Slashdot Mirror


Intel Breathes New Life Into Pentium

angry tapir writes "Intel is giving new life to its Pentium processor for servers, and has started shipping the new Pentium 350 chip for low-end servers. The dual-core processor operates at a clock speed of 1.2GHz and has 3MB of cache. Like many server chips, the Pentium 350 lacks features such as integrated graphics, which are on most of Intel's laptop and desktop processors."

207 comments

  1. This actually makes sense by jonwil · · Score: 3, Interesting

    A chip like this would work good for servers that are limited more by network bandwidth and disk IO than by CPU load.

    1. Re:This actually makes sense by Ethanol-fueled · · Score: 1, Interesting

      What they would be good for is trying to make money from the inventory of chips that were too defective to be part of a first-pass yield and sold with a more expensive name. I wonder if they were already written off as losses for tax purposes?

      *cough*

    2. Re:This actually makes sense by zwarte+piet · · Score: 1

      I still don't understand why a windows computer is used for fileserver purposes. Something embedded with for example an ARM processor on it sounds must more robust and energy efficient to me.

    3. Re:This actually makes sense by realityimpaired · · Score: 5, Informative

      Actually, a server is one of the few places I actually want an integrated graphics chip, because it means I don't have to take up a PCI slot and associated energy and I/O load with a low-end graphics card just to provide a console, not to mention the space considerations and form factor requirements to be able to put in a PCI card. While I like that it's not built into the CPU (freeing up those transistors for, you know, CPU things), I'd still be buying a motherboard that has built in cheap graphics.

      If you're in a situation where network and disk bandwidth is the limiting factor, then why wouldn't you put in a ULV celeron chip? My laptop has a 1.2GHz dual core with 2MB of cache, and a TDP of 18W, and while that is 3W higher than the processor in TFA, that's also including the graphics card, which this one isn't. And failing that, try putting in an Atom... I have built Atom-powered fileservers before, and they run very well: even with an Atom, the limiting factor is disk I/O, not CPU power for a fileserver.

    4. Re:This actually makes sense by Jeff+DeMaagd · · Score: 1

      Who says it has to be running Windows?

      While you can get ARM processor might be nice, but the few that I've used also had pretty cut-down and altered operating systems at best, I'd rather run an OS configuration I'm more familiar with.

    5. Re:This actually makes sense by egamma · · Score: 1

      I still don't understand why a windows computer is used for fileserver purposes. Something embedded with for example an ARM processor on it sounds must more robust and energy efficient to me.

      Sometimes you want features. I suspect Linux has equivalents to these, but maybe not quite as easy to set up:

      DFSR multi-master file replication

      BranchCache for file caching either on your Win7 machine or distributed Win2k8R2 servers

    6. Re:This actually makes sense by Anonymous Coward · · Score: 0

      Pedantry and insightfulness are not the same thing.

    7. Re:This actually makes sense by Calos · · Score: 0

      Agh, posting to undo moderation, wrong post.

      --
      I vote based on politicians' actions, unless contrary to my preconceptions. Often wrong, never uncertain. #iamthe99%
    8. Re:This actually makes sense by antifoidulus · · Score: 1

      But pedantry and incitefullness are :P

    9. Re:This actually makes sense by Anonymous Coward · · Score: 0

      Yes, Linux does have DFSR. It's called rsync and has been around since 1996...

    10. Re:This actually makes sense by dfghjk · · Score: 1

      Because integrated and PCI slot are the only two choices. ;)

    11. Re:This actually makes sense by swalve · · Score: 1

      Because that's what your software runs on?

    12. Re:This actually makes sense by colsandurz45 · · Score: 1

      Actually, a server is one of the few places I actually want an integrated graphics chip

      You don't need graphics on a server, just use ssh

    13. Re:This actually makes sense by Joce640k · · Score: 1

      I still don't understand why a windows computer is used for fileserver purposes.

      Because most servers are serving to Windows desktop machines. Even for pure 'files' that still means lots of Microsoft proprietary protocols.

      Plus most people also want Exchange, Access, etc.

      --
      No sig today...
    14. Re:This actually makes sense by Joce640k · · Score: 1

      You can argue the cleverness of it all day but Windows servers don't work without graphics. AFAIK they don't even boot.

      Given that, you want the graphics with the lowest power consumption.

      --
      No sig today...
    15. Re:This actually makes sense by LordLimecat · · Score: 1

      It doesnt have CPU graphics, but whether or not it has integrated graphics is up to the motherboard manufacturer. I have been looking at a number of server builds with various Xeon chips which dont have the Intel HD graphics, and a lot of them have Nuvoton graphics chips with a paltry amount of video RAM.

      It boils down to, do you really want Sandy Bridge graphics chewing up an extra 10w of electricity in your CPU when you could just use a much more modest chip?

    16. Re:This actually makes sense by Bacon+Bits · · Score: 1

      You can argue the cleverness of it all day but Windows servers don't work without graphics. AFAIK they don't even boot.

      All x86 computers don't boot without a video adapter. It's not a matter of the operating system. It's the POST. The motherboard must detect a CPU, must detect RAM, and must detect a video display device. Those are the devices all PC compatible style x86 computers require to complete the POST. This is why "no display" is a common beep code and also why the video BIOS show is before the motherboard BIOS show. One can hardly blame MS for expecting the POST to complete successfully.

      --
      The road to tyranny has always been paved with claims of necessity.
    17. Re:This actually makes sense by swalve · · Score: 1

      I think for machines like that, what they mean is the graphics aren't integrated into the processor or the chipset like they would be on a consumer machine. Just a single chip hanging off a PCIe lane or something.

    18. Re:This actually makes sense by datavirtue · · Score: 1

      Yeah, because a Windows server is sooooo easy to setup. Have you used a modern Windows server?

      --
      I object to power without constructive purpose. --Spock
    19. Re:This actually makes sense by datavirtue · · Score: 1

      Microsoft is going full CLI for servers soon. This a great opportunity to learn Linux instead of relearning Microsoft "Windows?" It makes no sense to relearn all this stuff--powershell, commands--when the system is proprietary and you can spend your time learning something that will benefit you (Linux/Unix environment) no matter what you run. Every time I learn to deal with another Microsoft server product I miss any feeling of accomplishment because all it qualifies me for is babysitting someones crappy Microsoft environment, not terribly rewarding work. If I learn how to configure a Linux server I know I have something then and feel proud to master a system that is not meant to control and cajole me. I've played with Microsoft Windows for twenty years and I'm done. People keep using the excuse "because it just works" to explain their preference for Windows. This excuse is no longer working; I switched to Mint for my new build over Windows 7 and haven't looked back (I'm a software developer). The only reason they want to hang onto Windows is baecause they don't know much and they have no desire to improve. It is sickening.

      --
      I object to power without constructive purpose. --Spock
    20. Re:This actually makes sense by Anonymous Coward · · Score: 1

      Isn't the part of the POST that concerns itself with testing the video card actually *on* the video card's ROM? I don't think a lack of video card should stop a PC from booting, but we'll try to find a way to blame MS for it anyway.

    21. Re:This actually makes sense by jandrese · · Score: 2

      I'm pretty sure if I were building a server class box with no video, I would configure the POST to not stop if no video were detected. Plenty of server class machines have the option to output their boot messages to a serial port if need be, there's no compelling reason to require video on a rack box.

      --

      I read the internet for the articles.
    22. Re:This actually makes sense by Anonymous Coward · · Score: 0

      Actually, no. Many BIOSes can be configured to boot a PC without a video card. Oh, my good old 486 linux server with serial console...

    23. Re:This actually makes sense by hey · · Score: 1

      I wonder if there is a market there. If some body could make a really small graphics card. With very lower power consumption. It might even turn itself off once the system is booted. Perhaps it might not even do graphics - but just provide what the POST is looking for.

    24. Re:This actually makes sense by washu_k · · Score: 1

      This is completely untrue. I have a few desktop class boards that boot up fine without any video cards. They do make the no video card beep but they continue booting just fine. Windows server (at least 2008 R2) seems quite fine with this and doesn't care about the lack of video.

    25. Re:This actually makes sense by YA_Python_dev · · Score: 1

      All x86 computers don't boot without a video adapter.

      That's not true, there's nothing in the x86 architecture that requires a video card. And in fact you can buy x86 servers without one, discrete or integrated.

      It's not a matter of the operating system. It's the POST.

      It's actually the BIOS, and there are open source implementations that you can change the way you like, e.g. to not require a graphics adapter.

      --
      There's a hidden treasure in Python 3.x: __prepare__()
    26. Re:This actually makes sense by Anonymous Coward · · Score: 0

      Just wanted to point out real quick: Some x86 systems do boot without a video adapter of any kind. They're specifically designed with this in mind, and the BIOS is usually ported for this, often using a serial console instead. BIOS video calls get redirected to this console.

