AMDs Vega Graphics Cards Could Kick Off a War. Every nerd loves a good tech war Windows vs Mac, Apple vs Android, Intel vs AMD. They give us something to armchair argue about over beers with friendsor to rant over in the comments of illustrious tech blogs. After spending the weekend playing with AMDs new Vega 6. Vega 5. 6 graphics cards, I think I can safely say an old tech war is back oneven if AMDs latest salvo feels paltry. Nvidia might be leading the discrete graphics card industry, but AMDs two newest cards are cheap and fast enough to finally compete. And that can only mean good things for PC users. AMD Vega 6. 4 AMD Vega 5. Price4. 00 Vega 5. Vega 6. 4What is it Discrete GPUs that are competing against graphics titan Nvidia. No Like. Dont expect any flash. AMD, which purchased Nvidias previous competitor, ATI Graphics, has been losing the GPU war for a while. Nvidia is currently producing the majority of discrete graphics cards found in computers today. According to Jon Peddie Research, by the end of 2. The Internationale Funkausstellung Berlin, or more simply IFA, started out as a showcase for new TV and radio technology and has debuted things like the first. InformationWeek. com News, analysis and research for business technology professionals, plus peertopeer knowledge sharing. Engage with our community. Every nerd loves a good tech war Windows vs Mac, Apple vs Android, Intel vs AMD. They give us something to armchair argue about over beers with friendsor to rant. Nvidia had more than 7. AMD trailed far behind with just 2. So AMD decided to focus on building cheap cards to go in cheap machineslike the 5. I reviewed back in April. The AMD Vega 6. 4 and Vega 5. Nvidia 1. 08. 0 with cheaper options that start to approximate its performance. When the Vega microarchitecture the cards are based on was announced back at CES, people didnt immediately leap out of their seats. AMD didnt have a cool hook like when Nvidia announced it had spent billions to develop its latest card. All AMD had was a promise of speed when the cards arrived this summer. Eight months later, the Vega 6. The AMD Vega 6. 4 retails for 5. AMD Vega 5. 6 retails for 4. Both also have 8. GB of RAM built in. The big difference between the two is the number of compute unitsthink of those like the cores in a CPU the more the better. The Vega 6. 4 has 6. Vega 5. 6 has 5. 6. When I compared the Vegas to the 1. I found was far less exciting that what Id hoped for. While the AMD Vega 6. Instead it was sort of like going to the car lot and having to choose between a Honda Civic or a Toyota Corolla. Theyre both very nice, cost the same, and do the same damn thing. What was really astounding was the performance of the 4. AMD Vega 5. 6. Despite being considerably cheaper than both the Vega 6. Nvidia 1. 08. 0, it played Overwatch and Civilization VI only marginally slower. The 1. 08. 0 managed 1. Overwatch on Ultra at 4. K, the the Vega 6. Vega 5. 6 pulled of 9. K with the graphics cranked to Ultra. Thats not just a little respectable, thats really damn good. In Civilization VI the difference was even smaller, with the Vega 5. Vega 6. 4. In one case the Vega 5. Intel Iron Lake Mobile Graphics Chipset Update' title='Intel Iron Lake Mobile Graphics Chipset Update' />Nvidia 1. Vega 6. When I rendered a frame in Blender, graphics software that allows you to create large 3. D images that heavily tax a discrete GPU, the Nvidia 1. The Vega 6. 4 rendered the same frame in 9 minutes and 2. The Vega 5. 6 Just 9 minutes and 2. With that kind of neck and neck performance theres no reason to really buy either a Nvidia 1. Vega 6. 4 over the Vega 5. But speed isnt the only factor to consider when buying a discrete graphics card. See, the cards that go in your desktop PC are very power hungry. If your power supply cant provide enough juice, the GPU is worthless, and that is one place Nvidia performs far better than AMD every time. The Nvidia 1. 08. Intel Iron Lake Mobile Graphics Chipset Update' title='Intel Iron Lake Mobile Graphics Chipset Update' />The Vega 5. Vega 6. 4 requires a whopping 2. All the extra juice means you have to use not one, but two 8 pin power connectors from your power supply. The Nvidia 1. 