issue120:labo_linux1
Différences
Ci-dessous, les différences entre deux révisions de la page.
| Les deux révisions précédentesRévision précédenteProchaine révision | Révision précédente | ||
| issue120:labo_linux1 [2017/05/15 08:18] – d52fr | issue120:labo_linux1 [2017/05/16 14:46] (Version actuelle) – andre_domenech | ||
|---|---|---|---|
| Ligne 7: | Ligne 7: | ||
| At the time, I was still paying back student loans and couldn’t afford to replace the system with anything close to what I had, so I picked up a used Pentium II for $150CDN (which was actually a good deal at the time). The used computer came with no operating system – which wasn’t an issue since I had several FreeBSD discs and a number of Linux distributions.** | At the time, I was still paying back student loans and couldn’t afford to replace the system with anything close to what I had, so I picked up a used Pentium II for $150CDN (which was actually a good deal at the time). The used computer came with no operating system – which wasn’t an issue since I had several FreeBSD discs and a number of Linux distributions.** | ||
| - | En 2002, j' | + | En 2002, j' |
| - | Puis une tragédie arriva, un court-circuit causé par un aspirateur branché sur une mauvaise prise murale grilla non seulement l' | + | Puis une tragédie arriva, un court-circuit causé par un aspirateur branché sur une mauvaise prise murale grilla non seulement l' |
| - | A ce moment-là, j' | + | A ce moment-là, j' |
| **Whether you’re in a rut, or helping someone else who’s in a rut, or just wondered how badly old technology fares against something newer, this is the article that will hopefully put things in perspective. For this hardware showdown I selected a range of mostly older hardware available on the used market. I also tried selecting a mix of commodity hardware and custom-built hardware across lower and higher-end systems. There’s nothing brand new in the mix and nothing with a solid state drive. Here’s a low-down of the hardware: | **Whether you’re in a rut, or helping someone else who’s in a rut, or just wondered how badly old technology fares against something newer, this is the article that will hopefully put things in perspective. For this hardware showdown I selected a range of mostly older hardware available on the used market. I also tried selecting a mix of commodity hardware and custom-built hardware across lower and higher-end systems. There’s nothing brand new in the mix and nothing with a solid state drive. Here’s a low-down of the hardware: | ||
| Ligne 22: | Ligne 22: | ||
| • Lenovo Thinkcentre M82 - Intel Core i5-3470 @ 3.60GHz (4 cores), 4GB RAM, 500GB HDD, Intel HD 2500 Graphics (onboard)** | • Lenovo Thinkcentre M82 - Intel Core i5-3470 @ 3.60GHz (4 cores), 4GB RAM, 500GB HDD, Intel HD 2500 Graphics (onboard)** | ||
| - | Si vous êtes dans une impasse, ou si vous aidez quelqu' | + | Si vous êtes dans une impasse, ou si vous aidez quelqu' |
| - | • Dell Optiplex 745 - Pentium D @ 3,4 GHz (2 cœurs), 2 Go de RAM, disque dur de 80 Go, processeur graphique Intel 965 (intégré à la carte-mère) | + | • Dell Optiplex 745 - Pentium D @ 3,4 GHz (2 cœurs), 2 Go de RAM, disque dur de 80 Go, processeur graphique Intel 965 (intégré à la carte-mère). |
| - | • Personnalisé - Phenom II 955 @ 3,2 GHz (3 cœurs), 4 Go de RAM, disque dur de 500 Go, processeur graphique Radeon HD 4290 (intégré) | + | • Personnalisé - Phenom II 955 @ 3,2 GHz (3 cœurs), 4 Go de RAM, disque dur de 500 Go, processeur graphique Radeon HD 4290 (intégré). |
| - | • Personnalisé - AMD Athlon 64 X2 4400+ @ 2,30 GHz (2 cœurs), 4 Go de RAM, disque dur de 160 Go, carte graphique Radeon HD 4450 (1 Go) | + | • Personnalisé - AMD Athlon 64 X2 4400+ @ 2,30 GHz (2 cœurs), 4 Go de RAM, disque dur de 160 Go, carte graphique Radeon HD 4450 (1 Go). |
| - | • Dell Precision T3500 - Xeon W3520 @ 2,67 GHz (4 cœurs), 8 Go de RAM, disque dur de 500 Go, carte graphique NVidia Quadro FX 580 (512 Mo) | + | • Dell Precision T3500 - Xeon W3520 @ 2,67 GHz (4 cœurs), 8 Go de RAM, disque dur de 500 Go, carte graphique NVidia Quadro FX 580 (512 Mo). |
