Killing the Business Desktop PC Softly
by Johan De Gelas on July 19, 2007 3:00 AM EST- Posted in
- IT Computing
Workstation Blades
The IT experts among our readers are probably protesting: CCI was launched in 2005 and blade PCs have been on the market since 2004-2005. So what's new? Besides the fact that the technology has matured and been upgraded, the technology is now able to meet the needs of more demanding applications like those used by CAD users, data miners, financial traders, and even 3D artists. So basically, the blade PCs are also now offering higher graphics and CPU performance. The "magic" to make this work is to replace the normal "remote protocols" such as ICA (Citrix) or RDP (Microsoft) by a proprietary lossless compression and encryption protocol. IBM has not yet given it a name to our knowledge; HP calls it Remote Graphics Software or RGS.
The blade workstation performs all 2D and 3D calculations and compresses and encrypts each frame before it's sent to the thin client. This kind of network transmission of 3D and 2D graphics at high "refresh" rates requires between 2 and 4 Mbit/s of bandwidth on average. The goal is to make it feel like the manipulation of 3D CAD is actually being done on the thin client. For this you need at least 50-60 frames per second and a response time of less than 20ms. With an excellent broadband internet connection, a leased line, or a LAN connection, it should be possible to get good performance even if your thin client is 2000 miles away from the blade PC or workstation blade. As long as your network connection doesn't add more than 10-15 ms to your frame time, it feels like you are working on a normal workstation. RGS works on both Red Hat Linux and Windows XP. You can also license RGS for blade PCs: HP's blade PCs contain a mobile version of AMD's 690G chipset.
Both the IBM and HP blades are based on the Intel 5000p chipset platform. The fastest supported CPU is a dual core Xeon 5150 at 2.66 GHz. The HP blade supports 16GB, the IBM blade 8GB. However, the IBM has the faster graphics card: it can support up to the modern Quadro FX1600 (256-bit interface), while the HP is limited to the relatively low-end older FX540M 128MB (128-bit memory interface). Notice that both IBM and HP are using lower clocked mobile versions, a result of using the cramped blade chassis. Both IBM and HP make quite a bit of noise claiming that their workstation blades require quite a bit less power than a typical workstation, around 150-200W versus >300W for a typical PC workstation. Although the workstation blades are slightly more efficient thanks to the fact that several blades use one or two big PSUs, the biggest gains are the result of using mobile but slower performing video chips.
The IT experts among our readers are probably protesting: CCI was launched in 2005 and blade PCs have been on the market since 2004-2005. So what's new? Besides the fact that the technology has matured and been upgraded, the technology is now able to meet the needs of more demanding applications like those used by CAD users, data miners, financial traders, and even 3D artists. So basically, the blade PCs are also now offering higher graphics and CPU performance. The "magic" to make this work is to replace the normal "remote protocols" such as ICA (Citrix) or RDP (Microsoft) by a proprietary lossless compression and encryption protocol. IBM has not yet given it a name to our knowledge; HP calls it Remote Graphics Software or RGS.
The blade workstation performs all 2D and 3D calculations and compresses and encrypts each frame before it's sent to the thin client. This kind of network transmission of 3D and 2D graphics at high "refresh" rates requires between 2 and 4 Mbit/s of bandwidth on average. The goal is to make it feel like the manipulation of 3D CAD is actually being done on the thin client. For this you need at least 50-60 frames per second and a response time of less than 20ms. With an excellent broadband internet connection, a leased line, or a LAN connection, it should be possible to get good performance even if your thin client is 2000 miles away from the blade PC or workstation blade. As long as your network connection doesn't add more than 10-15 ms to your frame time, it feels like you are working on a normal workstation. RGS works on both Red Hat Linux and Windows XP. You can also license RGS for blade PCs: HP's blade PCs contain a mobile version of AMD's 690G chipset.
 |
 |
| IBM's HC10 and HP's XW460c | |
Both the IBM and HP blades are based on the Intel 5000p chipset platform. The fastest supported CPU is a dual core Xeon 5150 at 2.66 GHz. The HP blade supports 16GB, the IBM blade 8GB. However, the IBM has the faster graphics card: it can support up to the modern Quadro FX1600 (256-bit interface), while the HP is limited to the relatively low-end older FX540M 128MB (128-bit memory interface). Notice that both IBM and HP are using lower clocked mobile versions, a result of using the cramped blade chassis. Both IBM and HP make quite a bit of noise claiming that their workstation blades require quite a bit less power than a typical workstation, around 150-200W versus >300W for a typical PC workstation. Although the workstation blades are slightly more efficient thanks to the fact that several blades use one or two big PSUs, the biggest gains are the result of using mobile but slower performing video chips.
Facebook
Twitter
RSS
39 Comments
View All Comments
TA152H - Thursday, July 19, 2007 - link
Actually, I had no steam until I read the Compaq stuff, so don't patronize me.I worked at IBM, and man did we hate Compaq. When I read stuff like that, I guess I still have some true blue blood in me, and it makes me mad.