08. With the new Vega cards AMD is trying to get around the power constraints of it GPUs by offering some software solutions. The first, and most notable, is the Radeon Chill feature, and its actually sort of clever. It operates under the assumption that people dont really need the fastest video card, they just need one fast enough for their monitor. Both AMD and Nvidia have a technology that allows cards to sync with monitors to deliver top level graphics without straining the GPU Nvidia calls its tech G Sync, while AMD calls it Free. Sync. But you have to have a special monitor that works with the syncing technology and no monitor works with both the AMD and Nvidia sync tech. Radeon Chill works with any monitor. You simply tell it how many frames you actually want to see per second. Got a monitor that refreshes 6. Set the max to 6. Radeon Chill makes magic happen. In addition, the software doesnt try and churn out 6. Instead it recognizes moments with static visuals and dramatically cuts down on how much power is being usedand you can tweak the number as well. Want it to never go below 3. Just set the slider in the AMD software. AMD added other software features that give users a better ability to throttle their cards without diminishing visuals. Theres a power saver mode which allows you to force the cards to sip as little as 1. Frame Rate Target Control mode that lets you cap the frames per second and an Enhanced Sync mode that figures out the optimal frame to display on screen, even if it means skipping a few other frames. All these features make the AMD Vega cards feel incredibly practical versus the excess and overclocking shenanigans of Nvidia. Its as if AMD is trying to say if you want nose bleed speed go with Nvidia, but if you want control and a nice experience join us. Thats a reasonable sale to an old woman such as myself. I play my games on 4. K TVs that have no syncing technology though Microsoft suggests that could change next year and can only show between 6. I dont need dual video cards or crazy numbers, I need good enough. And the AMD Vega 6. AMD Vega 5. 6 are good enough. If youre looking for a reasonably priced card the AMD Vega 5. But these cards arent enough to lure Nvidia loyalists or the power hungry away. As the latest salvo in the war between Nvidia and AMD goes, the new Vega cards are pretty weak. READMEThese cards are neck and neck with the 5. Nvidia 1. 08. 0 in the speed department. But they natively draw a lot more power. A bevy of software features try to resolve the power issue, but it means youre effectively throttling your shiny new GPU. SPEC DUMPAMD Vega 6. GB of RAM 6. 4 compute units 1. MHz base GPU clock 4. GBs memory bandwidth 2. TFLOPS 3 x Display. Port 1 x HDMIAMD Vega 5. GB of RAM 5. 6 compute units 1. MHz base GPU clock 4. GBs memory bandwidth 2. TFLOPS 3 x Display. Itanium Wikipedia. For more details on the technical architecture, not chip implementations, see IA 6. Itanium eye TAY nee m is a family of 6. User Manual Baker Universal 2000 Scale. Intelmicroprocessors that implement the Intel Itanium architecture formerly called IA 6. Intel markets the processors for enterprise servers and high performance computing systems. The Itanium architecture originated at Hewlett Packard HP, and was later jointly developed by HP and Intel. Itanium based systems have been produced by HP the HP Integrity Servers line and several other manufacturers. In 2. 00. 8, Itanium was the fourth most deployed microprocessor architecture for enterprise class systems, behind x. Power Architecture, and SPARC. In February 2. Intel released the current generation, Kittson, to test customers, and in May began shipping in volume. It is the last processor of the Itanium family. Historyedit. Itanium Server Sales forecast history56Development 1. In 1. 98. 9, HP determined that Reduced Instruction Set Computing RISC architectures were approaching a processing limit at one instruction per cycle. HP researchers investigated a new architecture, later named Explicitly Parallel Instruction Computing EPIC, that allows the processor to execute multiple instructions in each clock cycle. EPIC implements a form of very long instruction word VLIW architecture, in which a single instruction word contains multiple instructions. With EPIC, the compiler determines in advance which instructions can be executed at the same time, so the microprocessor simply executes the instructions and does not need elaborate mechanisms to determine which instructions to execute in parallel. The goal of this approach is twofold to enable deeper inspection of the code at compile time to identify additional opportunities for parallel execution, and to simplify processor design and reduce energy consumption by eliminating the need for runtime scheduling circuitry. HP believed that it was no longer cost effective for individual enterprise systems companies such as itself to develop proprietary microprocessors, so it partnered with Intel in 1. IA 6. 4 architecture, derived from EPIC. Intel was willing to undertake a very large development effort on IA 6. HP and Intel initiated a large joint development effort with a goal of delivering the first product, Merced, in 1. During development, Intel, HP, and industry analysts predicted that IA 6. RISC and complex instruction set computing CISC architectures for all general purpose applications. Compaq and Silicon Graphics decided to abandon further development of the Alpha and MIPS architectures respectively in favor of migrating to IA 6. Several groups ported operating systems for the architecture, including Microsoft Windows, Open. VMS, Linux, HP UX, Solaris,1. Tru. 64 UNIX,1. 