| - | • HP 6000 Pro SFF - Intel Core 2 Duo E8400 @ 3,00 GHz (2 cœurs), 4 Go de RAM, disque dur de 250 Go, processeur graphique Intel 4 (intégré) | + | • HP 6000 Pro SFF - Intel Core 2 Duo E8400 @ 3,00 GHz (2 cœurs), 4 Go de RAM, disque dur de 250 Go, processeur graphique Intel 4 (intégré). |
| - | • Lenovo Thinkcentre M82 - Intel Core i5-3470 @ 3,60 GHz (4 cœurs), 4 Go de RAM, disque dur de 500 Go, processeur graphique Intel HD 2500 (intégré) | + | • Lenovo Thinkcentre M82 - Intel Core i5-3470 @ 3,60 GHz (4 cœurs), 4 Go de RAM, disque dur de 500 Go, processeur graphique Intel HD 2500 (intégré). |
| **I fully expected the 3.60GHz 4 core i5 to win at pretty much every benchmark. Of course synthetic benchmarks don’t say everything about a machine, but they can produce some interesting results. Phoronix Test Suite has been my goto benchmarking system for several years, but there are over 200 possible tests in the suite. For the showdown, I chose the following Phoronix tests: | **I fully expected the 3.60GHz 4 core i5 to win at pretty much every benchmark. Of course synthetic benchmarks don’t say everything about a machine, but they can produce some interesting results. Phoronix Test Suite has been my goto benchmarking system for several years, but there are over 200 possible tests in the suite. For the showdown, I chose the following Phoronix tests: | ||
| Ligne 39: | Ligne 39: | ||
| • Ut2004-demo - is a measurement of older 3D operations using the Unreal Tournament 2004 engine. I chose this test over others because it ran on all the hardware.** | • Ut2004-demo - is a measurement of older 3D operations using the Unreal Tournament 2004 engine. I chose this test over others because it ran on all the hardware.** | ||
| - | Je m' | + | Je m' |
| - | • Encodage x264 - Ceci mesure combien | + | • Encodage x264 - Ceci mesure combien d' |
| - | • RAMSpeed (Vitesse de la RAM) - mesure les Mégaoctets | + | • RAMSpeed (Vitesse de la RAM) - mesure les mégaoctets |
| - | • Aio-Stress - est un comparatif des entrées/ | + | • Aio-Stress - C'est un comparatif des entrées/ |
| - | • X11perf - est une mesure pour des opérations en 2D, en Opérations | + | • X11perf - C'est une mesure pour des opérations en 2D, en opérations |
| - | • Ut2004-demo - est une mesure d' | + | • Ut2004-demo - C'est une mesure d' |
| ======2====== | ======2====== | ||
| Ligne 56: | Ligne 56: | ||
| - | Pour les tests X11perf et ut2004, j'ai utilisé le même écran Del P2213t qui a une résolution maximale de 1680x1050. j'ai testé Unreal Tournament à 800x600 (à peine le minimum pour jouer). 1024x768 et 1680x1050. J'ai essayé d' | + | Pour les tests X11perf et ut2004, j'ai utilisé le même écran Del P2213t qui a une résolution maximale de 1680x1050. j'ai testé Unreal Tournament à 800x600 (à peine le minimum pour jouer), 1024x768 et 1680x1050. J'ai essayé d' |
| Étant donné que le i5-3470 est un processeur i5 de troisième génération, | Étant donné que le i5-3470 est un processeur i5 de troisième génération, | ||
| - | Comme attendu, le i5 réalise le meilleur encodage vidéo, bien que la station Xeon « hyperthreadée » (sous-processus en parralèle) ne soit pas loin derrière. La vrai différence entre les stations i5 et Xeon n'est que de 9 Mo/s. Voici (ci-dessous) les valeurs réelles de chaque machine. | + | Comme attendu, le i5 réalise le meilleur encodage vidéo, bien que la station Xeon « hyperthreadée » (sous-processus en parallèle) ne soit pas loin derrière. La vraie différence entre les stations i5 et Xeon n'est que de 9 Mo/s. Voici (ci-dessous) les valeurs réelles de chaque machine. |
| **Our AMD Phenom had one core switched off, but came nowhere near either the Xeon or i5 processors. It outperformed the 3GHz Core 2 Duo by 18 FPS, which is significant if you’re ripping DVDs or simply transcoding video files. Even though the custom Athlon 64 X2 has a 1GB video card, that card didn’t help with the FPS score (as it might in a real test with OpenCL enabled - ripping and encoding using Handbrake). I mentioned earlier that this test is supposed to be strictly a CPU test and the numbers seem to reflect this. | **Our AMD Phenom had one core switched off, but came nowhere near either the Xeon or i5 processors. It outperformed the 3GHz Core 2 Duo by 18 FPS, which is significant if you’re ripping DVDs or simply transcoding video files. Even though the custom Athlon 64 X2 has a 1GB video card, that card didn’t help with the FPS score (as it might in a real test with OpenCL enabled - ripping and encoding using Handbrake). I mentioned earlier that this test is supposed to be strictly a CPU test and the numbers seem to reflect this. | ||
