Despite what many other commenters will say, no, it isn't a power hog compared to ARM. Or at least it doesn't have to be. Intel/AMD/VIA don't yet offer processors that have as low power as ARM (although some are pretty power/performance efficient depending on your workload), but they will within the next year for smartphones and tablets. On modern manufacturing processes the "x86 tax" becomes almost non-existant.
Atom has 64-bit support built in, and Intel disables it on most parts to a) to provide market segmentation and b) to save a bit of power. If MS had made Win7 64-bit only, Intel simply would have flipped the switch and sold the Atoms as 64-bit.
Yes, you could have stopped after the word "think".
We have and use VNC. It's supported and we depend on it. I did install and enable unsupported VNC clients like TightVNC to try to get some more speed. NX is the next step. You can argue whether or not users should be installing potentially insecure networking servers, and you can argue about productivity.
When working remotely everything is over VPN anyway.
I work for a large corporation that uses VNC,and several years ago I tried to install NX at work, hoping to get a speed boost when working remotely. Unfortunately, the creation of a user "nx" was required. I'm not in the IT department, I don't have root access, and they IT department had no interest in deploying NX. So I gave up.
I saw this announcement and hoped that an "nx" user would no longer be required, but it appears this is still necessary. If I could get it installed and it actually worked better I know the other engineers would jump on it and eventually IT would be forced to support it.
Completely disagree. A CompE master's gets you nothing except a bit of a pay increase (but it will take years before the increase makes up for the year(s) you could have been working with a BS), and many many microprocessor designers have a BS.
But this ISN'T that old, craptastic, power-hungry chipset used by most Atom netbooks. It's a new chipset code-named Poulsbo designed specifically to go with Atom. Quoting a tomshardware.com article:
"The Atom Z500 has a TDP that varies between 0.85 W (for the 800 MHz version without HyperThreading) and 2.64 W (for the 1.86 GHz model with HyperThreading enabled). The SCH consumes approximately 2.3 W in its most evolved version, which brings the SCH + CPU together to under 5 W. By comparison with existing solutions, thatâ(TM)s obviously a big step forward â" the Via Nano, for example, is announced at 25 W for the 1.8 GHz version and a Celeron-M ULV at 5 W at 900 MHz."
In addition, the Atom Z-series/Poulsbo combo supports the C6 idle power state where the CPU saves away its architectural state in a small SRAM which remains powered up while the rest of the CPU shuts off entirely. Idle power for the processor is somewhere from.01W -.1 W (this is from what I remember reading somewhere, but I can't find a link right now). Not sure what the chipset's powercomsumption is like when idle.
The biggest known downside to this chipset is that it supports 1 GB of RAM max.
From what I've read elsewhere, the chipset involved does have video decode acceleration support. After googling, anandtech.com has an article that says that the chipset can support 1080i and 720p decode. A tomshardware.com article says that it can do hardware decode of H.264, MPEG2, MPEG4, VC1, and WMV9 formats.
>I'm suspicious of the suggestion that a log-based >filesystem will cure all the ills of the limited flash- >controller based wear leveling.
Yeah. Total bull.
Anybody who thinks the filesystem can do really well has bought into the crud from most existing vendors about how you have to use those things differently. If you really do believe that, you shouldn't touch an SSD with a ten-foot pole.
If the flash vendor talks about "limits" in the wear levelling, and how you have to write certain ways, just start running away. Don't walk. Run away as fast as you can.
>A question keeps coming up in my mind about what happens >when you split an SSD into multiple partitions, and what >*you want to happen*. I use separate partitions for root, >boot, and var, because I tend to make root and boot >read-only.
Again, if your SSD vendor says "align to 64kB boundaries" or anything like that, you really should tell them to go away, and you should do what Val said - just get a real disk instead. Let them peddle their crap to people who are stupider than you, but don't buy their SSD.
So what you want to happen if you split an SSD into multiple partitions is exactly nothing. It shouldn't matter one whit. If it does, the SSD is not worth buying. If it is so sensitive to access patterns that you can't reasonably write your data where you want to, just say "No, thank you".
Anyway, I have a good SSD now, so I can actually give some data: - Most flash-based SSD's currently suck.
I don't have these ones myself, but last week we had the
yearly kernel summit here in Portland, and a flash
company that shall remain nameless (but is one of the
absolute biggest and most recognizable names in flash)
was selling their snake-oil about how you need to write
in certain patterns.
So I called them on it, and called them idiots. Probably
one reason why I didn't get one of the drives they were
handing out, but one of the people who did get a drive
was the Linux block system maintainer. So he ran some
benchmarks.
