, at 5400RPM vs 7200RPM respectively.

Now, You can go on about how the economics of it are different.. and you may have a point to an extent... But my main point is that RPM's do not always mean faster performance... And my second main point is that RPM's do not always mean better. It depends on the circumstances.. When you are talking about ultra light laptops or midsized to smaller 14in laptops, you have to take into account battery life, heat, and power drain... In these circumstances, 7200RPM drives are definately overkill and will be significantly detremental to battery life and drive reliability (being that they aren't cooled properly) and are also un-needed considering that a 5400RPM 80gb drive performs the same or better than a 7200RPM 80GB desktop drive...

That is my argument.. and I'm sticking to it :)

Re:speed boost... but detremental power savings...

[mp3phish]

naa.. It is just a matter of benchmark methods... My 40MB/s numbers come from unbuffered sequential reads of the outter track in a SATA 7200RPM 160GB drive (that is about 3 months old)... I used sandra to do the benchmarks... I even compared my drive to other drives in the sandra database and my 7200RPM drive is marginally faster than other 160GB 7200RPM drives...

Unless you run the same exact benchmark, you will never get consistant results... You may get 60MB/s in sequential buffered reads if they are shorter bursts, etc.. I am not arguing this...

In the end, You all have points about my data being "wrong" in the circumstances you give... But in other circumstances (namely, the ones i give in my (1) paragraph with equivilently sized desktop vs laptop drives, my data is correct and conclusive. I may not be able to point to hard data on an independant reviewer's website but that is only because no reviewer to date has done a comparison spec'd out as I have in my origional post. (one day I will post hard numbers onto my website but that is for another day and time long after this discussion will be over)..

All in all, i'm not really disagreeing with your points.. I'm just saying that my points are different than what you are making, and don't conflict with what you are saying. It just so happens that the experiments you point to do not use the same initial assumptions as I have..

Wether or not you agree with my initial assumptions in point (1) in my post are another matter and can be discussed in a different way. But just for the sake of clarity:

1) drive density and RPM speed are directly proportional to sustained sequential read speed.

2) platter radious and RPM speed is directly proportional to seek time.

3) Multiple platters in the same drive read in parallel while accessing data, thus doubling the density of a single platter

4) Knowing the above, and working the numbers in the specific instances I am claiming in my origional post (namely, 5400RPM 2.5in disk vs 7200RPM 80gb disk) you can determine that the 5400RPM 2.5in disk will have faster sequential read speeds on the outter track of said disk.

5) Adding point (4) above with the fact that 5400RPM laptop drives are more reliable, generate less heat, use up less battery power, and have a longer MTBF (due to heat issues in plastic chassis and due to bumping the drive while being accessed is more detremental on a 7200RPM drive) and the fact that this is a discussion on the AMD Turion processor (which is ment to go into thinner, lighter, and more portable and longer battery life notebooks, and My claims about 5400RPM drives being better in this scenerio stand in my professional opinion. (I won't give my credentials because I don't feel this discussion should be about me justifying my knowledge of the subject).
You don't have to trust me, you just have to take what I claim and understand it.. And if you don't agree you can only run the tests yourself and come to your own conclusions.. But pointing to a website which doesn't run the tests as I described will not get you anywhere in this discussion.

Re:speed boost... but detremental power savings...

[mp3phish]

"A two platter disk does not have a higher transfer rate then a single platter disk! "

In all modern drives that I am aware of, yes they do. Multi platter disks read from both platters at once in parallel fashion (you might call it "striped" fashion...

There may be some cases where you can't read from both platters at once, such as when the drives S.M.A.R.T logic moves data to different sections of one platter because of known bad sectors. But in the general case of a new drive (or newly formatted and defrag'd drive) you can normally assume that the drive reads both platters in parallel when accessing one single large file.

If you don't believe me run the benchmarks on 2 drives from the same MFG where one drive is say.. 80GB and the other is 160.. (but watch out and make sure that they use the same platters but 2 of them in the larger one.. Sorry I can't name off any model#'s right now cos i don't have the matrices handy) and you will see that the 160 drive will have a faster sustained sequential read speed.

Re:speed boost... but detremental power savings...

[moonbender]

Along the lines of your argument, a 30 GB 3.5in 7200 rpm drive should be a lot slower than an 80 GB 2.5in 5400 rpm drive, due to the difference in density. Well, the review I was referring to did not test a 30 GB 7200 rpm drive, but I did quote the numbers for a 10,000 rpm drive, and despite the capacity difference, it was almost three times as fast. Figuring in the advantage of a higher rpm, at 7200 rpm it probably would still be twice as fast. Sorry, I just don't see your argument holding water given those numbers. It also goes against my daily experience. That said I obviously agree with your other points, like I've said in my first post to this thread, I wouldn't put one of those mad 10krpm drives in my subnotebook, either.

Re:speed boost... but detremental power savings...

[juhaz]

naa.. It is just a matter of benchmark methods...

No. It's not. It's a matter of your number being totally different than those of anyone else with any methods. Are you sure you're reading outer tracks? You know that the "logical" beginning of disk is in outer tracks, not inner?