| Ligne 68: | Ligne 68: | ||
| The x264 test was pretty much as expected, but the RAMSpeed test (below) shows a lot more variation.** | The x264 test was pretty much as expected, but the RAMSpeed test (below) shows a lot more variation.** | ||
| - | Notre Phenom d'AMD avait un cœur à l' | + | Notre Phenom d'AMD avait un cœur à l' |
| - | Quand j'ai choisi ces machines, j'ai essayé de choisir des machines | + | Quand j'ai choisi ces machines, j'ai essayé de choisir des ordinateurs |
| Le test x264 est bien comme attendu, mais le test RAMSpeed (ci-dessous) montre beaucoup plus de fluctuations. | Le test x264 est bien comme attendu, mais le test RAMSpeed (ci-dessous) montre beaucoup plus de fluctuations. | ||
| Ligne 79: | Ligne 79: | ||
| As with the other hardware, there was a real mixed amount of hard drives in the various systems, but generally lower end machines had smaller-sized drives. The smallest drive was in the Optiplex, which had a tiny 80GB drive, and the largest in the Phenom, Xeon and i5 which all had 500GB drives. To provide a bit more variety, the Athlon 64 had a 160GB hard drive and the HP 6000 had a 250GB hard drive. The previous page (bottom right) has the results of the aio-stress test.** | As with the other hardware, there was a real mixed amount of hard drives in the various systems, but generally lower end machines had smaller-sized drives. The smallest drive was in the Optiplex, which had a tiny 80GB drive, and the largest in the Phenom, Xeon and i5 which all had 500GB drives. To provide a bit more variety, the Athlon 64 had a 160GB hard drive and the HP 6000 had a 250GB hard drive. The previous page (bottom right) has the results of the aio-stress test.** | ||
| + | |||
| + | Gardez en tête que je n'ai testé qu'un aspect : un test de moyenne en virgule flottante. Si j' | ||
| + | |||
| + | Je me serais attendu à ce que le Thinkcentre i5 gagne du fait de son architecture DDR3 plus récente, mais les deux stations Phenom et Xeon battent le i5. Alors que le Xeon a plus de RAM, ce supplément de RAM n'est pas la clé pour le meilleur résultat. Le Phenom et le i5 ont la même quantité de RAM DDR3. La RAM du i5 est aussi plus rapide que celle du Phenom, avec une horloge à 1600 MHz contre 1333 MHz pour le Phenom. La différence semble venir du fait que le i5 opère avec une carte RAM unique de 4 Go là où le Phenom a 2 cartes RAM de 2 Go fonctionnant en mode double canal. De la même façon, la vitesse de la RAM du Precision T3500 n'est que de 133 MHz, mais elle fonctionne aussi en mode double canal, alors que ce sont deux cartes de 4 Go. Le Xeon a un résultat nettement meilleur que les i5 et Phenom. Je ne saurais pas dire si ceci est dû au 4 Go supplémentaires ou à l' | ||
| + | |||
| + | Pour les autres matériels, il y a un vrai mélange de tailles de disques durs dans les différents systèmes, mais en général, les machines bas de gamme ont des disques de plus petite taille. Le plus petit disque est dans l' | ||
| ======3====== | ======3====== | ||
| Ligne 85: | Ligne 91: | ||
| One thought was that perhaps the i5 Thinkcentre had a bad drive, but all drives underwent a basic SMART test and tested good. I’m not sure why the i5 drive did so poorly and the only co-relationship I can draw is that both the two worst performing drives are manufactured by Seagate. I actually love Seagate drives and have 2 x 3TB and a 2TB drive in our KODI server and all have performed well over a couple of years. Newness doesn’t seem to matter either as the Phenom’s 500GB hard drive (Western Digital) got edged out by the 160GB Western Digital drive in the Athlon 64 X2. The i5 Seagate drive is rated at 7200 RPMs, 6 GB/s and has a 16MB cache. The 160GB drive in the Athlon 64 X2 is also a 7200RPM drive and only has an 8MB cache. Clearly something was bottlenecking the i5 drive. Then there’s the Precision T3500 drive, a Hitachi HDS72105. I was able to explain this crazy result by looking at the results of other machines on openbenchmarking.org. Looking at all the results, Xeon-based computers seemed to fare much better than any other system, including i5’s with Solid State Drives (SSDs). The architecture of the Xeon seems to explain why it did much better than the rest. In those openbenchmarking systems all the systems were in the thousand plus range while other systems were in the hundreds or sub-hundred range.