Those things suck. You will never get any decent
performance of anything but a very specialized filesystem
out of them, unless you use them as essentially read-only
devices.
For a basic 4kB blocksize random write test, the SSD got
around 10 IOps. That's ten, as in "How many fingers do
you have?" or as in "That's really pathetic". It means
that you cannot actually use it as a disk at all, and
you need some special filesystem to make it worthwhile,
and certainly means that wear levelling is probably not
working right.
(For the math-challenged, 10 IOps at a 4kB blocksize
means 40kB/s throughput and 100ms+ latencies for those
things. It also means that even if some operations are
fast, you can never trust the drive)
- In contrast, the Intel SSD's are performing exactly as
advertised.
I did get one of these, with warnings about how
if I want to get low-power operation etc I need to make
sure that disk-initiated power management is enabled etc.
Whatever. The important thing is that the Intel SSD does
not care one whit where you write stuff, or how you do
it. With the same 4kB random write b
I agree. The "performance" number he's quoting are complete crap.
"IBM's "Broadway" CPU is clocked at 729MHz, according to updated Nintendo documentation. By comparison, GameCube's Gekko CPU ran at 485MHz. The original Xbox's CPU was clocked at 733MHz. Meanwhile, Xbox 360 runs three symmetrical cores at 3.2GHz."
and
"Revolution's ATI-provided "Hollywood" GPU clocks in at 243MHz. By comparison, GameCube's GPU ran at 162MHz, while the GPU on the original Xbox was clocked at 233MHz."
and
"Clearly, numbers don't mean everything, but on paper Revolution's CPU falls performance-wise somewhere well beyond GameCube and just shy of the original Xbox."
THE MAJORITY OF HIS COMPARISON IS BASED ON CLOCK SPEED. Yet he's comparing completely different architectures. Gamecube had an IBM flipper chip (some sort of Power-based core), Xbox had an x86, X360 has 3 simplified Power-based cores running at high clock speeds. Gamecube had an ATI graphics, XBox had NVIDIA graphics. You can't just throw random MHz numbers out there and draw any type of conclusion. Ok, I suppose there's one type: an invalid one.
Completely Backwards
on
Brute Force
·
· Score: 2, Informative
Intel's specs can't come anywhere close to AMD in ANYTHING. AMD has Intel beat in: price, performance, heat, power, maximum operating temperatures, and (lower) heat output.
Not true. The opteron's pricing is similar to the xeons (depends which model you're comparing to (mp, dp, regular)). The performance benchmarks go back and forth, with the opteron generally leading in multiprocessor configurations. The power dissipation is a flat out win for opteron.
Not like your enterprise is going to care which company is making more money.
Companies will care. They need suppliers that have the cash to remain around to give support and take responsibility for problems. Sure x86-64, wave of the future... unless it dies with AMD a year from now.
Intel can't take enough of a loss for a long period of time to be competitive with AMD.
Not sure what this means exactly, but Intel so far isn't losing anything. They're still profitable (in fact, I've read that they've been profitable every single quarter since they went public 20-30 years ago). And for many quarters in a row AMD has been losing money.
The hammer chips are nice, but so far they're don't kill Intel in performance, they're just competitive. And while they have potential to do more, it hasn't been realized yet except in certain niche markets.
As you mention, there's more than performance (like power consumption), but it's still all wait and see... what will Intel bring out next, how well will AMD be able to supply chips and rally support?
From Page 4 "Because of objectives and subject, paper surveys sent by mail were used to avoid built-in sample bias from internet-based study"
From Page 6 Age mix - 35% Age 55+ - 20% Age 45 to 54 - 21% Age 35 to 44 - 24% Age 34 and under
Gender blend - 38% Male, 62% Female
It looks like the ended up with a bias in the sample anyway. 55% over 55 years old, 62% female... I think it was already understood that technology confuses them.
Actually, I believe Carmack has said that the true baseline for Doom 3 is the original GeForce, not the GeForce3. In other words, it should be playable on one if you turn down the resolution and a lot of the effects.
I'm pretty sure that this is one of those quotes that's been debunked as not being true. Unfortuantely, so many people quote it that when I search I can't find any pages refuting it. Of course, no guarantee they'd be correct anyway.
Your comments are misleading. The extra registers CAN improve performance when code is compiled to use them, but will not necessarily. AMD has previously quoted numbers that the expect to get about 15% additional performance with code that used the registers over code that does not. However, we do not know which apps they're using...
Also, I believe that interview you mention does talk about a 30% increase in performance, but it does NOT say it is from the registers (I can't find the link to the interview). It's not very clear what they were comparing to, but if I remeber correctly it looked like they were comparing to a regular Athon. Thus, the 30% increase would be coming from a) a new core with micro-architectural enhancements b) onboard memory controller c) extra registers.