      It's actually pretty common on a lot of specialty servers and embedded systems, since a graphics chip would just be unnecessary, when a serial console and network would do fine. Most of the systems I've seen run some linux variant, and I know that RHEL will default to only using a serial console during the install if it doesn't detect a graphics device.

      Just my two cents!

    27. Re:This actually makes sense by JDG1980 · · Score: 1

      Microsoft is going full CLI for servers soon.

      Citation needed.

      I'm sure that Microsoft is working on improving its command line interface, but that's very different than going "full CLI." I very much doubt the standard GUI admin tools for Windows Server are going away any time soon.

    28. Re:This actually makes sense by rrohbeck · · Score: 1

      And most of all you want a video chip that supports IPMI. A server without IPMI is not up to contemporary standards IMHO.

    29. Re:This actually makes sense by egamma · · Score: 1

      Yeah, because a Windows server is sooooo easy to setup. Have you used a modern Windows server?

      I'm a windows server administrator, and have 185 2003 servers, 10 2008 servers, and 49 2008 R2 servers in my environment. I find it quite easy to set up and maintain windows servers. On the other hand, using vi on the 4 AIX servers I have to deal with occasionally--ugh.

    30. Re:This actually makes sense by realityimpaired · · Score: 1

      It's not a question of whether it'll boot for me, it's a question of how easy it is to troubleshoot. Most issues can be fixed with a remote connection, and in fact, I've never needed physical access to my computers in colocation to troubleshoot a problem before (though one of 'em has a dying CMOS battery and I'll need to make an appointment to replace it in the near future: its clock loses a few seconds every day and I have ntpd updating every 60 minutes to keep it in synch), but try troubleshooting a POST hang or failure to detect the network card issue without a physical display. When you're given a certain amount of space for your colocation, which is ample space to include a video card (either a full tower form factor, or a blade in a rack -- most blades I've seen have a VGA out built into the motherboard), and when you pay an hourly rate for physical access to your system, it's a no-brainer to include a video card so that you can connect the monitor/keyboard in the data center and see right there whether it's going to be a 5 minute fix or whether you need to take your computer back and return at a later date. :)

    31. Re:This actually makes sense by Just+Some+Guy · · Score: 1

      I almost never work from a server's physical console, but when I do it's because there's a hardware problem and I can't get it into an SSH-able status. At those times, a server-room-mounted KVM console is a nice thing to have.

      (Serial consoles are also nifty, sure, but there are still times when I'd rather be able to look at a screen physically plugged into the ailing machine.)

      --
      Dewey, what part of this looks like authorities should be involved?
    32. Re:This actually makes sense by gatkinso · · Score: 1

      Wrong.

      --
      I am very small, utmostly microscopic.
    33. Re:This actually makes sense by washu_k · · Score: 1

      If you have a proper server then you can do all that troubleshooting over ILO or whatever remote access card your vendor supplies, no physical display required. Some x86 servers have console serial ports, but that is of limited use for a Windows server.

      My point was that neither x86 or Windows require a video card to function. I used the example of a desktop board specifically because it lacks an ILO emulating a video card (no on board video either).

    34. Re:This actually makes sense by cthulhu11 · · Score: 1

      Video on "servers" pisses me off to no end. The answer is not to integrate the graphics, but to not have them at all. A "server" needs a working serial console that can be used to bootstrap from bare metal. Sun has done this for years, even on today's latest Sun/Oracle x64 servers. Systems from Dell, IBM, HP, etc. all effectively require that someone plug a monitor and keyboard into each physical server to set it up, and it sucks big fat hairy rocks.

    35. Re:This actually makes sense by Anonymous Coward · · Score: 0

      Microsoft is going full CLI for servers soon.

      Citation needed.

      I'm sure that Microsoft is working on improving its command line interface, but that's very different than going "full CLI." I very much doubt the standard GUI admin tools for Windows Server are going away any time soon.

      The standard GUI admin tools lack many administration abilities that can now only be done with powershell commands. And the syntax of those commands or the commands themselves keep changing from hotfix to rollup to service pack. Either try to roll a 2008 R2 environment from scratch or browse any Admin or Exchange 2010 forum and the evidence of this is abundant.

    36. Re:This actually makes sense by ttong · · Score: 2

      Because it's SO HARD to install joe, nano, ne or just edit files over a mounted sftp. Oh wait, you can't mount sftp in Windows.

    37. Re:This actually makes sense by WorBlux · · Score: 1

      They will probably still stick one on the motherboard, but they can use a bigger (hence cheaper) manufacturing process.

    38. Re:This actually makes sense by WorBlux · · Score: 1

      They are making it so many installation are command line only, though they've set it ups so the standard GUI's can operate remotely.It reduces resources needed for virtual instances of the OS.

  2. So, exept from the name.... by Ch_Omega · · Score: 5, Insightful

    ... What exactly does this have to do with the older pentium architechtures?

    1. Re:So, exept from the name.... by bill_mcgonigle · · Score: 1

      ... What exactly does this have to do with the older pentium architechtures?

      The current Intel line of decent chips is descended from the Pentium III line. While the marketing department was running the US operation with its Pentium 4 ("burn baby burn") line, a small group in Israel took the Pentium III and made it power efficient ('Core').

      Pentium 350 sounds like the end of Atom. Yay, I guess, since they handicapped Atom on purpose.

      --
      My God, it's Full of Source!
      OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
    2. Re:So, exept from the name.... by Anonymous Coward · · Score: 0

      Pentium 350 sounds like the end of Atom.

      No it doesn't. It's amazing that you decided it replaces Atom when it's targeted at servers (Atom isn't) and has a TDP of 15W (typical Atoms are 2.5W or less).

      Yay, I guess, since they handicapped Atom on purpose.

      Only in the sense that they designed a very low power, low cost x86 core. Atom's design point was TDPs of 0.5W to 5W and a die size of about 25mm^2 (typical full performance x86 chips are 100mm^2 minimum). Guess what, low power budget and low cost equals less performance and features. Who knew?

  3. Smart move by intel by Anonymous Coward · · Score: 0

    Their chips became to powerful and you can run 4-8 vm's off a core i7. That's only 300 bucks. We don't need that much power, scale it back and sell chips that would do the same as a p4 2.4ghz that appeal to the masses. Kudos intel.

    1. Re:Smart move by intel by bkmoore · · Score: 1

      How does a dedicated server appeal to the masses?

    2. Re:Smart move by intel by leenks · · Score: 5, Informative

      HP Microservers sold like hotcakes, and were based around AMD's Athlon II Neo N36L processor - which is 64bit, dual core, 25W TDP, VT-x etc. No doubt Intel want part of this pie

    3. Re:Smart move by intel by Anonymous Coward · · Score: 2, Interesting

      It's pretty obvious that the low TDP means it's meant for something fanless/low-noise like MacMini servers. Though the MacMini already runs a laptop i5 chip for the same reason.

      I hope, but have little faith in Intel when it comes to putting out cooler chips with lower TDP. When they have 100W+ TDP, you can only stuff 4 of them in a 15A rack. When they're 15W, that comes up to 24 ,machines or even blade systems.

      But I think it might actually be an attempt to beat Calxeda before it sells any servers. http://www.calxeda.com/products/energycore/ecx1000 1.5Watts per chip (5Watts at maximum power.) The point to notice is the Pentium VTd feature is missing, which means it's not meant for VM's

    4. Re:Smart move by intel by macs4all · · Score: 1

      It's pretty obvious that the low TDP means it's meant for something fanless/low-noise like MacMini servers. Though the MacMini already runs a laptop i5 chip for the same reason.

      Actually, you can BTO a dual-core i7 on the higher-end (server-intended) Mac mini. But yes, they do normally run an i5 in the 'minis.

    5. Re:Smart move by intel by JDG1980 · · Score: 1

      I hope, but have little faith in Intel when it comes to putting out cooler chips with lower TDP.

      Ever since they came to their senses and shitcanned NetBurst, Intel has beaten AMD in performance per watt on both desktops and servers.

      But I think it might actually be an attempt to beat Calxeda before it sells any servers. http://www.calxeda.com/products/energycore/ecx1000 1.5Watts per chip (5Watts at maximum power.) The point to notice is the Pentium VTd feature is missing, which means it's not meant for VM's

      I think this may be a preemptive strike against all the talk of ARM servers from various companies. Calxeda is just one of many places which are planning something like this. A cheap, low-power x86 that supports server features like ECC will dissuade many users from switching to ARM servers if and when they materialize.

  4. Dual core for servers? by aglider · · Score: 2

    That'd be a very very low end server!
    You can buy more powerful hardware, a desktop actually, with 4 cores and call it your server.
    Naaa, Intel is killing the Pentium.

    --
    Sent as ripples into the electromagnetic field. No single photon has been harmed in the process.
    1. Re:Dual core for servers? by the+linux+geek · · Score: 5, Informative

      Desktops don't have ECC. This does. Two cores with hyperthreading and ECC, only drawing 15W, isn't such a bad idea for the lowest of low-end servers.