** | One thought was that perhaps the i5 Thinkcentre had a bad drive, but all drives underwent a basic SMART test and tested good. I’m not sure why the i5 drive did so poorly and the only co-relationship I can draw is that both the two worst performing drives are manufactured by Seagate. I actually love Seagate drives and have 2 x 3TB and a 2TB drive in our KODI server and all have performed well over a couple of years. Newness doesn’t seem to matter either as the Phenom’s 500GB hard drive (Western Digital) got edged out by the 160GB Western Digital drive in the Athlon 64 X2. The i5 Seagate drive is rated at 7200 RPMs, 6 GB/s and has a 16MB cache. The 160GB drive in the Athlon 64 X2 is also a 7200RPM drive and only has an 8MB cache. Clearly something was bottlenecking the i5 drive. Then there’s the Precision T3500 drive, a Hitachi HDS72105. I was able to explain this crazy result by looking at the results of other machines on openbenchmarking.org. Looking at all the results, Xeon-based computers seemed to fare much better than any other system, including i5’s with Solid State Drives (SSDs). The architecture of the Xeon seems to explain why it did much better than the rest. In those openbenchmarking systems all the systems were in the thousand plus range while other systems were in the hundreds or sub-hundred range.** | ||
| + | |||
| + | Ce graphique est beaucoup plus difficile à interpréter car les résultats sont biaisés. J'ai lancé aio-stress plusieurs fois sur le système Dell Precision T3500 à base de Xeon, et, à chaque fois, le résultat était au-dessus de 1200 Mo/s. Les résultats de toutes les autres machines pour ce test étaient inférieurs à 100, avec l' | ||
| + | |||
| + | Une idée était que, peut-être, le Thinkcentre i5 avait un mauvais disque, mais tous les disques avaient passé un test SMART de base et paraissaient bons. Je ne suis pas sûr de la raison pour laquelle le disque du i5 était si mauvais et la seule corrélation que j' | ||
| **Graphics are an integral part of any system, and 2D graphics, the X11 subsystem is important to daily desktop-Linux users. The x11perf, like the RAMspeed test, has several tests embedded in the single test. I chose to focus on the 500x500 pixel scroll test. There were some surprising and not-so-surprising results (shown bottom left). | **Graphics are an integral part of any system, and 2D graphics, the X11 subsystem is important to daily desktop-Linux users. The x11perf, like the RAMspeed test, has several tests embedded in the single test. I chose to focus on the 500x500 pixel scroll test. There were some surprising and not-so-surprising results (shown bottom left). | ||
| Ligne 91: | Ligne 101: | ||
| Given that these are the only two AMD-based systems on the list I wondered if it had something to do with older AMD architecture, | Given that these are the only two AMD-based systems on the list I wondered if it had something to do with older AMD architecture, | ||
| + | |||
| + | L' | ||
| + | |||
| + | La meilleure machine était le Precision T3500 avec sa carte vidéo NVidia Quadro FX 580. Bien que le FX 580 n'ait pas la plus grande quantité de RAM vidéo dans ce comparatif (la carte Radeon avec 1 Go de l' | ||
| + | |||
| + | Étant donné que ce sont les deux seuls systèmes de la liste basés sur AMD, je me demande si ça avait quelque chose à voir avec l' | ||
| **With that in mind, I loaded up the Unreal Tournament 2004 demo and ran it on each machine. Below right are the results at 800x600. | **With that in mind, I loaded up the Unreal Tournament 2004 demo and ran it on each machine. Below right are the results at 800x600. | ||
