• Tag Archives HP

  • Digital Archaeology: HP Pavilion p7-1012

    Posted on by Azrael Comment

    The Hewlett Packard p7-1012 is a great example of why you don’t want to buy “big box” store computers off the shelf. On the one hand, it looks ok and while it has a fairly low-end CPU even for the time, there is at least a fairly substantial upgrade path. However, on the other hand, it only has a 250 watt power supply and while it can be upgraded from a dual-core cpu to a hexa-core CPU, the top of the range is off limits because of motherboard power restrictions. Worst of all, this computer has no PCI x16 slot. That limits you to something like the nVidia GT 710 if you are thinking about a video card upgrade. Given the pitiful power supply, you couldn’t do a whole lot better even if you did have an x16 slot.

    Nevertheless, even with its relatively low end dual-core Athlon II 220 CPU, with the RAM maxed out at 8 GB it still makes a useable Windows 10 or Linux box provided you don’t want to do anything too intensive (or anything resembling modern gaming…”modern” in this case being 2009 or later). Because of that, I’m not sure this really quite fits into the category of “vintage” or even “retro” but it’s getting close. Generally, I consider anything older than 10 years to be getting close to the “retro” category when it comes to computers. This one fits with that criteria but just barely. A better criteria might be anything with a single core CPU or anything with a 32-bit CPU, etc. I guess it is pretty subjective though.

    Some more detailed stats:

    • Motherboard – FOXCONN 2AB7 / H-Apricot-RS780L-uATX (Apricot) (Socket AM3)
    • Chipset – AMD 780G
    • CPU – AMD Athlon II X2 220
    • Graphics – AMD Radeon 3000 (RS780L) / Taurus (integrated)
    • RAM – 8 GB DDR3-1066
    • Hard Drive – Seagate ST3320413AS 7200 RPM, 320 GB
    • Optical Drive – hp DVD-RAM GH60L

    For more detailed specs, check out the full output of HWiNFO.

    The AMD 780G chipset was introduced in 2009 with an integrated graphics solution that was meant to provide low cost DirectX 10.1 graphics capabilities to value PCs. While the chipset itself supports one PCIe 2.0 x16 slot, apparently this low end AMD solution wasn’t cheap enough for HP so they didn’t even provide that on their implementation. Also, while the chipset itself was designed to support AM3+ CPUs (or at least able to support them), this motherboard will only accept AM3 CPUs with a max 95 watt TDP. This means the best processor you can put in it is a Phenom II X6 1065T. The 1075T and beyond have a TDP of 125 watts. Still, I suppose the 1065T would be a reasonably large upgrade over the Athlon II X2 220 that is in it.

    Speaking of which, The Athlon II X2 220 is an AM3 processor with a 65 watt TDP and 2 cores running at 2.8 GHz. This Athlon II is one of a few that have only 1 MB of L2 cache (most have 2 MB) so they pretty much picked near the lowest end for an already lower end CPU line. The Athlon II is essentially the same as the Phenom II but with less cache. The Phenom II has 6 MB of L3 cache whereas the Athlon II has none. However, the Athlon II did have 2 MB of L2 cache (well, most of them anyway) instead of the 1 MB of L2 cache that the Phenom II had. I guess it could have been worse…They could have put a single core Sempron in it. Generally speaking, I think the Athlon II would have been a direct competitor to the Core 2 Duo era Pentium which was essentially a Core 2 Duo with less cache. The Athlon II was somewhat slower per MHz but also cheaper.

    As far as the graphics capabilities go, I don’t think there has ever existed an integrated motherboard solution that has risen above absolute crap when it comes to gaming. I don’t know exactly what discrete solution the Radeon 3000 IGP would be closest to but at least according to Techpowerup, it is nearly 10x slower than the GeForce GT 430 and Radeon HD 5570. That combined with the aforementioned lack of PCI x16 slot and pitiful power supply made this pretty useless for gaming purposes even when it was brand new. I guess as a relatively low power office machine or for basic web browsing/e-mail/office tasks at home it would work well enough though.

    Like any computer I turn on, I installed BOINC on this one. Being a 64-bit CPU, this CPU is able to process work units from most projects, including the ones I participate in most including Einstein@home, MilkyWay@home, Rosetta@home, Universe@home and World Community Grid. Click on the links to see how it is doing in those projects. Sadly, most rosetta@home work units require VirtualBox and gobs of memory and “regular” work units are few and far between. World Community Grid has also been down longer than expected after a recent move. Check out BOINC Stats or Free-DC for overall BOINC statistics for this computer.