    2. Re:Dual core for servers? by beelsebob · · Score: 5, Interesting

      What this looks perfect for is a NAS... now if only anyone would release a mITX C202/C204 board with 6 SATA ports on it.

    3. Re:Dual core for servers? by smash · · Score: 5, Informative

      Depending on the company, this would even be fine for an ESXi host to run 5-6 VMs on, given enough ram. As any ESX admin will tell you, you'll run out of IO and memory LONG before you get anywhere near running out of CPU these days, for all but the most cpu-demanding tasks (like VDI, code breaking, rendering, etc).

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    4. Re:Dual core for servers? by fa2k · · Score: 2

      Desktops don't have ECC. This does.

      Pretty much all AMD desktop chips support unregistered ECC memory. It's great for situations where you load up a desktop with RAM and hard drives instead of building a separate NAS. If Intel has two memory controller designs, I can understand that they don't ship ECC on desktops, but I suspect that they just flip a bit in the microcode, which is really annoying. AMD did it once as well, when they disabled a perfectly good core on quad core chips to get the cheaper 3-core versions, but they don't seem to disable features as much as Intel. I can totally understand that the Pentium ships with worse integrated graphics, because that's a large amount of silicon. The other stuff seems like a failure of capitalism to me -- Companies should be encouraged to make the most of their resources. People who buy i3s may realise that they want to do full disk encryption, why not give them the AES instructions?

    5. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      My HP desktop workstation at work has ECC and can take up to 32GB of it. ECC is available for desktops if you want it and it makes sense if you have a lot of RAM.

    6. Re:Dual core for servers? by serviscope_minor · · Score: 3, Informative

      Desktops don't have ECC. This does.

      Really? Because even a low end AMD motherboard supports (certainly the 40GBP range ASUS ones do) ECC. It's very hard to find such a low-clocked processor as the 1.8GHz Pentium 350, but processors don't put out all that much heat if you don't use them much. It's also a motherboard which allows for easy underclocking, should you wish to reduce the power draw.

      --
      SJW n. One who posts facts.
    7. Re:Dual core for servers? by __aardcx5948 · · Score: 1

      That'd be perfect. 6x 3TB in RAID6 on that baby... 12TB in a small box!

    8. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      Actually, it is quite nice to use as the MAC in a hardware-based router or switch. You must have a hardware forwarding engine to go with it, though... still, this chip is easily and cheaply connectable to ~4GB of ECC DDR3, and therefore is quite capable of holding large routing tables (such as several copies of the full Internet routing table, which is ~400k routes for each copy) and therefore it could work well in a border router.

    9. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      wtf happened to the other 6TB?

    10. Re:Dual core for servers? by PhrstBrn · · Score: 2

      Intel desktop (Core/Pentium/Celeron) processors don't. You need to get a Xeon UP/DP workstation processor (same socket as the server processors, but have a larger TDP and run at a higher clockrate) in order to get ECC. This is a "Pentium" branded processor, and not a Xeon that has ECC.

    11. Re:Dual core for servers? by Anonymous Coward · · Score: 2, Informative

      RAID 6 = 2 disks for "parity"

    12. Re:Dual core for servers? by beelsebob · · Score: 0

      RAID-z2 thanks

    13. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      By my calcs, the difference in cost at 12 cents per kwh, between a 15 watt CPU and a 45 watt CPU in a 24/7 system is, like, $32/yr. I mean, hey, if you're generating your own power from solar panels or something, that kind of power micro-management is perhaps significant, but really, power savings alone don't seem to justify their use.

    14. Re:Dual core for servers? by hairyfeet · · Score: 4, Interesting

      "Intel Desktops don't have ECC" FTFY

      AMD desktops all support ECC and have for ages. When you figure in the cost of the board and Intel's price for the chip one could easily get get one of the low power Phenom variations and still have money in your pocket. One of the places i like to get chips from (great bunch to shop with BTW) has the AMD Phenom X4 9150e which is a 65w quad at 1.8GHz for $55. Slap it in a nice cheap business class board, I prefer the ECS business class myself, and you have a nice cheap server that will be quiet as a churchmouse while being cheaper and more powerful than the Intel Pentium dual.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    15. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      It wouldn't be a bad idea if there was also a good and cheap motherboard to go with it. I checked on Newegg and the cheapest C204 motherboard is $165 and it doesn't even have USB 3.0 nor a PCI-E x16 slot.

    16. Re:Dual core for servers? by Agripa · · Score: 1

      My Phenom II desktop has ECC. The cost difference between it and an equivalent Intel system more than paid for a hardware raid controller and 4 big drives. My previous two desktops, a P4 and P3, both have ECC as well.

    17. Re:Dual core for servers? by CheshireFerk-o · · Score: 0

      I still run my dual 2.0ghz p4xeon with agp... shes strong as an ox if its not graphics you want.... I certainly want to build another 10 year box....

    18. Re:Dual core for servers? by CheshireFerk-o · · Score: 0

      self-reply...ugh... my mobo is a supermicro x5dal-tg2 and shes max'd =P

    19. Re:Dual core for servers? by LordLimecat · · Score: 1

      There are already Xeons with 20w and 45w TDP-- the Xeon E3 1220L and 1260L.

    20. Re:Dual core for servers? by LordLimecat · · Score: 1

      Comparing a 65w TDP to a 15w TDP is such a stretch its not even funny.

      Im sorry, I love AMD, how cheap they are, their graphics, the competition they bring to intel, etc; but theyre so uncompetitive right now its not even funny. For $200 I can get an intel processor that can do over 1GB/sec of AES encryption (thats around 10gbit vpn tunnels, if youre keeping score) in a 20w package. They seriously need to get their act together, especially as regards power draw.

    21. Re:Dual core for servers? by NewWorldDan · · Score: 1

      It would be more than adequate for most of the servers in my rack. Granted, I'm moving in the direction of virtualization, but I need to get a better storage infrastructure. That brings me to another question, where does one go to get an updated education on server/stoage technologies? I'm primarily a programmer, but being a small shop, I do everything, and my server knowledge is kind of lagging behind.

    22. Re:Dual core for servers? by PhrstBrn · · Score: 1

      Those are Xeon branded processors, which has nothing to do with the Pentium brand.

    23. Re:Dual core for servers? by swalve · · Score: 1

      Workstations, yes. Desktops, not usually.

    24. Re:Dual core for servers? by Anonymous Coward · · Score: 0

      That'd be a very very low end server! You can buy more powerful hardware..

      This is the kind of thing that happens when people think the word "server" really means anything useful. It doesn't. "Server" doesn't imply anything about the machine's load, so different people infer wildly different things and go off in completely different directions.

      Anyway, to address your main point, to some people, "very very low end server" is the only thing that makes sense to implement, because the server is the machine with the lowest CPU requirements. Those people don't want "more powerful hardware"; they want more efficient hardware, because the machine is on 24x7 and at any given moment it typically has load average 0.12 or something like that.

      I almost like the sound of this particular chip, except not quite. For all of Intel's dizzying array of SKUs they still haven't quite come up with the perfect mix of features for what I want out of a server (specifically, a file server).

      Here's what I'd like to see:

      1. Latest core, of course. SB. And only one or maybe two of them.
      2. Completely gone L3 cache, or at least a lot smaller. i.e. copy what distinguishes the Athlon II from the Phenom II. L3 is all about performance at expense of power efficiency, and the job I have in mind doesn't need that.
      3. DO have the entire fucking instruction set (this is what enrages my about Intel still), especially the AES stuff.
      4. Use the most popular/common socket of the day, which would be 1155 I guess.
      5. Integrated graphics, but the lowest end of current generation, so I guess that'd be HD2000.

      C'mon, Intel, where the hell is this part? You have so many that come close but miss. Oh, and leaving the AES stuff out of Core i3 makes me hate you. I know you're trying to make me "upgrade" to the more profitable for you Core i5, but it's not as clear of an upgrade to the buyer and involves more of a tradeoff than just price. You're fucking up.

    25. Re:Dual core for servers? by rrohbeck · · Score: 1

      This.
      I just built a new workstation around an Asus mobo (Crosshair V) and a Bulldozer with ECC RAM. I would have gone with Intel but non-ECC RAM is a no-no for work for me unless it's a laptop that works basically as a thin client.

    26. Re:Dual core for servers? by hairyfeet · · Score: 1

      Actually 65w is the rating, the TDP is 41 watts for a quad. So you are only using 11w more than your $200 chip and for the same price you are paying just for the chip i could have a complete server as long as I use a smallish SSD instead of the crazy priced HDDs.

      So if its anything its you sir that are comparing Apple & Oranges. you are comparing $200 chips to a $55 chip that can build you a whole system for less than the cost of just the chip and motherboard for your solution. Most of my customers would be happy to pay that extra 11w in return for having 4 machines for less than the price of two of yours.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    27. Re:Dual core for servers? by LordLimecat · · Score: 1

      Sometimes TDP is the most important thing. There is a datacenter Im looking at with incredible pricing ($400 / mo, with 100mbit line on a backbone included) who has a maximum power draw of 2 amps. Thats 240 watts.