My 40MB/s numbers come from unbuffered sequential reads of the outter track in a SATA 7200RPM 160GB drive (that is about 3 months old)... I used sandra to do the benchmarks...

Sandra is not a good benchmark, but if that's what you use, let's try the same to get "matter of benchmark methods" dealt with. As for the database, Sandra isn't capable of testing zones of a drive, only partitions, given that the db doesn't mention anything except the drive, we can't know ANYTHING about what zone(s) those readings are for.

My 40MB/s numbers come from unbuffered sequential reads of the outter track in a SATA 7200RPM 160GB drive (that is about 3 months old)...

My number come from unbuffered sequential reads of the outer track in a ATA 7200RPM 120GB drive (that is about 16 months old). AVG read speeds across all zones 49MB/s. Hitachi 7K250, 80GB/platter, 2 platters. Just like yours, just older, and much faster.

39MB/s outer zones, 35.8MB/s average on some years old Maxtor D740DX (7200RPM), 60GB, 40GB platters.

I've only got one partition on my laptop so Sandra can't tell anything about it for specific zones, but the speeds of the two other discs have mirrored HD tach results very closely so this one probably will to. 30-32MB/s outer 26.3MB/s average. Samsung Spinpoint M, about two months old, 80GB, 40GB platters, 5400 RPM.

Transfer graphs for all three drives http://www.cc.puv.fi/~e0000274/hdtach.png here, just in case you're interested.

You may get 60MB/s in sequential buffered reads if they are shorter bursts, etc..

I'm not stupid, sequential unbuffered is what was talked about and sequential unbuffered is what I give you. 57MB/s. Sequential. Unbuffered. Sandra. Got it? (Buffered around ~90M/s, btw)

In the end, You all have points about my data being "wrong" in the circumstances you give...

No. Not just in the circumstances I give, _any_ circumstances. 2.5" drives, whether similar size or not, are slower than desktop drives.

namely, the ones i give in my (1) paragraph with equivilently sized desktop vs laptop drives, my data is correct and conclusive.

Unfortunately your data is in question, and even if it's correct (which I doubt) it certainly doesn't seem to represent vast majority of drives out there, and you haven't provided data for a notebook drive that is faster than even that relatively slow desktop drive. That's not conclusive.

1) drive density and RPM speed are directly proportional to sustained sequential read speed.

Agreed.

2) platter radious and RPM speed is directly proportional to seek time.

Agreed.

3) Multiple platters in the same drive read in parallel while accessing data, thus doubling the density of a single platter

False. Reading data from multiple platters may yield a small speed boost, but it's nowhere as large as actual doubling of single platter density would be. For example, 250GB drive with three 80GB platters performs almost identifically to my 120GB version with 1½.

4) Knowing the above, and working the numbers in the specific instances I am claiming in my origional post (namely, 5400RPM 2.5in disk vs 7200RPM 80gb disk) you can determine that the 5400RPM 2.5in disk will have faster sequential read speeds on the outter track of said disk.

And this is where we disagree. Based on overwhelming evidence of all benchmarks ever published and my own testing, 5400RPM 2.5" disk can't beat even several generations older 7200RPM 3.5" disk, can do absolutely NOTHING against current generation 7200RPM

Lilo & Stitch

[spoggle]

Obviously this will power the computers of the Galactic Council.

Thats not their job

[m00j]

What do you want them to do? Tell you that AMD is the greatest thing since sliced bread and you would be an idiot for buying Intel when you can get AMD cheaper? The salespersons job is *not* to give you advice and help you find what it is that suits you best; it is to give you advice on which of the products they sell would suit you best.

I work (part time while at uni) in retail and while I will try to sell the AMD computers (when we get stock of them, which is almost never) I primarily have to convince people to buy Intel since that is all we have available.

I would also love to point out to them that they if they shopped online they could get a no-brand machine with much higher specs for the same price but that's not my job. My main computer is better than all the ones we have on display yet it was cheaper than half of them all because I built it myself. 90% of the customers I see could also get away with the cheapest computers (why would you need more than a 2.8 GHz Celeron for the internet and word?) yet it is our job to sell them a more expensive box.

Morally I don't like to be the sleazy salesperson type - if it is an old granny wanting to get her emails from her children I won't try to sell her a 3Ghz P4 Gaming machine. I try to get them to spend the money on things that will be of better use to them, like a decent printer rather than the cheapest Lexmark they can find - The kind that comes with only a colour cartridge, making you shell out the cost of the printer again straight away for black.

In your case though it sounds like he was trying to unload some old stock - there was probably a bonus for him if he unloaded it.

Sigh....

[Chanc_Gorkon]

CENTRINO IS NOT A PROCESSOR! The Pentium-M is the processor and Centrino is the WHOLE package.....chipset, Wireless-G and other parts make whe whole of a Centrino laptop. Everyone may refer to Centrino as a processor, but that does not mean it's right!

When in Rome. . .