    📂This entry was posted in Computer Arcana 📎and tagged

  • Digital Archaeology: HP Pavilion DV2700

    Posted on by Azrael Comment

    The Pavilion DV2700 is laptop that was released by Hewlett Packard in the 2008 time frame. However, like many of their models, it refers to a fairly wide variety of hardware configurations. In fact, there were versions of the DV2700 with both Intel and AMD processors. However, despite the different hardware configurations, the physical design was the same. The features that stick out were ones to make watching movies more convenient. The DV2700 has a row of touch sensitive buttons with blue backlighting along the top for controlling DVD playback and volume. There was a wide variety of HP models from this time period that had these features and the same basic physical design overall.

    Hardware in this particular DV2700 includes the following:

    • CPU: AMD Turion 64 X2 Mobile TL-60 @ 2GHz, Stepping BH-G2, Codename Tyler
    • Memory: 2GB DDR2
    • Graphics: nVidia GeForce 7150M/nForce 630M
    • Motherboard: Wistron 30D6
    • Chipset: nVidia nForce 560
    • Screen: 1280×800
    • Hard Drive: WDC WD3200LPVX-75V0TT0, 320 GB, 5400 RPM, 8 MB cache, SATA 3.0

    While AMD has leapfrogged Intel from time to time in terms of performance on the desktop, they’ve had a harder time competing in lower power spaces like laptops. The Turion was essentially AMD’s answer to the Core Duo and Core 2 Duo. While it competed pretty evenly performance wise, at least for a while, it wasn’t able to quite match Intel in terms of lower power consumption. This CPU performed roughly the same as a 1.9 GHz Core Duo.

    There were two versions of the TL-60. The first was code-named Trinidad, built on a 90nm process and had a 35 watt TDP. The second version (the one in this laptop) was code-named Tyler, built on a 65nm process and had a 31 watt TDP. Otherwise they were essentially identical. These were based on the K8 architecture with minor modifications so they were more closely related to the Athlon 64 line than the Phenom line.

    The chipset, or more specifically the built-in GPU made it difficult to find drivers for. Depending on the OS, they seemed to be non-existent or a closely guarded secret. I first tried installing a recent version of Ubuntu and failed miserably. All I could get was a blank screen (or maybe it was a scrambled screen…I forget now). I spent quite some time trying to get it to work but couldn’t. I tried the default Noveau drivers and every nVidia driver version I could find and nothing worked. After trying another distribution or two, I finally got Debian to work…kind of. It still required manual alteration of config files and I could only get a 1024×768 resolution instead of the native 1280×800. That seemed to be the best I was going to get with Linux. Perhaps with an old enough distribution and old enough nVidia drivers I would have better luck…

    Instead, I decided to give Windows a try. Originally, this laptop came with Windows Vista (almost certainly the 32-bit version despite having a 64-bit chip). It also, in theory, supported Windows XP. And it was also likely possible to install Windows 7 on it. Windows 10 might work with more RAM…maybe…but I wasn’t going to put in more RAM at the moment and even if I did it would probably be painfully under-performing. I didn’t really want Vista and while I love XP I felt something a bit newer was better for this laptop so I decided to go with Windows 7.

    Installation went fine and it was immediately operable. But again, the resolution wasn’t right and it didn’t have correct drivers for the video (and a few other things). Finding the right nVidia drivers turned out to be more difficult than it should have been. Normally, you can go to nVidia’s web site, choose the appropriate category and find the drivers you need. However, the 7150M/630M seemed to be a black sheep of the family or something. No selection combination produced drivers that were appropriate. I finally found a link somewhere else for the correct drivers (back to an apparently obscure nVidia page no less) and once I installed them, things worked fine and I could finally get that 1280×800 resolution.

    It wasn’t too long before this chipset was released that AMD bought ATI. That probably explains to some degree why this chipset had a relatively short life and little support. That and it really wasn’t that great of a performer, at least as far as the GPU was concerned. It was better than Intel’s built-in graphics of the same time period but that really isn’t saying much. It only supported DirectX 9.0c and there’s only so much you can do with 2 pixel shaders, 1 vertex shader, 2 texture mapping units and 2 ROPs. Even at the time that was pretty lousy.