| As expected, the onboard graphics of the Pentium D-based Optiplex 745 fared the worst with only 23.16 FPS at 800x600. I expected the Athlon 64 with it’s 1GB video card to do much better than it did. The HP 6000 Pro’s onboard video was almost 11 FPS better than the Athlon 64 which leads me to believe the card is pretty much wasted in this system. The Phenom fared much better in the Unreal Tournament test than it did in the x11perf test, helping the theory that the motherboard architecture of the Athlon 64 is at least partly to fault for the crippled card. Both AMD systems use Asus motherboards, | As expected, the onboard graphics of the Pentium D-based Optiplex 745 fared the worst with only 23.16 FPS at 800x600. I expected the Athlon 64 with it’s 1GB video card to do much better than it did. The HP 6000 Pro’s onboard video was almost 11 FPS better than the Athlon 64 which leads me to believe the card is pretty much wasted in this system. The Phenom fared much better in the Unreal Tournament test than it did in the x11perf test, helping the theory that the motherboard architecture of the Athlon 64 is at least partly to fault for the crippled card. Both AMD systems use Asus motherboards, | ||
| + | |||
| + | Avec ceci en tête, j'ai chargé la demo de Unreal Tournament 2004 et l'ai lancé sur chaque machine. Les résultats en 800x600 sont en bas à droite. | ||
| + | |||
| + | Comme prévu, le module graphique intégré de l' | ||
| ======4====== | ======4====== | ||
| Ligne 101: | Ligne 121: | ||
| Generally speaking most games are acceptable at 30 FPS. Some people argue they can’t live without 60 FPS, but most people tend to forget FPS once immersed in a game. This would rule out the Dell Optiplex 745 as a gaming machine, and as a refurbisher, | Generally speaking most games are acceptable at 30 FPS. Some people argue they can’t live without 60 FPS, but most people tend to forget FPS once immersed in a game. This would rule out the Dell Optiplex 745 as a gaming machine, and as a refurbisher, | ||
| + | |||
| + | Malheureusement, | ||
| + | |||
| + | Pour généraliser, | ||
| + | |||
| **At 1680x1050 (previous page, bottom right), our Frames Per Second drops the most across machines (again except the Athlon 64 X2 which saw a slight increase). Hardest hit is the leader of the pack, the Xeon with the NVidia Quadro FX card, which suffered a 38 FPS drop. At 1680x1050, three of our six test systems fall below the 30 FPS minimum. | **At 1680x1050 (previous page, bottom right), our Frames Per Second drops the most across machines (again except the Athlon 64 X2 which saw a slight increase). Hardest hit is the leader of the pack, the Xeon with the NVidia Quadro FX card, which suffered a 38 FPS drop. At 1680x1050, three of our six test systems fall below the 30 FPS minimum. | ||
| Ligne 107: | Ligne 132: | ||
| Why go through all the trouble for what amounts to mostly low-end older systems? Mostly because they’re still out there being repurposed. As a refurbisher, | Why go through all the trouble for what amounts to mostly low-end older systems? Mostly because they’re still out there being repurposed. As a refurbisher, | ||
| + | |||
| + | En 1680x1050 (page précédente, | ||
| + | |||
| + | Le Phenom d'AMD personnalisé passe sous le seuil des 30 FPS à 1680x1050. Une explication pour cette faible performance pourrait être le fait que tous les affichages graphiques sur les systèmes peu performants sont des puces intégrées sur des cartes-mère anciennes. Le Thinkcentre i5 a aussi un jeu de puces Inel 2500 HD intégré, mais, comme c'est un jeu de puces plus récent, il est mieux supporté et tout simplement meilleur. Bizarrement, | ||
| + | |||
| + | Pourquoi faire tout ce travail pour de telles valeurs sur d' | ||
| **On eBay, people are selling similar HP 6000 Core 2 Duo based systems for between £45 - £60 including shipping. The decent performing Precision T3500 sells for around £140 when you include shipping. Local brick and mortar refurbishers sell systems like the HP 6000 for around $99CDN (around £58) as spec’d in this article (with a Windows 10 OEM refurbisher license). At the newest end of the spectrum, we found an ad on our local Kijiji for a Lenovo M82 with 12GB of RAM that was selling for $260 (£160). Sometimes, it takes a bit to find really good deals on used hardware, but, around here, there are always ads (by both brick and mortar stores and individuals) for graphics cards