I've seen a lot of comments here asking why Intel would do such a thing, why they're trying to prevent overclocking even though it voids the warrantee.
They really aren't concerned so much with enthusiasts... the percentage of people who over clock in the total PC market is very small (they just speak loudly online).
The problem they have is with resellers (ie whitebox shops) taking a slow processor (say a P4 2.0 GHz), overclocking it, and selling it in a system as, say, a P4 2.8 GHz and marking up the price as such. To clarify, these resellers do not tell their customers the system has a P4 2.0 overclocked to 2.8 GHz and that the warrantee is voided, they say it has a P4 2.8 GHz part in it, and pocket the extra cash. So Intel loses money on sales of its higher end parts, and customers aren't getting what they paid for: they end up with an overclocked part that may or may not be completely stable.
They can't market the Itanium 1 successfully at 800 MHz, even if it compares with a 2 GHz chip because of the perceived differential. The Itanium 2 fares better, but it's still a power hog. The companies that focuses on a balance between clock-cycles and efficient design are the only winners (namely IBM) because their chips have a wider application. You won't see an Itanium 2 in a laptop, but you might see a PPC 970.
You couldn't possibly have a worse understanding of the markets involved here. Itanium is targeted at technical computing workstations and massively parallel processing supercomputers. The people buying these things know exactly what they're looking for, they're not Joe Consumer "tricked" by MHz over what constitutes actual performance.
I can't believe so many posters here believe that pushing MHz in the Desktop space troubles Intel in the high end space where clock speeds are lower. It doesn't. People in the desktop space buy on MHz, people in the high end space buy on performance, reliability, scalability, and more (not necessariliy in that order either). Power usually isn't a concern (it's accepted that a costly cooling solution will be necessary).
By the way, the reason the Itanium 1 has problems is because its performance is not good. The Itanium 2 is much, much better. Get a clue.
I think a lot of people are misreading (or in/. fashion, not reading) this. Or just not understanding. This does not mean IBM and AMD are working on some new type of processor together (ie no new processor architecture).
What this means is that they will work together on having manufacting technoligies in the future. Fabs and fab equptment are extremely expensive and it is generally hard to move from one manufacturing process to another. This alliance should help shave costs and improve manufacturing quality on the process (I believe it said 0.65 micron) in question. Each will continue to design cpus separately.
Slashdot Mirror
Despite what many other commenters will say, no, it isn't a power hog compared to ARM. Or at least it doesn't have to be. Intel/AMD/VIA don't yet offer processors that have as low power as ARM (although some are pretty power/performance efficient depending on your workload), but they will within the next year for smartphones and tablets. On modern manufacturing processes the "x86 tax" becomes almost non-existant.
Atom has 64-bit support built in, and Intel disables it on most parts to a) to provide market segmentation and b) to save a bit of power. If MS had made Win7 64-bit only, Intel simply would have flipped the switch and sold the Atoms as 64-bit.
Thank you, I will check this out.
Yes, you could have stopped after the word "think".
We have and use VNC. It's supported and we depend on it. I did install and enable unsupported VNC clients like TightVNC to try to get some more speed. NX is the next step. You can argue whether or not users should be installing potentially insecure networking servers, and you can argue about productivity.
When working remotely everything is over VPN anyway.
I work for a large corporation that uses VNC,and several years ago I tried to install NX at work, hoping to get a speed boost when working remotely. Unfortunately, the creation of a user "nx" was required. I'm not in the IT department, I don't have root access, and they IT department had no interest in deploying NX. So I gave up.
I saw this announcement and hoped that an "nx" user would no longer be required, but it appears this is still necessary. If I could get it installed and it actually worked better I know the other engineers would jump on it and eventually IT would be forced to support it.
Anyone have a workaround?
Completely disagree. A CompE master's gets you nothing except a bit of a pay increase (but it will take years before the increase makes up for the year(s) you could have been working with a BS), and many many microprocessor designers have a BS.