    When I acquired this laptop, it had 2 GB of RAM. However, given the fact that the two modules were different brands and had different specs, I’m pretty sure it came stock with 1 GB when new and was later upgraded. One GB seems a less than ideal amount for Windows Vista. Even 2 GB can be a struggle with modern web browsing on Windows 7 and Windows 7 is a bit more efficient with RAM than Vista. Having said that, I know that some computers that came with Vista (basic) only had 512 MB. Sounds like torture to me.

    For detailed specs, see https://www.megalextoria.com/DigitalArchaeology/da_Pavilion-DV2700/PAVILION-DV2700.HTM


    📂This entry was posted in Computer Arcana 📎and tagged

  • Stupid Patent of the Month: HP Patents Reminder Messages

    Posted on by Azrael 1 Comment

    On July 25, 2017, the Patent Office issued a patent to HP on reminder messages. Someone needs to remind the Patent Office to look at the real world before issuing patents.

    United States Patent No. 9,715,680 (the ’680 patent) is titled “Reminder messages.” While the patent application does suggest some minor tweaks to standard automated reminders, none of these supposed additions deserve patent protection.

    Claim 1 of the patent states (comments in brackets):

    A non-transitory computer-readable storage medium containing instructions, the instructions when executed by a processor causing the processor to [use a computer to]:

    receive at a first computer system, via a network, event data descriptive of an event to occur at an event time [get event and time information];

    receive via the network, reminder data descriptive of a reminder time to occur on or before the event time [get the reminder time];

    at a time after receipt of the event data, receive via the network article data descriptive of an article to be associated with the event, the article data created during an electronic scanning operation [receive some additional information (created by scanning) relating to the event]; and

    at the reminder time send via the network a reminder message describing the event and the article to a second computer system, for presentation at the second computer system [at the reminder time, send the reminder message].

    Although this claim uses some obscure language (like “non-transitory computer-readable storage medium” and “article data”), it describes a quite mundane process. The “article data” is simply additional information associated with an event. For example, ‘buy a cake’ might be included with a birthday reminder. The patent also requires that this extra information be input via a “scanning operation” (e.g. scanning a QR code).

    The ’680 patent comes from an application filed in July 2012. It is supposed to represent a non-obvious advance on technology that existed before that date. Of course, reminder messages were standard many years before the application was filed. And just a few minutes of research reveals that QR codes were already used to encode information for reminder messages. For example, QRickit suggested using QR codes for calendar events and reminders (with the option of adding additional information beyond the event descriptor). This 2011 article suggests using QR codes to embed information such as “assignments for the week.” The only even arguable difference from the prior art is that the patent’s claims require the “article data” to be received after the event data. In our view, that is not a distinction that warrants the government-granted monopoly power inherent in a patent.

    The Patent Office reviewed HP’s application for years without ever considering any real-world products. Indeed, the examiner considered only patents and patent applications. We have complained before that the Patent Office seems to operate in an alternative universe where only patents provide evidence of the state of the art in software. The fact that the Patent Office doesn’t take developments in real software into account in its assessment of prior art speaks poorly for its ability to determine whether patent applications actually reflect new inventions.

    In addition to failing to consider real products, the Patent Office gives little weight to common sense and takes an extremely rigid approach to evaluating whether or not a patent application is obvious. This leads to patents on things like taking photos against a white background, filming a yoga class, voting for a favorite photo, and out-of-office email. Much of the responsibility for this mess rests with the Federal Circuit, which has failed to apply a Supreme Court case called KSR v. Teleflex that calls for a flexible, common sense approach to obviousness. Together with Public Knowledge, EFF recently filed an amicus brief [PDF] asking the Supreme Court to consider the obviousness standard in patent law and to reaffirm that examiners can reject common sense combinations of known elements.

    Even leaving obviousness aside, HP’s patent application still should have been rejected under Alice v. CLS Bank. In Alice, the Supreme Court ruled that an abstract idea does not become eligible for a patent simply by being implemented on a generic computer. As with many software patents, the patent goes out of its way to explain that its method can be implemented on a generic computer, or, as the patent puts it “generally any computer.” Despite this, the prosecution history [PDF] reveals that the examiner never even mentioned Alice, even in office actions written well after the Supreme Court’s decision came down. We have written many times (e.g. 1, 2, 3, and 4) to protest that the Patent Office is not doing enough to diligently apply the Alice decision. The ’680 patent provides yet another example of abstract software patents being issued despite the Supreme Court’s ruling.

    In case you want to set a reminder, the ’680 patent will expire on December 16, 2035.

    Source: Stupid Patent of the Month: HP Patents Reminder Messages | Electronic Frontier Foundation


    📂This entry was posted in News and Politics 📎and tagged