      So I could either go get a kicking Xeon E3 processor with a 20w TDP and performance that slaughters anything AMD could throw at it, or I could go AMD in the name of saving about $150 up front and use at least double the power for around half the performance-- which in reality means quadruple the energy consumption.

      Theres not even a comparison there, its not even funny.

    28. Re:Dual core for servers? by toddestan · · Score: 1

      Your HP desktop workstation is undoubtedly running a Xeon CPU, which seems to be a pretty common solution in the high end workstation market (Dell and Apple do the same thing).

  5. It's only a matter of time. by drunkennewfiemidget · · Score: 5, Interesting

    Until Intel brings back the Pentium brand in general.

    Unless they're stupid.

    I'll never understand why they killed their most visible, most recognised brand.

    1. Re:It's only a matter of time. by yourdeadin · · Score: 1

      They had too much money and apparently AMD gave them enough rope to hang them self.

    2. Re:It's only a matter of time. by Tastecicles · · Score: 1, Insightful

      Their whole Pentium M, Centrino and Core brands are based on the Pentium III "Tualatin".

      Pentium never died, although P4 came close. That was a dead end.

      --
      Operation Guillotine is in effect.
    3. Re:It's only a matter of time. by SuricouRaven · · Score: 1

      Because the ageing pentium architecture was a mess, and they needed to redesign from scratch for the Core 2 architecture - which was a great improvement. They stopped using the pentium brand because they stopped selling chips with any pentium-based technology in.

    4. Re:It's only a matter of time. by Anonymous Coward · · Score: 0

      you mean like any of these?

      http://www.newegg.com/Product/Product.aspx?Item=N82E16819116381

      http://www.newegg.com/Product/Product.aspx?Item=N82E16819116399

      http://www.newegg.com/Product/Product.aspx?Item=N82E16819116398

      Pentium has always been here in name, its just not their primary chips any more. Pentium is more along the lines of "budget" chips these days

    5. Re:It's only a matter of time. by korgitser · · Score: 4, Interesting

      I'll never understand why they killed their most visible, most recognised brand.

      For me that hasn't been much of a question. For what I gather, the Intel brand is way stronger than the Pentium brand. You don't buy Pentium or Core, you buy Intel. Their changing the processor name only signifies that they are moving forward (and leading) as usual.

      --
      FCKGW 09F9 42
    6. Re:It's only a matter of time. by bhcompy · · Score: 4, Funny

      It's all about the Pentiums, baby

    7. Re:It's only a matter of time. by Guy+Harris · · Score: 5, Informative

      Because the ageing pentium architecture was a mess, and they needed to redesign from scratch for the Core 2 architecture - which was a great improvement. They stopped using the pentium brand because they stopped selling chips with any pentium-based technology in.

      What is the "Pentium architecture"? The microarchitecture of the original Pentium (P5) was different from the microarchitecture of the Pentium Pro/Pentium II/Pentium III (P6), and P6 was different from the microarchitecture of the Pentium 4 (NetBurst), and NetBurst was different from the microarchitecture of the Pentium M (which was, I think, P6-derived). The microarchitecture of the Core 2 (Core) was, I think, Pentium M-derived.

      So there's Pentium-the-chip (P5), and there's Pentium-the-brand, which was first used with the P5 chip but was also used with chips with significantly different microarchitectures from the P5 chip.

      The Pentium 350 apparently uses the Sandy Bridge microarchitecture, along with a bunch of other microprocessors, some named Core, some named Xeon, some named Celeron, and some named Pentium. Some of the ones named Pentium were launched in Q3 2011, before the Pentium 350, so "Intel Breathes New Life Into Pentium" is, to use the technical term, a "complete bullshit headline".

    8. Re:It's only a matter of time. by smash · · Score: 2

      Where "redesign from scratch" means "tweak the old p6 core a bit more"

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    9. Re:It's only a matter of time. by unitron · · Score: 3, Insightful

      "You don't buy Pentium or Core, you buy Intel."

      The more technically aware perhaps, but people like the dudes who got Dells (i.e., the ones who had no idea that there were any OS'es besides Windows) knew they wanted a Pentium even if they didn't know if it was made by Intel or Mat-tel.

      --

      I see even classic Slashdot is now pretty much unusable on dial up anymore.

    10. Re:It's only a matter of time. by Anonymous Coward · · Score: 1

      You know that a technology has become a commodity when even the nerds focus more on brand issues than on technical details.

    11. Re:It's only a matter of time. by Neil+Boekend · · Score: 1

      The impression I got from P4's was that they were always overheating and the consessions made to get these high frequencies were unacceptable (looooong pipelines) and limiting to the actual speed.
      It was an overall recognised brand, but not always in a good way.

      --
      Well, I might have a way, but it only works on a semi spherical planet in a vacuum.
    12. Re:It's only a matter of time. by SuricouRaven · · Score: 1

      It is complicated, yes. I can't keep track of it all myself.

    13. Re:It's only a matter of time. by unixisc · · Score: 1

      32-bit x86 CISC instruction set based CPUs should be called Pentiums - as you point out, the core architectures were all over the place. I'd consider Celerons, Xeons and others as subsets of Pentiums.

      As I said elsewhere, I'd call the x64 based CPUs something else other than Pentiums.

    14. Re:It's only a matter of time. by unixisc · · Score: 1

      Only 'redesigned from scratch' CPU from Intel was the Itanium. And to an extent, their EM-64s

    15. Re:It's only a matter of time. by drunkennewfiemidget · · Score: 2

      To most people, Pentium was just a brand name. The technology behind it was irrelevant. Even if tomorrow Intel starts making their CPUs out of tungsten and unicorn tears, they could still call it Pentium.

    16. Re:It's only a matter of time. by the_humeister · · Score: 1

      That makes absolutely no sense. You would consider a 32-bit Pentium 4 505 as a "True Pentium" and a 64-bit Pentium 4 506 as something else even though they are both based on Prescott microarchitecture but the 505 processor has 64-bit capabilities disabled???

    17. Re:It's only a matter of time. by Anonymous Coward · · Score: 0

      George: Who's keys are these?
      Jerry: Not my keys?
      Kramer: Who's are these?

    18. Re:It's only a matter of time. by unixisc · · Score: 1

      Yeah. I'd not base the naming on the internal similarities, but on the external differences. The way I'm suggesting it, one wouldn't be able to run 64-bit instructions on Pentiums, and so if the same CPU had the 64-bit capabilities enabled, I'd not call it a Pentium.

    19. Re:It's only a matter of time. by Anonymous Coward · · Score: 0

      The more technically aware preferred AMD over Intel back when AMD were selling their AMD Athlon XP desktop processors. Those processors ran hotter than Intel equivalents but they sure offered a lot more performance for compute-intensive applications and games.

    20. Re:It's only a matter of time. by Larryish · · Score: 1

      It's all about the Pentiums, baby!

      https://www.youtube.com/watch?v=qpMvS1Q1sos

    21. Re:It's only a matter of time. by mikkelm · · Score: 1

      You have a lot of naming to cover if you genuinely want to distinguish between instruction sets by name, rather than model number.

    22. Re:It's only a matter of time. by reub2000 · · Score: 1

      Until intel doubled the L2, increased the FSB to 800MHz, and really ramped up the clock speed. Then AMD was stuck with no other option than to inflate the ratings on it's processors. (AMD Athlon XP 3400+, sure....) AMD couldn't do much until the hammers fell.

    23. Re:It's only a matter of time. by the_humeister · · Score: 1

      Well, it can't possibly be any worse than Intel's current naming scheme!

    24. Re:It's only a matter of time. by Sloppy · · Score: 1

      You don't buy Pentium or Core, you buy Intel.

      You don't buy Intel; you buy Sandy Bridge.

      --
      As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
    25. Re:It's only a matter of time. by LordLimecat · · Score: 1

      I would buy that processor.

    26. Re:It's only a matter of time. by Guy+Harris · · Score: 1

      32-bit x86 CISC instruction set based CPUs

      All x86 processors are CISC (regardless of whether they internally turn the instructions into sequences of 1 or more RISCy micro-ops and schedule and execute those independently).

      should be called Pentiums

      Presumably meaning "Only 32-bit x86 CPUs should be called Pentiums", as you presumably don't intend to call the 80386 and 80486 Pentiums.

      (Not that Intel's marketing department has any good reason whatsoever to care what any of us think is the proper use for their "Pentium" brand.)

    27. Re:It's only a matter of time. by hawk · · Score: 1

      It Would sell like hoecakes on slashdot--it's well known that only virgins can cach unicorns . . :)

      hawk

    28. Re:It's only a matter of time. by Anonymous Coward · · Score: 0

      then shut the fuck up

  6. Easy answer, dude! by aglider · · Score: 1, Troll

    Because they're stupid! And stupidly dumb.