2. Never use, or even think of, the word "goddess" in reference to yourself or other women.
But this ISN'T that old, craptastic, power-hungry chipset used by most Atom netbooks. It's a new chipset code-named Poulsbo designed specifically to go with Atom. Quoting a tomshardware.com article:
"The Atom Z500 has a TDP that varies between 0.85 W (for the 800 MHz version without HyperThreading) and 2.64 W (for the 1.86 GHz model with HyperThreading enabled). The SCH consumes approximately 2.3 W in its most evolved version, which brings the SCH + CPU together to under 5 W. By comparison with existing solutions, thatâ(TM)s obviously a big step forward â" the Via Nano, for example, is announced at 25 W for the 1.8 GHz version and a Celeron-M ULV at 5 W at 900 MHz."
http://www.tomshardware.com/reviews/intel-atom-cpu,1947-3.html
In addition, the Atom Z-series/Poulsbo combo supports the C6 idle power state where the CPU saves away its architectural state in a small SRAM which remains powered up while the rest of the CPU shuts off entirely. Idle power for the processor is somewhere from .01W - .1 W (this is from what I remember reading somewhere, but I can't find a link right now). Not sure what the chipset's powercomsumption is like when idle.
The biggest known downside to this chipset is that it supports 1 GB of RAM max.
From what I've read elsewhere, the chipset involved does have video decode acceleration support. After googling, anandtech.com has an article that says that the chipset can support 1080i and 720p decode. A tomshardware.com article says that it can do hardware decode of H.264, MPEG2, MPEG4, VC1, and WMV9 formats.
>I'm suspicious of the suggestion that a log-based
>filesystem will cure all the ills of the limited flash-
>controller based wear leveling.
Yeah. Total bull.
Anybody who thinks the filesystem can do really well has
bought into the crud from most existing vendors about how
you have to use those things differently. If you really
do believe that, you shouldn't touch an SSD with a ten-foot
pole.
If the flash vendor talks about "limits" in the wear
levelling, and how you have to write certain ways, just
start running away. Don't walk. Run away as fast as you
can.
>A question keeps coming up in my mind about what happens
>when you split an SSD into multiple partitions, and what
>*you want to happen*. I use separate partitions for root,
>boot, and var, because I tend to make root and boot
>read-only.
Again, if your SSD vendor says "align to 64kB boundaries"
or anything like that, you really should tell them to go
away, and you should do what Val said - just get a real
disk instead. Let them peddle their crap to people who are
stupider than you, but don't buy their SSD.
So what you want to happen if you split an SSD into multiple
partitions is exactly nothing. It shouldn't matter
one whit. If it does, the SSD is not worth buying. If it is
so sensitive to access patterns that you can't reasonably
write your data where you want to, just say "No, thank you".
Anyway, I have a good SSD now, so I can actually
give some data:
- Most flash-based SSD's currently suck.
I don't have these ones myself, but last week we had the
yearly kernel summit here in Portland, and a flash
company that shall remain nameless (but is one of the
absolute biggest and most recognizable names in flash)
was selling their snake-oil about how you need to write
in certain patterns.
So I called them on it, and called them idiots. Probably
one reason why I didn't get one of the drives they were
handing out, but one of the people who did get a drive
was the Linux block system maintainer. So he ran some
benchmarks.
Those things suck. You will never get any decent
performance of anything but a very specialized filesystem
out of them, unless you use them as essentially read-only
devices.
For a basic 4kB blocksize random write test, the SSD got
around 10 IOps. That's ten, as in "How many fingers do
you have?" or as in "That's really pathetic". It means
that you cannot actually use it as a disk at all, and
you need some special filesystem to make it worthwhile,
and certainly means that wear levelling is probably not
working right.
(For the math-challenged, 10 IOps at a 4kB blocksize
means 40kB/s throughput and 100ms+ latencies for those
things. It also means that even if some operations are
fast, you can never trust the drive)
- In contrast, the Intel SSD's are performing exactly as
advertised.
I did get one of these, with warnings about how
if I want to get low-power operation etc I need to make
sure that disk-initiated power management is enabled etc.
Whatever. The important thing is that the Intel SSD does
not care one whit where you write stuff, or how you do
it. With the same 4kB random write b
I agree. The "performance" number he's quoting are complete crap.
"IBM's "Broadway" CPU is clocked at 729MHz, according to updated Nintendo documentation. By comparison, GameCube's Gekko CPU ran at 485MHz. The original Xbox's CPU was clocked at 733MHz. Meanwhile, Xbox 360 runs three symmetrical cores at 3.2GHz."
and
"Revolution's ATI-provided "Hollywood" GPU clocks in at 243MHz. By comparison, GameCube's GPU ran at 162MHz, while the GPU on the original Xbox was clocked at 233MHz."
and
"Clearly, numbers don't mean everything, but on paper Revolution's CPU falls performance-wise somewhere well beyond GameCube and just shy of the original Xbox."