They didn't do anything. Compaq was certainly not the only company to reverse engineer IBM's BIOS, or really ROM as it was called then. It was done by many companies. Ever hear of Phoenix?
They were not an important company in any way, there were plenty of clone makers out, although Compaq did make a reliable clone, but it was expensive as Hell too. They did use the 8086 instead of the 8088, and were the first to use the external cache on the 386, but then, Intel created the cache chip (82385)and setup so I'm still not sure what credit they deserve. What did Compaq do that would not have been without them? Nothing! They invented nothing. They just sold an expensive clone, that was ugly. IBM, by contrast, did a lot. Apple too, really.
I don't know how to make the analogy for the parent, but Compaq was a meaningless company that in no way shaped the industry. I would call IBM the mother, and maybe Intel the father. Or maybe even Apple, because if you look at the IBM PC, it did seem to borrow a few things from the Apple II. Maybe it was a half grandfather or something. Tandy was an extremely important player too, not only because of the PC compatibles, but before. They brought the PC home, and did a lot of minor innovations too, like putting the OS in ROM, and adding their own low-end GUI (Deskmate). They also released the PC that Bill Gates said was what he used to develop Windows (The Tandy 2000, an odd bird based on the 80186, and MS-DOS compatible but not truly PC Compatible since the ROM calls were different).
The reality is, the PC was really just an IBM product. No other company deserves too much credit, and frankly, IBM didn't even make anything too revolutionary, when you consider things. They made probably the biggest mistake in business history in fact, using a microprocessor from Intel, and an operating system from Microsoft. They thus lost control of the computer industry, slowly at first, but inevitably as microprocessors became too powerful for IBM's liking. They also made huge mistakes by degrading their PCs so as not to compete with their other lines. Not only in performance (why did the PC/AT have a wait state???), but also in price. Would anyone have bought that junk from Compaq if IBM didn't artificially inflate prices on their PS/2s? Would there have been clones if IBM had used proprietary technology, instead of using an inferior microprocessor, and off the shelf parts?
Really, IBM only should get credit because of their name. It created a standard, and that it was a standard was more important than the actual company that created it. There was nothing special about the machines, although the PS/2s were incredible machines when released in 1987, and the RT PC was a technical masterpiece (although, in real world performance, was a failure). OS/2 was a great operating system too. But, by then their mistakes were carved in stone, and the rest of the world is now suffering from them. Sub-optimal microprocessors from AMD and Intel because of horrible instruction set, and a miserable OS from Microsoft that is bloated, buggy, and slow. Compaq, by virtue of it's irrelevance, bears no responsibility for that. IBM does, with arguably the worst choices ever made by a company, when considering the consequences of them.
Chunga29 - Thursday, July 19, 2007 - link
Dude, you've got SERIOUS issues. I won't argue about whether or not Compaq was the mother of the business PC or not, but they were at least a player and basically you're having a fit about a short sentence that serves as the intro to the article. HP is now the biggest business PC maker I believe (barely ahead of Dell), so at least they're somewhat relevant. Anyway, get past the intro and read the article rather then going off on a little comment that was basically there to try and get people interested (or in your case perhaps, riled up).TA152H - Thursday, July 19, 2007 - link
Yes, my issue is I dislike misinformation and revisionism. Yours is you do not. Which do you think is worse?I never said anything bad about HP. Compaq being the mother of the PC though, it just wrong. If no one points it out, then it tacitly is accepted as true, and it was not. If you think truth is irrelevant, then we have a fundamental difference of opinion.
Chunga29 - Thursday, July 19, 2007 - link
This isn't even revisionism, I don't think. I mean come on, Johan is over in the Netherlands testing this stuff, and it was just an intro. The main gist (I believe) is that HP/Compaq and IBM are trying to push PCs out of the business sector. The conclusion is that at present they're still pushing proprietary, expensive technology that really doesn't benefit *most* companies. There are instances where it could be useful, but for most large companies a cheaper PC is still easy enough to support.I work at a fortune 500 company with thousands of PCs and laptops throughout the corporation. I haven't heard much about anyone pushing blades for us (though I'm sure IBM has tried). We have about 200 PCs and 20 laptops at my location, and about 10 spares. The spares are imaged and ready to roll in the computer room. If a PC has issues, we go out, swap PCs, take the old PC back and start troubleshooting. Takes about 15 minutes, 10 of which involves us carting the PC from the data center to the desk of whoever needs it.
FWIW, Compaq did create the "luggable brick" PC, didn't they? I have fond memories of playing Rogue on a small 4-6" screen with a fold-down keyboard. Hahaha... those weighed about 40 pounds, I think! You still have major isses, though (as you indicate above). I mean, seriously, who gives a rip about whether or not Compaq was one of the major founders of the PC world? They were, along with many others, but it's pretty irrelevent.