[Ohreally_factor]

Centrino + Turion = Centurion

Throne room

[Ohreally_factor]

No, the real benefit of a laptop is being able to read slashdot while I'm taking a crap. 802.11 is nice, but if there was no such thing, I would have installed a jack next to the toilet.

Answer

[turgid]

The virtual addressable space will still be 4GB, however the 36-bit address bus will allow for a theoretical limit of 64GB.

That's true, however it's a dirty hack. The extra memory is only addressable in 4GB (or 2GB) segments. Therefore, if you have an application that needs more than 4GB in a contiguous chunk, you are out of luck. You could probably fake it with operating system calls and complicated wrpper functions to hide the memory address arithmetic. However, you loose a lot of performance.

It's a bit like the old days of the 8086 and 80286 where memory was in 64k segments and there were segment registers used in conjunction with index registers to calculate addresses. As you can imagine, writing programs with datastructures larger than 64k was complicated and bug-prone due to the added complexity. It also slowed the program down significantly due to the extra calculations required.

So you see, a much better, cleaner, more efficient solution to the problem is to have a flat 64-bit address space, like the AMD64 architecture.

Of course, intel tried very hard to pretend that no-one "needs" 64-bits on the desktop for many years while they tried to peddle the dreadful itanium for servers and workstations. They hoped that naieve users would believe them long enough until they could get itanium PCs out in the mass market. Then along came AMD with a much better processor...

Re:Answer

[andreyw]

Not sure where you're getting the 2GB. Setting the base of a GDT descriptor to 0, limit to 0xFFFFF, and granularity to page-wide yields a 4GB segment. YOu might be trying to consider things like.... memory-mapped devices or something... but that has nothing to do with the 4GB addressability.

In 80286 days, the bus was 24-bits wide, with the maximum addressable space being 16MB. There was not concept of "page-wide" granularity... so segments defined by GDT/LDT descriptors could have a maximum size of 64kb, correct. However, this was flat-out 16-bit PMODE, with priviledge levels et al. Very much different from real 8086 mode with its 'tarded address calculation - the segment registers contained selectors into the GDT/LDT.

Acer Ferrari Rules

[turgid]

It runs 64-bit Solaris 10 very well indeed.

hm...

[Mihai331]

For a few years now (since first Athlon processor) AMD is much stronger than Intel (in cpu brain cells count) with lower prices, frequencies and less cache, lower FSB and power consumption, and for a strange reason people keep buying Intel in 95 cases in 100. I hope AMD will have a greater share of the market in the years to come. I have a great respect for AMD, with such a low budget for research and publicity (because with money you can sell gastric juice as coke and have a profit) they managed such remarkable results. AMD 64 is as strong as 2 intel processors (HT).. all benchmarks show that! :o)) and my 1.6 GH AMD Duron processor with Asus motherboard kicks the s#&t out of P4 2.4 GH.

Alan Turion

[Anonymous Coward]

Was Alan Turion the gay cryptologist who invented the internet during nazi occupied Britan? Is that who it's named after?

Mod parent up "Interesting"

[pgilman]

AMD's making the "(Cen)Turion" to compete with Intel's "Centrino?" God bless AMD, I love 'em, but man, they need a better marketing department...

Yes smarty pants I know...

[turgid]

Not sure where you're getting the 2GB.

Depending on the OS being used, the amount of memory space useable by an application varies between 3.5GB and 2GB on the x86 architecture.

I realise that the 80286 and 8086 had very different addressing modes. The 286 had a very primitive form of protected mode, but my point was about the 64k segement size and how inconvenient it was to the programmer, hurting program reliability and performance due to increased complexity.

Further more, my point was that with the segmented 36-bit addressing of the Pentium II and later proecssors, intel has, in effect, brought back a segmented memory hach with many of the drawbacks and failings of the old memory models on the processors of 25 years ago

Compare this with all the processors of the last 15 years which have 64-bit address registers (some only capable of physically addressing 40 bits of memory) which engineer out these problems in a simple and elegant way.

The reason intel didn't extend the x86 architecture in this manner earlier (c.f. AMD64) was to artificially restrict the market of the Pentium range to try to persuade people to buy itanic.

Now, we could go on to the relative merits of extending the x86 architecture to 64-bits or starting with a clean slate (c.f. Alpha, MIPS, PA-RISC, UltraSPARC), however, the instruction set wars are over, as AND has demonstrated with the AMD64 architecture (Opteron etc.) getting all the performance of the big RISC chips with effectively a 64-bit RISC with an x86 ISA translation layer on chip (yes, I know Pentium II, K6 and Cyrix M1 started this off years ago, but Opteron was the first 64-bit one).

Re:Too much embedded crap

[Dashing Leech]

"...maybe it's time to regain some of that currently wasted heat back into electricity"

Good idea, I've always thought that it seemed like a waste of energy, and I know people are working on it. The problem is really that a system that uses the waste heat to generate energy (thereby reducing both heat output and power requirements) is generally big, heavy, and not too portable. I know there are small versions, but I'm guessing they haven't gotten them to the point where they are practical or economic for laptops. Someone can't point me to a reference if I'm wrong about this. Still, I expect it isn't too far in the future.