THE MAJORITY OF HIS COMPARISON IS BASED ON CLOCK SPEED. Yet he's comparing completely different architectures. Gamecube had an IBM flipper chip (some sort of Power-based core), Xbox had an x86, X360 has 3 simplified Power-based cores running at high clock speeds. Gamecube had an ATI graphics, XBox had NVIDIA graphics. You can't just throw random MHz numbers out there and draw any type of conclusion. Ok, I suppose there's one type: an invalid one.
Actually, SETI@Home runs like http://www.distributed.net./
Check out this post at Ars Technica.
Basically, there's a clause in there that states allows a certain part of the bill to be exempt from judicial review.
Just an FYI on some of the who did what first arguments here.
The guy who started The Register (Mike Magee) left and started The Inquirer. So, it's not really a wanna-be...
Intel's specs can't come anywhere close to AMD in ANYTHING. AMD has Intel beat in: price, performance, heat, power, maximum operating temperatures, and (lower) heat output.
Not true. The opteron's pricing is similar to the xeons (depends which model you're comparing to (mp, dp, regular)). The performance benchmarks go back and forth, with the opteron generally leading in multiprocessor configurations. The power dissipation is a flat out win for opteron.
Not like your enterprise is going to care which company is making more money.
Companies will care. They need suppliers that have the cash to remain around to give support and take responsibility for problems. Sure x86-64, wave of the future... unless it dies with AMD a year from now.
Intel can't take enough of a loss for a long period of time to be competitive with AMD.
Not sure what this means exactly, but Intel so far isn't losing anything. They're still profitable (in fact, I've read that they've been profitable every single quarter since they went public 20-30 years ago). And for many quarters in a row AMD has been losing money.
The hammer chips are nice, but so far they're don't kill Intel in performance, they're just competitive. And while they have potential to do more, it hasn't been realized yet except in certain niche markets.
As you mention, there's more than performance (like power consumption), but it's still all wait and see... what will Intel bring out next, how well will AMD be able to supply chips and rally support?
The bottom line should say 55% over 45 years old, not 55 years old.
The PDF of the survey can be found here.
From Page 4
"Because of objectives and
subject, paper surveys sent
by mail were used to avoid
built-in sample bias from
internet-based study"
From Page 6
Age mix
- 35% Age 55+
- 20% Age 45 to 54
- 21% Age 35 to 44
- 24% Age 34 and under
Gender blend
- 38% Male, 62% Female
It looks like the ended up with a bias in the sample anyway. 55% over 55 years old, 62% female... I think it was already understood that technology confuses them.
Actually, I believe Carmack has said that the true baseline for Doom 3 is the original GeForce, not the GeForce3. In other words, it should be playable on one if you turn down the resolution and a lot of the effects.
I'm pretty sure that this is one of those quotes that's been debunked as not being true. Unfortuantely, so many people quote it that when I search I can't find any pages refuting it. Of course, no guarantee they'd be correct anyway.
Your comments are misleading. The extra registers CAN improve performance when code is compiled to use them, but will not necessarily. AMD has previously quoted numbers that the expect to get about 15% additional performance with code that used the registers over code that does not. However, we do not know which apps they're using...
Also, I believe that interview you mention does talk about a 30% increase in performance, but it does NOT say it is from the registers (I can't find the link to the interview). It's not very clear what they were comparing to, but if I remeber correctly it looked like they were comparing to a regular Athon. Thus, the 30% increase would be coming from a) a new core with micro-architectural enhancements b) onboard memory controller c) extra registers.
I've seen a lot of comments here asking why Intel would do such a thing, why they're trying to prevent overclocking even though it voids the warrantee.
They really aren't concerned so much with enthusiasts... the percentage of people who over clock in the total PC market is very small (they just speak loudly online).
The problem they have is with resellers (ie whitebox shops) taking a slow processor (say a P4 2.0 GHz), overclocking it, and selling it in a system as, say, a P4 2.8 GHz and marking up the price as such. To clarify, these resellers do not tell their customers the system has a P4 2.0 overclocked to 2.8 GHz and that the warrantee is voided, they say it has a P4 2.8 GHz part in it, and pocket the extra cash. So Intel loses money on sales of its higher end parts, and customers aren't getting what they paid for: they end up with an overclocked part that may or may not be completely stable.
I think a lot of people are misreading (or in /. fashion, not reading) this. Or just not understanding. This does not mean IBM and AMD are working on some new type of processor together (ie no new processor architecture).
What this means is that they will work together on having manufacting technoligies in the future. Fabs and fab equptment are extremely expensive and it is generally hard to move from one manufacturing process to another. This alliance should help shave costs and improve manufacturing quality on the process (I believe it said 0.65 micron) in question. Each will continue to design cpus separately.
150 Mb/sec seems to imply megabits. SerialATA can transer 150 megabytes/sec.