0 and Monterey6. The latter three were canceled before reaching the market. By 1. 99. 7, it was apparent that the IA 6. Merced began slipping. Technical difficulties included the very high transistor counts needed to support the wide instruction words and the large caches. There were also structural problems within the project, as the two parts of the joint team used different methodologies and had slightly different priorities. Since Merced was the first EPIC processor, the development effort encountered more unanticipated problems than the team was accustomed to. In addition, the EPIC concept depends on compiler capabilities that had never been implemented before, so more research was needed. Intel announced the official name of the processor, Itanium, on October 4, 1. Within hours, the name Itanic had been coined on a Usenet newsgroup, a reference to the RMS Titanic, the unsinkable ocean liner that sank on her maiden voyage in 1. Itanic has since often been used by The Register,1. Itaniumand the early hype associated with itwould be followed by its relatively quick demise. Itanium Merced 2. Itanium MercedItanium processor. Produced. From June 2. June 2. 00. 2Common manufacturersMax. CPUclock rate. 73. MHz to 8. 00 MHz. FSB speeds. 26. 6 MTs. Instruction set. Itanium. Cores. 1Core namesL2 cache. KBL3 cache. 2 or 4 MBSocketsBy the time Itanium was released in June 2. RISC and CISC processors. Itanium competed at the low end primarily four CPU and smaller systems with servers based on x. IBMs POWER architecture and Sun Microsystemss SPARC architecture. Intel repositioned Itanium to focus on high end business and HPC computing, attempting to duplicate x. The success of this initial processor version was limited to replacing PA RISC in HP systems, Alpha in Compaq systems and MIPS in SGI systems, though IBM also delivered a supercomputer based on this processor. POWER and SPARC remained strong, while the 3. Only a few thousand systems using the original Merced Itanium processor were sold, due to relatively poor performance, high cost and limited software availability. Recognizing that the lack of software could be a serious problem for the future, Intel made thousands of these early systems available to independent software vendors ISVs to stimulate development. HP and Intel brought the next generation Itanium 2 processor to market a year later. Itanium 2 2. 00. Itanium 2 Mc. KinleyItanium 2 processor. Produced. From 2. Designed by. Intel. Common manufacturersMax. CPUclock rate. 90. MHz to 2. 5. 3 GHz. Instruction set. Itanium. Cores. 1, 2, 4 or 8. Core namesMc. Kinley. Madison. Hondo. Deerfield. Montecito. Montvale. Tukwila. Poulson. L2 cache. 25. 6 KB on Itanium. KB D 1 MBI or 5. KB I on Itanium. L3 cache. 1. 5 3. MBSockets. Intel Itanium 2 9. The Itanium 2 processor was released in 2. The first Itanium 2, code named Mc. Kinley, was jointly developed by HP and Intel. It relieved many of the performance problems of the original Itanium processor, which were mostly caused by an inefficient memory subsystem. Mc. Kinley contains 2. CMOS process with six layers of aluminium metallization. In 2. 00. 3, AMD released the Opteron CPU, which implements its own 6. AMD6. 4. Opteron gained rapid acceptance in the enterprise server space because it provided an easy upgrade from x. Under influence by Microsoft, Intel responded by implementing AMDs x. IA 6. 4 in its Xeon microprocessors in 2. Intel released a new Itanium 2 family member, codenamed Madison, in 2. Madison uses a 1. Itanium processors until Montecito was released in June 2. In March 2. 00. 5, Intel announced that it was working on a new Itanium processor, codenamed Tukwila, to be released in 2. Tukwila would have four processor cores and would replace the Itanium bus with a new Common System Interface, which would also be used by a new Xeon processor. Later that year, Intel revised Tukwilas delivery date to late 2. In November 2. 00. Itanium server manufacturers joined with Intel and a number of software vendors to form the Itanium Solutions Alliance to promote the architecture and accelerate software porting. The Alliance announced that its members would invest 1. Itanium solutions by the end of the decade. In 2. 00. 6, Intel delivered Montecito marketed as the Itanium 2 9. Intel released the Itanium 2 9. Montvale, in November 2. In May 2. 00. 9, the schedule for Tukwila, its follow on, was revised again, with release to OEMs planned for the first quarter of 2. Itanium 9. 30. 0 Tukwila 2. Intel Itanium 9. 30. CPU LGA. Intel Itanium 9.