    --
    Sent as ripples into the electromagnetic field. No single photon has been harmed in the process.
  7. Cool! by Anonymous Coward · · Score: 0

    Will this one be defective and unable to do math like the pentium 60?

    1. Re:Cool! by Tastecicles · · Score: 5, Funny

      it was a 59.97, actually...

      --
      Operation Guillotine is in effect.
    2. Re:Cool! by unitron · · Score: 1

      I thought that was the video sweep on color NTSC sets.

      --

      I see even classic Slashdot is now pretty much unusable on dial up anymore.

    3. Re:Cool! by smash · · Score: 3, Funny

      A mate had a bugged P90 that he got cheap. For 99.9% of users, there was no issue. It certainly made for a cheap machine that kicked arse at quake, back when he got it for about the price of a 486.

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    4. Re:Cool! by Anonymous Coward · · Score: 0

      59.94, or to be more precise 60*1000/1001, it's the framerate of NTSC.

    5. Re:Cool! by Anonymous Coward · · Score: 0

      Or to be more precise is the fieldrate of NTSC :-)

    6. Re:Cool! by Tastecicles · · Score: 1

      I thought NTSC framerate was 23.976? Such an odd number simply because the field refresh rate (originally 60.0Hz) had to be reduced by a factor of exactly 1000/1001 (to give a FiR of 59.94Hz which with a bit of math and timecode drops (1 frame every 1000 count, on average) gives the aforementioned standard NTSC colour framerate) to overcome some constraint or other of the transition between monochrome and colour signal. I think, and stand to be corrected, that the colour signal was vulnerable to interference between horizontal, sound and colour frequencies and how they interacted with line frequencies.

      --
      Operation Guillotine is in effect.
  8. Intel by Cherubim1 · · Score: 1

    I assume this is for low duty servers ? In any case, Intel has made a complete mess of its processor naming conventions. They lack any form of consistency or logic.

  9. Atom.. by Severus+Snape · · Score: 2

    Is to try and stop people using Atom chips on the server due to their low profit margins? Doesn't have me fooled atleast. On the subject of the Pentium brand, it's best off where they left it. I think of Pentium I think slow, old and crap. To they extent I was put off the second I read the name.

  10. Nothing by Sycraft-fu · · Score: 5, Informative

    And this is also nothing new. They've been selling "Pentiums" for some time now. The Pentium G6950 is one for the last generation Core i series systems (LGA1156). The Pentium G600 and G800 series are for this generation core i series (LGA1155, Sandy Bridge). They are the same architecture as the i series chips, just more cut down.

    So for example with the current LGA1155 offerings:

    -- The i7-2600/2700 are the quad core, hyperthreaded chips with 8MB cache.
    -- The i5s are quad core, non-HT, 6MB cache.
    -- The i3s are dual core, hyperthreaded, 3MB cache. They also lack AES-NI instructions.
    -- The Pentiums are dual core, non-HT, 3MB cache and have slower graphics and clockspeed. They also lack AVX instructions (and AES-NI).
    -- The Celerons are even slower, and 2 or 1MB of cache, and the lowest end one is single core.

    In all cases they are all Sandy Bridge. They are 32nm chips with that core architecture. The lower end ones just have less features, cache, clockspeed, and so on and thus can be made cheaper.

    Basically these days "Core" is Intel's mainstream and high end brand. Everything from about $120 up is branded Core. Pentiums are their budget brand, the $60-100 range. Celerons are their extreme budget brand. $40-50 (only sold to OEMs).

    1. Re:Nothing by Shinobi · · Score: 5, Interesting

      The difference with the Pentium 350 is that it is HT and supports VT-x and ECC. And has a TDP of 15W.

      I'm trying to dump the Zacate I bought about a month ago onto someone now, and buy a Pentium 350 instead.... The Zacate gets rather hot(noticed 67 degrees Celsius from on-die sensor) when decoding a movie for example, even with a fan. With the Pentium 350 and a GT 520 for example, I could go completely fanless, and not reach those temperatures.

    2. Re:Nothing by rapidreload · · Score: 5, Funny

      Sycraft-fu, your posts are consistently accurate, informative and insightful. Could you at least pretend to troll once in a while? It's much more fun!

      --
      To all newcomers - people here are very close-minded and can't handle complaints about Linux. Keep this in mind.
    3. Re:Nothing by jawtheshark · · Score: 1

      Celerons are their extreme budget brand. $40-50 (only sold to OEMs).

      Sounds incorrect. I can get a Celeron here. Not that I'd ever buy a Celeron or a Pentium. Those names are tarnished for me.

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    4. Re:Nothing by swalve · · Score: 1

      I don't know what to think about this new Pentium line, but Celerons are fine. They are the same processor cores as the mainline processors, just with less cache usually. And possibly with some features turned off, like virtualization. There is nothing to be tarnished, it's just a $ per performance calculation. Some workloads don't need as much cache, and if you aren't using a feature, why pay for it?

    5. Re:Nothing by Anonymous Coward · · Score: 0

      The Pentium 350 is just a Core-i3 with some changes. If you compare the Pentium G6950 (Column 1), Pentium 350 (Column 2) and then most of the Desktop Core i3 CPU, it is easy to see that the Pentium 350 is just a Desktop Core-i3 with ECC added and graphics removed. It can be priced agressively without being easily compared to the desktop line.
      http://ark.intel.com/compare/43230,61272,53422,53423,53424,55448,53426,53427,59080,53428

    6. Re:Nothing by macs4all · · Score: 1

      The difference with the Pentium 350 is that it is HT and supports VT-x and ECC. And has a TDP of 15W.

      I'm trying to dump the Zacate I bought about a month ago onto someone now, and buy a Pentium 350 instead.... The Zacate gets rather hot(noticed 67 degrees Celsius from on-die sensor) when decoding a movie for example, even with a fan. With the Pentium 350 and a GT 520 for example, I could go completely fanless, and not reach those temperatures.

      Except for one thing: Isn't the max die temp for most Pentiums in the 45-50 deg. C world? (I couldn't find the 350 specifically).

    7. Re:Nothing by Shinobi · · Score: 1

      Don't know about the others, but for the 350 it's 76 degrees Celsius.

      http://ark.intel.com/products/61272/Intel-Pentium-Processor-350-(3M-Cache-1_2-GHz)

    8. Re:Nothing by hawk · · Score: 1

      Separate product lines again, huh?

      Once more, pentium sets out to divide and conquFOOF

  11. It makes some sort of sense by jwijnands · · Score: 2

    I suppose. Although these days I see most serious customers by a few high end servers and use vmware. I've yet to see a single low end low power server in a datacentre in this country

    1. Re:It makes some sort of sense by leenks · · Score: 2

      This kind of product is ideal for SMBs though, and even individuals who want a high performance NAS box. I've got AMD's equivalent processor (as I see it anyway) in a HP Microserver, and it runs a couple of Linux VMs and a Windows XP VM without a problem (for the odd bits of Windows stuff I have to do), as well as providing me with a fast 4x 3.5" removable HDD storage solution.

    2. Re:It makes some sort of sense by smash · · Score: 2

      This will likely be aimed at the small 10-20 employee shops to run vmware (or hyperV - blech) on. Free copy of ESXi, low end cpu like this with plenty of RAM = win. Add nodes/vcenter license and CPU/RAM as you grow.

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    3. Re:It makes some sort of sense by Prof.Phreak · · Score: 1

      There are corps who buy a ``a few high end servers and use vmware'' and then there are corps who buy thousands of low-end cheap boxes and build hadoop clusters.

      --

      "If anything can go wrong, it will." - Murphy

    4. Re:It makes some sort of sense by mjwalshe · · Score: 1

      actually from what my colleges a LN Risk tell me HPC clusters are not normally built with cheap as chips parts - they use commodity but fairly decent kit dual quad cores plus infiniband to tie it together

    5. Re:It makes some sort of sense by Anonymous Coward · · Score: 0

      I've yet to see a single low end low power server in a datacentre in this country

      That's because you're looking at datacentres. That biases what you're seeing, toward solutions which are based on a strategy of concentrating and sharing resources.

      A datacentre is precisely the kind of place where if a certain business really only needs n% of an awesome machine, then they can actually get away with literally using n% of an awesome machine. Now let's say we're talking about a server in an office with a DSL link to the internet. The office really only needs n% of awesome machine, but can't effectively rent or share (100-n)% out to other parties, so an awesome machine would be wasteful overkill. They're better off with a machine that n% awesome.

  12. Retroactively rebrand... by unixisc · · Score: 2

    ... all their 32-bit x86 CPUs as Pentiums, including the recent names, such as Pentium i3 Dual Cores, etc, an come up w/ a new name for all their x64 CPUs - maybe call it Hexiums, or Sexiums, and append them w/ their current names, such as Xeon, Core2Quad, et al, so that they'd have a good branding strategy. And come up w/ low cost versions of the Itanium, since it's obviously going nowhere in servers, and they might as well get some lower cost versions of that CPU and offer systems on that loaded w/ things like FreeBSD, Debian, et al.