and systems. Of course there’s always “buyer beware” when dealing with individuals and dealers that haven’t been around long. ** | **On eBay, people are selling similar HP 6000 Core 2 Duo based systems for between £45 - £60 including shipping. The decent performing Precision T3500 sells for around £140 when you include shipping. Local brick and mortar refurbishers sell systems like the HP 6000 for around $99CDN (around £58) as spec’d in this article (with a Windows 10 OEM refurbisher license). At the newest end of the spectrum, we found an ad on our local Kijiji for a Lenovo M82 with 12GB of RAM that was selling for $260 (£160). Sometimes, it takes a bit to find really good deals on used hardware, but, around here, there are always ads (by both brick and mortar stores and individuals) for graphics cards and systems. Of course there’s always “buyer beware” when dealing with individuals and dealers that haven’t been around long. ** | ||
| + | |||
| + | Sur eBay, les gens vendent des systèmes HP 6000 à base de Core 2 Duo similaires pour 45 à 60 £, livraison incluse. Le Precision T3500 aux performances honorables est vendu à environ 140 £ si la livraison est incluse. Les rénovateurs locaux en boutique vendent des systèmes comme le HP 6000 pour environ 990 $ CDN (env. 58 £) comme spécifié dans cet article (avec une licence Windows 10 OEM de rénovateur). Dans la partie la plus récente du spectre, nous avons trouvé une annonce sur notre Kijiji local pour un Lenovo M82 avec 12 Go de RAM qui était vendu pour 260 $ (160 £). Quelquefois, | ||
| + | |||
| ======5====== | ======5====== | ||
| Ligne 117: | Ligne 151: | ||
| Although the Precision T3500 scored better in gaming performance than the i5, we’d probably still pick it over the T3500 for 3D gaming simply because we could add a newer PCIe video card to the system to outpace the T3500. Buying a slightly newer card can help with gaming performance on an old system like the Athlon 64 X2 4400+, but be aware that the card might be capped by a slower (PCIe 1) bus speed and may not do nearly as well as it would in a newer system. Generally the higher the gaming resolution, the fewer frames per second, so if you want to game higher than 1680x1050, be sure to get the best video card you can buy.** | Although the Precision T3500 scored better in gaming performance than the i5, we’d probably still pick it over the T3500 for 3D gaming simply because we could add a newer PCIe video card to the system to outpace the T3500. Buying a slightly newer card can help with gaming performance on an old system like the Athlon 64 X2 4400+, but be aware that the card might be capped by a slower (PCIe 1) bus speed and may not do nearly as well as it would in a newer system. Generally the higher the gaming resolution, the fewer frames per second, so if you want to game higher than 1680x1050, be sure to get the best video card you can buy.** | ||
| + | |||
| + | En résumé | ||
| + | |||
| + | Si vous construisez une machine pour l' | ||
| + | |||
| + | Bien que le Precision T3500 ait un meilleur score en performance de jeu que le i5, nous le choisirions probablement plutôt que le 3500 pour des jeux en 3D parce que nous pourrions ajouter une nouvelle carte vidéo PCIe au système pour surpasser le T3500. L' | ||
| **Systems like the Pentium D-based Optiplex 745 and Athlon 64 X2 4400+ are really near the end of their life, but two years ago, a Pentium D was all we had in our KODI server and it was okay paired with a GeForce 210 1GB video card. The Pentium D wasn’t as responsive as the Core 2 Quad chip we switched it with, but we were able to play our locally stored HD content. As a refurbisher, | **Systems like the Pentium D-based Optiplex 745 and Athlon 64 X2 4400+ are really near the end of their life, but two years ago, a Pentium D was all we had in our KODI server and it was okay paired with a GeForce 210 1GB video card. The Pentium D wasn’t as responsive as the Core 2 Quad chip we switched it with, but we were able to play our locally stored HD content. As a refurbisher, | ||
| + | |||
| + | Les systèmes comme l' | ||
| + | |||
| + | |||
| + | |||
| + | |||
issue120/labo_linux1.1494829091.txt.gz · Dernière modification : 2017/05/15 08:18 de d52fr
