As most people are aware, AMD dropped the first GPUs to utilize HBM (High Bandwidth Memory). These GPUs use a form of HBM called 2.5D which requires the use of an interproser layer than both the memory and the GPU sit on. This is opposed to the 3D stack in which the memory sits on top of the processor that owns it. The traditional stacking of 3D Memory provides significant performance benefits, but would require a different chip for every memory density you plan on releasing. In the GPU world this can be a big problem and is why both AMD and NVIDIA have opted for the 2.5D method.
Some GPU news comes today from the AMD camp where we are hearing rumors that AMD’s Radeon 390X might ship with 8GB of High Bandwidth Memory (HBM). Now this sounds great when you take into account the fact that HBM has some serious muscle behind it (128GBps per memory chip). AMD could technically leapfrog nVidia’s new monster the GeForce Titian X with 12GB of memory onboard. The problem is that some of the information that is floating around does not seem to add up.
CES 2015, Las Vegas NV -
ADATA has had some very interesting products to show us at CES in the past and this year was no exception. Although much of what we saw was not new, it was still pretty interesting. We kicked off our guided tour with a couple of demonstrations. The first was one showing off the power loss protection on their SSDs. This is a handy little feature that helps prevent you from losing data that has not been written to “disk” in the event of a loss of power. Not really new, but still cool.
CES 2015 Las Vegas, NV Patriot Suite Bellagio
Stopping by the Patriot suite at the Bellagio we found Patriot talking about some pretty cool new items in addition to their normal memory products. We kicked off our meeting with a walk through of their flash memory products.
CES 2015 Las Vegas, NV Caesar’s Palace
One company that we always are sure to visit during CES is Kingston. Over the years they have had a number of interesting demos and products that are dropped on the world during CES. This year was no exception. David Leong was quick to get into the meat of the presentation and showed off the next generation of the Cloud gaming headset. This new headset features 53mm drivers and a power USB audio adapter to push 7.1 virtual surround sound.
The Flash Memory Summit wound down on Thursday after a four day run at the Santa Clara Convention Center in San Jose. The show floor was fairly crowded with over 5,000 attendees and a sold out exhibition space.
One of the items we have always beat AMD up on is there poor memory performance in their CPUs and APUs. This little issue is what has separated AMD from Intel since the AM2 days. It has always been understood that latency has a massive impact on an internal memory controller. As you latency increases your efficiency decreases. You can offset some of this by enlarging your cache and also optimizing the CPU to use it more efficiently. This is one area that AMD has traditionally had issues with, even going back to the Athlon 64 we saw them reducing cache sizes to remove problems and bump performance.
Samsung has started mass production of 4 Gb DDR3 memory chips in 20-nanometer manufacturing process. According to the manufacturer, production of DRAM chips in fewer nanometer technology is more demanding job in relation to the NAND flash memory chips. While NAND in every cell requires only a transistor, 0DRAM memory must have both transistor and capacitor.
According to sources coming from the company itself, Intel will during the third quarter of this year add support for DDR memory 4 of the new generation. New DDR 4 memory is developed for a long time, specifically over five years.
During yesterday an agreement was reached between Toshiba and CAA, by which Toshiba buys OCZ's technology related to solid state disk drives.