    1. Re:Retroactively rebrand... by fnj · · Score: 1

      Just what i3's can you name that are not 64 bit capable?

    2. Re:Retroactively rebrand... by unixisc · · Score: 1

      Capability is irrelevant: if a processor's 64-bit is disabled, call it a Pentium

  13. Re:GOOD GOD MAN !! by beelsebob · · Score: 1, Offtopic

    I want the best !!

    Why do you want bulldozer then?

  14. ...of time, 22nm, till intel hit brick wall by Anonymous Coward · · Score: 1

    Because IBM rename one of their O.S.'s "i", and released the POWER7 which dominates the sever market in performance. Microsoft renames windows version 6.2 as Windows 7. IBM announces their PowerPC A2 processor which also sets record levels in performance all under 50-watts, so Apple names their processor A4. IBM announced in the late 90's the huge performance upgrade to their mainframes would be G5, so Apple decides to refer to the PowerPC 74xx's as the G series. Apple, Intel, and Microsoft are all trying to defuse the technological dominance that IBM has.

    1. Re:...of time, 22nm, till intel hit brick wall by dbIII · · Score: 1

      and released the POWER7 which dominates the sever market in performance

      Yes, but when you can get so many of anything else that is slightly slower it's fairly irrelevant unless you have a black-ops budget to play with or some other reason to be immune from accountants.

    2. Re:...of time, 22nm, till intel hit brick wall by unixisc · · Score: 1

      Except that IBM's A2 was based on PowerPC, whereas Apple's A4 & A5 were based on ARM.

    3. Re:...of time, 22nm, till intel hit brick wall by not-my-real-name · · Score: 1

      PowerPC 74xx's

      I remember that the 7400 was a quad two input NAND gate and the rest of the 74xx series was assorted other TTL logic gates. There were variants like the 74Sxx, 74LSxx, and others. There was also the 4xxx CMOS series, but the numbers didn't map, so 4000 was not a quad two input NAND gate. I think that there was also a 74xx variant 74Hxx(?) that was also CMOS, but used TTL logic levels. Now, a PowerPC built out of 74xx series chips would be quite the sight to see, especially if you wired it up to some blinkenlights.

      --
      un-ALTERED reproduction and dissimination of this IMPORTANT information is ENCOURAGED
  15. Still selling the same chip? by Anonymous Coward · · Score: 0

    1999: 1.2 gHz
    2011: 1.4 gHz

    NO SALE

    1. Re:Still selling the same chip? by Mr+Z · · Score: 4, Insightful

      Hmmm... processors do do more work per clock nowadays as compared to 12 years ago. And, they do it at waaaaaaaay less power and cost. Think about your huge many-fanned nearly 1kW rig from the turn of the millennium vs. the cramped space of a 1U slot pulling maybe 100W. This ain't your father's Pentium.

    2. Re:Still selling the same chip? by Anonymous Coward · · Score: 0

      Processors have been budgeted for 100-130W since the dawn of high performance desktop computing. It is only recently that chip makers have made low-performance, low-power versions of their high-performance processors - and usually not on purpose - they're just re-labeled defects that work at lower clock speeds but not at the original intended speed.

    3. Re:Still selling the same chip? by CAIMLAS · · Score: 1

      This ain't your father's Pentium.

      That made me feel really, really old.

      Given how they're targeting them (servers), I'd think calling them "Pentium Pro" would be more apropos.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
  16. Re:GOOD GOD MAN !! by Anonymous Coward · · Score: 0, Offtopic

    Because Pentiums are far too small for earth-moving work?

  17. Pentium 350? by unitron · · Score: 2

    What happened to the Pentium 5 through 349?

    --

    I see even classic Slashdot is now pretty much unusable on dial up anymore.

    1. Re:Pentium 350? by thegarbz · · Score: 1

      The Pentium did have maths problems.

    2. Re:Pentium 350? by unixisc · · Score: 1

      Floating point divide problems, to be precise. But I think GP was asking what happened to all the integers between 5 and 349?

    3. Re:Pentium 350? by thogard · · Score: 2

      Their formula for picking the next number involved a sub expression of 4195835/3145727.

    4. Re:Pentium 350? by Anonymous Coward · · Score: 0

      I had a Pentium 66 and a Pentium 266 back in the day.

    5. Re:Pentium 350? by unitron · · Score: 3, Funny

      In other words you had a waffle iron and a 266.

      --

      I see even classic Slashdot is now pretty much unusable on dial up anymore.

    6. Re:Pentium 350? by timepilot · · Score: 1

      They were not entirely successful.

    7. Re:Pentium 350? by Neil · · Score: 1

      Damn. First time in months that I've been tempted to post, and you beat me to it ... :-)

    8. Re:Pentium 350? by Carewolf · · Score: 1

      Well some of them existed. I had a Pentium 100 back in the day where the clock-frequency was part of the chip name.

    9. Re:Pentium 350? by godrik · · Score: 1

      "What happened to the Pentium 5 through 349?"

      Their marketing guy comes from mozilla corporation.

    10. Re:Pentium 350? by Anonymous Coward · · Score: 0

      Bah, i'm sitting next to my Pentium2 350(MHz). It's from 1997.

    11. Re:Pentium 350? by leenks · · Score: 1

      No it isn't.

    12. Re:Pentium 350? by toddestan · · Score: 1

      Well the Pentium 60, 66, 75, 90, 100, 120, 133, 150, 166, and 200 were actual chips produced by Intel from the days when the clock speed was part of the processor's name which only makes this new name even more confusing despite the Pentium (P5) line never getting to 350 MHz.

  18. Integrated graphics? by TemperedAlchemist · · Score: 1

    More like integrated waste of money.

    1. Re:Integrated graphics? by gl4ss · · Score: 1

      on 1.2ghz server though, integrated graphics would be more reasonable than on most of the machines integrated graphics ship on.. so.. uh.. I don't think this is a chip for totally gpu'less installations either..

      --
      world was created 5 seconds before this post as it is.
  19. New legal battle on the horizon by alfredos · · Score: 2

    So if Intel is now pushing the Pentium brand, and having suffered Intel's legal belligerence myself, I feel sorry for all those who have brand names starting with P and having less than 12 letters.

  20. Doesn't support VMs in hardware by Anonymous Coward · · Score: 0

    So it won't be fine for an ESXi host.

    1. Re:Doesn't support VMs in hardware by smash · · Score: 1

      Hmm.. thought i saw someone else comment that it has VT instructions. If not, then too bad.

      Still, point remains I guess. Even if its pushed at low end physical servers - CPU these days just isn't required. Even less so if you aren't virtualising multiple servers onto one machine. Just retired a couple of blades last year that were pentium III 1.3ghz single cores. They were still running around 90-99% idle most of the time (web server for internal app + db server to go with it).

      --
      I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
    2. Re:Doesn't support VMs in hardware by macs4all · · Score: 1

      Hmm.. thought i saw someone else comment that it has VT instructions. If not, then too bad.

      Still, point remains I guess. Even if its pushed at low end physical servers - CPU these days just isn't required. Even less so if you aren't virtualising multiple servers onto one machine. Just retired a couple of blades last year that were pentium III 1.3ghz single cores. They were still running around 90-99% idle most of the time (web server for internal app + db server to go with it).

      Yep. I've never understood what the big need is for massive CPU when most of the heavy lifting in a file-server is handled through DMA transfers. Hell, all a file server does (for the most part) is lookup files in a directory (which I believe is a CPU-saving indexed file architecture in most, if not all, modern filesystems), then setup a DMA controller with source and destination addresses, and then stand back until a "Transfer Complete" interrupt is raised. Even if the DMA controller has to be setup for each block (likely), that means that the next block location and maybe the destination address and block length has to be stuffed. Whoop-de-doo.

      My gut says that a file server will run out of I/O (both internal and external) and disk bandwidth LONG before the CPU is maxed-out.

      I remember setting up an old 350MHz G3-based iMac with a (paltry) 512MB of RAM up as a streaming video server about 5 years ago. I was amazed to find that serving 10 simultaneous (but different) streams only took about 2% of that slow-ass CPU with a slow-ass bus (100MHz) and moderately-slow HD. I understand we aren't talking about 1000 simultaneous connections; but we're also not talking about 2.7GHz, multi-core, multi-threaded CPUs, with 1GHz frontside busses and 16GB of RAM, either.

  21. A cluster**** of unclearly-positioned brands by Dogtanian · · Score: 5, Informative

    Basically these days "Core" is Intel's mainstream and high end brand. Everything from about $120 up is branded Core. Pentiums are their budget brand, the $60-100 range. Celerons are their extreme budget brand. $40-50 (only sold to OEMs).

    Good grief, Intel's marketing department really needs a good slapping.

    Their brand positioning used to make sense when you knew that Celeron was their budget line (though sometimes quite decent) and Pentium-XXX (later replaced by Core-XXX) the standard midrange, with Xeon for servers.