I'd say Intel is the father (hardware), Microsoft is the mother (software), and IBM is basically the preist that married the two and then everyone more or less ignored. That you're one of the people who think PS/2 was a great system speaks volumes in my book. It was expensive, not truly much faster, filled with proprietary parts, and in the PC world it was doomed to failure. It was basically IBM trying to put the wine back in the bottle after pouring, and the market rejected the idea.
I remember doing some work on a guy's PS/2, and I was shocked at how much he paid for so little. $4000 or something crazy, and my little old $2000 clone 386 could run circles around it. Sure, it had SCSI (I think?), but a faster HDD subsystem with less RAM, a slower CPU, and all the other junk was meaningless. I think the biggest contribution the PS/2 made to the computer world is the PS/2 keyboard and mouse adapters!
JohanAnandtech - Thursday, July 19, 2007 - link
Than I am sorry I overemphasized the role of compaq, but I think we can agree that this article was not about giving credit to those who layed the foundation for the personal computer.The purpose was to show that both HP and IBM/Lenovo, IMHO the most important players in the pc industry (from business desktop to pc server) were thinking out alternatives to the pc.
Thanks for the historical insight.
TA152H - Thursday, July 19, 2007 - link
I still have a fundamental problem with giving Compaq any credit for the foundation of the personal computer, because they were one of many clones, but they did sell to businesses, so I guess that counts for something (what I'm not sure of).HP was a much more important company than Compaq, not so much for their PCs, which were essentially clones and not particularly original, but they did a lot of work on CISC based workstations (68030) and operating systems for them, as well as their own stuff. I have nothing at all against HP, they were and are a very important company and have pioneereed a lot of interesting technology, currently the Itanium. Also, their impact on printers can not be ignored, as they became at one point a de facto standard.
Compaq, invented?
With regards to the article, it was nice to see the Eden processors mentioned. I wish they received more attention than they do, although I do have an 800 MHz processor and it's a complete failure on every level. It uses as much power as a K6-III+ at 600 MHz, and underperforms it at everything. For that reason I stuck with my K6-III+s, which are capable of changing multiplier without even rebooting the machine.
However, now they have some 3.5 watt 1 GHz model which completely changes things. Their 1.5 GHz 7 watt model is also very impressive, and, I think, deserves a lot more press than it gets. These are certainly not speed kings, but for a lot of functions are perfectly adequate while having extraordinary power use.
One thing about HP and IBM is, they are both technology companies, and I think have an interest in kicking out the commodity market as much as possible. Dell, being little more than a distributor, has an interest in commoditizing as much as possible. So, it's not altogether surprising that companies with the technical ability to create new products would attempt to do so, in an attempt to usurp the usefulness of commodity products. It's almost surprising it hasn't taken off more than it has, because most of the time PCs on a business LAN are so restricted that it almost is irrelevant if they have a hard disk, except for the fact they whine and use power. I'm not even sure the term business PC makes any sense, it's something of an oxymoron.
As they say, the more things change, the more things remain the same. This is not so different from mainframes with 3270s attached to them. The processing can still be done on the client though, so in that respect it is. Certainly, however, things are a lot more centralized than they were 15 years ago. Good grief, I wonder if PL/1 will make a comeback. Oh, the humanity!
stmok - Thursday, July 19, 2007 - link
Compaq made the PC clone possible by reverse engineering IBM's BIOS. (clean room approach).So they did have some influence as to what has become today's PC. (or PC Clone).
Then again, the only use for the BIOS nowadays is for bootstrapping until the OS's drivers take over. (except for ACPI, you still need BIOS for that).
fic2 - Thursday, July 19, 2007 - link
So, Compaq were more like the long lost uncle that comes into town after a few years to take pictures of the niece/nephew. Nothing to do with the conception or raising the kid, just bragging that if he wanted to he could have done the same.bob661 - Thursday, July 19, 2007 - link
I'm still back on people being able to do CAD apps from 2000 miles away. LOL! The hardware may be there and the bandwidth may be there but there's NO WAY any company can guarantee that you'll get less than 20 ms of latency on an internet connection. Especially the more remote your users are. Are you willing to bet your engineering departments productivity on a promise from your ISP? I'm not. I'll pay the horrendous costs (LOL again!) for a dedicated desktop PC that's NOT at the mercy of an external company.BTW, if you want to eliminate viruses and such from the companies network, there's a free solution for this: IT'S CALLED ENFORCING COMPANY POLICY! This involves (gasp) training and if that fails, taking disciplinary action. If they STILL don't get it, fire them! It's FAR cheaper to fire an employee that's not productive (and inducing viruses on a the company's network reduces productivity) than it is to come up with some multi-million dollar scheme (thin clients and their respective high dollar IT admins) to outsmart them and put everyone's productivity at risk.