    So when they brought back Pentium, I was confused until (as you say) realising that it was meant as a kind-of-lower-priced line, but not as cheap as the Celeron (*). Confused partly because they still had the Core 2 (**) then i3/i5/i7 lines as their mainstream brand which Pentium used to represent.

    In other words, they brought back the Pentium name due (presumably) to some vague consumer recognition, but not for what it was used before and for some vaguely-defined semi-budget segment.

    Worse, it isn't even necessary because the current "Core" line is split into i3, i5 and i7, which is an easily-understood hierarchy, and along with the "Celeron", there's absolutely no need for another damn confusing name.

    *Now* they're making things even more of a cluster**** by using the Pentium name on low-end *server* (not mainstream) processors.

    Please note that I'm *not* talking about the underlying architecture, which marketing doesn't necessarily follow, and which the man on the street probably doesn't care about much. I'm simply talking about incompetent marketing and positioning in that there are a mess of names that no longer represent their intended price segment and/or use clearly.

    Then again, perhaps confusion is the aim of the game, as it makes it easier for sales people to bamboozle the public and upsell more expensive CPUs than they need? But I suspect not.

    (*) You say that Celeron is now an ultra-cheap OEM-only thing, but I can still apparently purchase boxed versions here and here, for example.

    (**) And while I'm here, "Core" and "Core 2" were absolutely stupid choices for a processor name, as "core" already had a technologically-defined use we all know well, and "Core" (the name) was thus guaranteed to confused anyone not in the field, e.g. a dual-core Core, etc. etc..... "Core 2" was even worse, as it's going to get easily confused with "dual core" and terms like "Core 2 Quad" (i.e. a four-core "Core 2"!) are just a confusing mess for Joe Public. I know of at least one alleged computer technician (i.e. someone who *could* be expected to know this) who thought that "Core 2" in itself meant that it was a dual-core processor! I'll give them a free pass on the fact that the original "Core" line didn't actually feature the "Core" architecture, as I was complaining about bad marketing, and marketing doesn't normally mention internal architectures anyway.

    --
    "Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
    1. Re:A cluster**** of unclearly-positioned brands by iviv66 · · Score: 5, Funny

      *Now* they're making things even more of a cluster**** by using the Pentium name on low-end *server* (not mainstream) processors.

      You forgot to include the footnote for explaining the **** there!

    2. Re:A cluster**** of unclearly-positioned brands by gallondr00nk · · Score: 2

      What's worse, up until fairly recently they had *two* different chips named after the Pentium. The Pentium D and the Pentium Dual Core.

      Who thought up these product ranges? Never mind naming them both Pentium, but giving them similar names? I've known lots of people confuse the two. I have no idea why we need all these product ranges. Celeron, Pentium and Xeon should be sufficient, with maybe something to differentiate the i7. But for god's sake, give them distinct and understandable names!

    3. Re:A cluster**** of unclearly-positioned brands by Namlak · · Score: 2

      You forgot to include the footnote for explaining the **** there!

      **** fsck**

    4. Re:A cluster**** of unclearly-positioned brands by macs4all · · Score: 1

      Worse, it isn't even necessary because the current "Core" line is split into i3, i5 and i7, which is an easily-understood hierarchy, and along with the "Celeron", there's absolutely no need for another damn confusing name.

      They're just borrowing a page from the Microsoft marketing play-book: Just keep changing the name of things so that dumb consumers (and lazy OEMs) will think its "something new".

    5. Re:A cluster**** of unclearly-positioned brands by Anonymous Coward · · Score: 0

      You forgot to include the footnote for explaining the **** there!

      Yeah, because "beowulf" doesn't fit in 4 characters.

    6. Re:A cluster**** of unclearly-positioned brands by CAIMLAS · · Score: 1

      It's not the "sorta Celeron" it's the "Celeron of their server CPUs", the little brother to the Xeon. people were (are) buying desktop boards and putting them in cheap, low-end servers. This fills that gap, but still lets you use a full-featured server board.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    7. Re:A cluster**** of unclearly-positioned brands by Anonymous Coward · · Score: 0

      I'm still holding out for the Sextium. I am a /.er, so that's likely as close as I'm ever gonna get.

    8. Re:A cluster**** of unclearly-positioned brands by Dogtanian · · Score: 1

      It's not the "sorta Celeron" it's the "Celeron of their server CPUs", the little brother to the Xeon. people were (are) buying desktop boards and putting them in cheap, low-end servers. This fills that gap, but still lets you use a full-featured server board.

      Perhaps, but you've missed the point- my rant wasn't about what the chips did nor who they were (supposedly) aimed at, it was a criticism of their poor and confusing naming scheme.

      Specifically:-

      (a) too many names in areas where it wasn't necessary (i.e. Celeron, i3, i5 and i7 are already more than enough to differentiate the consumer/business desktop segment),

      (b) reusing names for slightly different purposes to what they had stood for for years (e.g. Pentium-XXX used to be the mainstream/flagship desktop line, now it's a semi-budget brand in an area where it isn't needed anyway, see (a)), and

      (c) Using the same name for multiple distinct uses at the same time, e.g. Pentium is a sort-of-budget desktop chip, er... and it's also the "celeron of their server CPUs", so wouldn't "Celeron Xeon" have been clearer about that, given that Celeron was always de facto a budget or cut/down version of their mainstream chips?

      --
      "Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
    9. Re:A cluster**** of unclearly-positioned brands by Anonymous Coward · · Score: 0

      Pentium D *was* a Pentium in the old-fashioned sense: it was two Pentium 4 dies in a single package with a simple bus arbitration system to allow them to share a single external connection.

    10. Re:A cluster**** of unclearly-positioned brands by Guy+Harris · · Score: 1

      It's not the "sorta Celeron" it's the "Celeron of their server CPUs"

      "It" presumably means "the Pentium 350"; there are other, non-server, Pentiums in the Intel Sandy Bridge-architecture product line.

    11. Re:A cluster**** of unclearly-positioned brands by evilviper · · Score: 1

      It makes sense that yesterday's high-end brand is today's low-end. If you've been buying "Pentiums" in the past couple years, they've been old, out of date chips, so the old brand is tarnished, replaced by something newer and shinier, but still holding on, as warehouses of P4s and Pentium-Ds are still selling well, but only at fire-sale prices.

      That they're using the P brand only for servers, yeah, that's pretty damn stupid... I can't argue with that. Those buying servers will be looking at specs, and are interested in exact model numbers, to hell with vague marketing names...

      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    12. Re:A cluster**** of unclearly-positioned brands by toddestan · · Score: 1

      Which would be fine if that's all they did. But no, they decided later to call a bunch of chips "Pentium Dual Core" which were cut down Core 2 chips and had little to do with the Pentium D except for sharing a socket. Oh, and to make things more confusing they also had the Celeron D at the same time as the Pentium D, except it wasn't even a dual core chip but rather just a cut down single core Precott P4 chip.

    13. Re:A cluster**** of unclearly-positioned brands by Dogtanian · · Score: 1

      It makes sense that yesterday's high-end brand is today's low-end. If you've been buying "Pentiums" in the past couple years, they've been old, out of date chips,

      AFAIK the Pentium (brand) hadn't been dead for very long before they brought it back- the point at which you're suggesting this logic would have been applied in the repositioning.

      so the old brand is tarnished, replaced by something newer and shinier, but still holding on, as warehouses of P4s and Pentium-Ds are still selling well, but only at fire-sale prices.

      No offence, but this comes across as an over-thought post-rationalisation of Intel's rather badly-thought out choice.

      I doubt they applied that logic, and I doubt that the people Intel marketing had in mind when choosing names (i.e. mildly-informed computer buyers and business manager types who don't know nor care much about CPUs beyond intended purpose and price segment) would think about it that way either.

      Far more likely that someone in marketing recognised that- gasp!- the recently-buried Pentium name still had a lot of brand recognition and they wanted to somehow exploit that, but were too late to kill off the new "Core" name that had replaced it in terms of positioning, so contrived some pointless, half-baked new "sort-of-budget-but-not-as-cheap-as-the-Celeron" segment that they could use the "Pentium" name for.

      Really, who gives a s*** anyway, Intel's naming scheme is a mess, and that's all we really need to know. :-) Probably what you get with too many marketing twonks trying to justify their own pointless, over-salaried jobs and get noticed.

      --
      "Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
    14. Re:A cluster**** of unclearly-positioned brands by evilviper · · Score: 1

      AFAIK the Pentium (brand) hadn't been dead for very long before they brought it back- the point at which you're suggesting this logic would have been applied in the repositioning.

      Intel went about 2 years without any pentium branded products before reintroducing it. And to make it worse, that 2 year gap was after the brand had taken a serious beating with years of Pentium-4 CPUs being greatly inferior to AMD offerings.

      No offence, but this comes across as an over-thought post-rationalisation of Intel's rather badly-thought out choice. I doubt they applied that logic, and I doubt that the people Intel marketing had in mind when choosing names

      Actually, it's quite the opposite. There's no question Pentium was a damaged brand. YOU are the one who wants to paint it as an idiotic move with no thought put into it.

      Even if you are correct, then they still happen to have dummied their way into a mostly-correct decision, keeping the brand alive. Unlike you, I don't claim to know the rationale behind it, and I'd be the last person to try to rationalize or justify anything Intel does (I pretty well exclusively buy AMD gear, myself).

      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    15. Re:A cluster**** of unclearly-positioned brands by Dogtanian · · Score: 1

      Intel went about 2 years without any pentium branded products before reintroducing it.

      Unless the dates on this article are wrong, there doesn't appear to be anything approaching a two year gap.

      And to make it worse, that 2 year gap was after the brand had taken a serious beating with years of Pentium-4 CPUs being greatly inferior to AMD offerings.

      I doubt the man in the street knew diddly-squat about AMD- beyond maybe knowing they were Intel's competitor, and a small proportion of people possibly having heard of their smart new 64-bit chip- though they'd probably have known "Intel Inside" and "Pentium" quite well back then.

      You're confusing your own knowledge of the situation with that of Joe Public that Intel seem to design their name for.

      Actually, it's quite the opposite. There's no question Pentium was a damaged brand. YOU are the one who wants to paint it as an idiotic move with no thought put into it.

      If (as was apparent elsewhere) the intended significance of the Pentium brand was unclear to a significant proportion of Slashdot readers, it sure as hell isn't going to be clear to Joe Sixpack!

      then they still happen to have dummied their way into a mostly-correct decision, keeping the brand alive.

      The problem was they made that decision too late, after they'd replaced it with the "Core" name and couldn't backtrack. Hence the stupid new non-segment they shoehorned it into. They should have just kept it- what the hell was the point of a stupidly-confusing name like "Core" anyway?

      --
      "Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
  22. Re:GOOD GOD MAN !! by hairyfeet · · Score: 3, Insightful

    Actually I'm shocked nobody has tried cooking up a low cost low power server based on brazos. With OpenCL you could harness the built in Radeon GPU and at 18w for a dual core 1.6GHz part its a power sipper. You could slap 4 of them into a blade and only be using 72 watts for an 8 core with 4 Radeon GPUs you could run GP/GPU code on. Now that even Nvidia is supporting OpenCL it seems like that would be a better deal over the Pentium in TFA.

    --
    ACs don't waste your time replying, your posts are never seen by me.
  23. What about Windows Servers? by Anonymous Coward · · Score: 0

    Are these not for Windows servers? Or do they expect that a desktop system will work with integrated graphics, and a Windows server will require a real graphics card?

    I would expect it to be the other way around. The desktop user needs the real graphics card for his games, where as the server will do fine with integrated graphics.

  24. Weird Al had it right after all!! by philban · · Score: 1

    Had to be done: http://www.youtube.com/watch?v=qpMvS1Q1sos&ob=av3e *Long live the Pentium*

  25. Re:Not Cool, yet Hot Very HOT! by Anonymous Coward · · Score: 0

    In the mid 90's DOE built a cluster Pentium Pro's, yet they were always over heating and the electric bill was extremely prohibitive. So they scraped the entire system that was less then a year old, and brought in IBM to build the super's using POWER4's.

  26. Only by changing the architecture. by Anonymous Coward · · Score: 0

    So if you're right, and it gets far more done per clock, then it isn't a pentium. If it IS the pentium, then the OP is right, and it's a 200MHz speed pimple.

    1. Re:Only by changing the architecture. by Mr+Z · · Score: 2

      It's a Pentium in as much as they named it "Pentium," much like a Ford Taurus of today is just as much a Ford Taurus as one they made 20 years ago, even though they have no parts in common. It's a Taurus because Ford named it that.

      The Intel Atom probably has more in common with the original P54C than the Pentium Pro did.

  27. hopefully without any bugs by Errtu76 · · Score: 1

    f00f

  28. PowerPC 750 was already the G3 by tepples · · Score: 1

    IBM announced in the late 90's the huge performance upgrade to their mainframes would be G5, so Apple decides to refer to the PowerPC 74xx's as the G series.

    Or perhaps they called the 7400 series "G4" because the PowerPC 750 was already the G3. As marketed to Mac users, the first-generation PowerPC was 601, the second was 603 and 604, and the third was 750.

  29. Windows 95? by tepples · · Score: 1

    The same thing that happened to Windows 4 through 94.

    1. Re:Windows 95? by Anonymous Coward · · Score: 0

      The same thing that happened to Windows 4 through 94.

      And OS I to OS X.

    2. Re:Windows 95? by not-my-real-name · · Score: 1

      And OS I to OS X.

      I just pitched a t-shirt that mentioned OS 8.5, and I have used Mac OS 6, 7, 8, and 9.

      --
      un-ALTERED reproduction and dissimination of this IMPORTANT information is ENCOURAGED
    3. Re:Windows 95? by LocalH · · Score: 1

      Windows 4 through 94, hell. What about Windows 99 through Windows 2000? 1,901 versions is a huge gap.

      --
      FC Closer
  30. Summary again by msobkow · · Score: 2

    ...the Pentium 350 lacks features such as integrated graphics, which are on most of Intel's laptop and desktop processors.

    Somehow I doubt that integrated graphics are on "most" of their chips, unless you're talking about the volume shipped for laptops, and even then I thought the graphics were on a separate chip in most cases.

    --
    I do not fail; I succeed at finding out what does not work.
  31. So what? by kenh · · Score: 1

    I'm writing this post from an Acer very low-end notebook that sports a "Sandybridge" Celeron CPU - it is a dual core P4600 with dual cores, HD1000 integrated graphics, etc. and the entire laptop was $229 at Bestbuy - it can take 8 Gigs of DDR3 RAM and performs very nicely according to the Windows Experience Index (mid-5's and above on all ratings). I assume it is a Core i3 that failed some test, but it works just fine for me.

    I've also got a small Wolfdale E3400 "Celeron" desktops at home - a dual core Celeron that performs quite nicely, again, I assume it was a CPU that failed certification for higher-spec applications,

    There is a long tradition of "binning" parts based on test results - at least as far back as the 80386sx (where a failed math coprocessor section made the chips more affordable), no problem here...

    --
    Ken
    1. Re:So what? by Anonymous Coward · · Score: 1

      NOPE. No 386 from Intel ever had a built in coprocessor. You mean 486. It's also questionable whether the primary motivation in the 486 for the SX/DX lines resulted from manufacturing glitches or just plain market coverage.

  32. Re:GOOD GOD MAN !! by Thrawn7 · · Score: 1

    Brazos' GPU is low power but its also quite gutless. To use bitcoin mining as an example a single E-350 does 12 mhash/sec. Your 72 watt example would do 48 mhash/sec. A single Radeon 5770 would do 150 mhash/sec and use up only 110 watts (say 130 watts with a low power Sempron). Larger GPU cores are much more power efficient.

  33. TOP TEN new Intel slogans for the Pentium by Anonymous Coward · · Score: 0

    9.9999973251 It's a FLAW, Dammit, not a Bug
        8.9999163362 It's Close Enough, We Say So
        7.9999414610 Nearly 300 Correct Opcodes
        6.9999831538 You Don't Need to Know What's Inside
        5.9999835137 Redefining the PC -- and Mathematics As Well
        4.9999999021 We Fixed It, Really
        3.9998245917 Division Considered Harmful
        2.9991523619 Why Do You Think They Call It *Floating* Point?
        1.9999103517 We're Looking for a Few Good Flaws
        0.9999999998 The Errata Inside

  34. Re:GOOD GOD MAN !! by Khyber · · Score: 1

    The AMD GPU is far better at bitcoin mining than you suspect. A low-end 6XXX series MOBILITY Radeon hits around 80MHash/sec, ALONE.

    --
    Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  35. Re:GOOD GOD MAN !! by Thrawn7 · · Score: 1

    Not the 6310 in a Brazos it won't... The fastest 6xxxM in the mining wiki is 6630M 49 Mhash/sec (probably overclocked too). At least 30 watts to pull that off.

  36. Could be great for home server / NAS by JDG1980 · · Score: 1

    The Intel site indicates this new chip has a TDP of only 15 watts, and supports ECC RAM. This would make it great for a home server, such as a ZFS NAS, if the price is reasonable (and assuming it isn't OEM-only).

  37. Re:GOOD GOD MAN !! by JDG1980 · · Score: 1

    Actually I'm shocked nobody has tried cooking up a low cost low power server based on brazos.

    Brazos doesn't support ECC.

  38. masters of confusion by reiisi · · Score: 1

    Don't argue with them. They know what they're doing. They've built their empire on confusion.

    --
    Computer memory is just fancy paper, CPUs just fancy pens with fancy erasers; the 'net is just a fancy backyard fence.
  39. Over network by DrYak · · Score: 1

    on modern machines, it is even possible to route these serial boot message over the network (SoE - Serial over ethernet), or to have the display output to a virtual frame buffer and made available through an embed VNC server.

    --
    "Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]