DigiTimes have heard that NVIDIA hope to release the GTX 1050 at the end of the month or early in October but there are hints it may be somewhat of a paper launch. NVIDIA will have some silicon for sale but only a few who are quick enough on the draw will be able to purchase them; similar to the launch of the GTX 1080 and 1070 or even AMD's new cards for that matter. Both vendors had great difficulty providing retailers with cards to sell at launch, the supply has increased and prices are much closer to the MSRP than they were a month ago but there are still plenty of out of stock models when you start looking for deals.
Hopefully this situation will resolve soon, but for now patience is recommended for upgraders on a tight budget.
"Nvidia's Pascal-based GeForce GTX 1080, 1070 and 1060 graphics cards are seeing strong demand from the market and therefore are currently in tight supply. To further expand into the market, Nvidia is planning to release its mid-range GTX 1050 graphics card at the end of September at the earliest."
Here is some more Tech News from around the web:
- Humble Summer Sale Kicks Off With Free Tropico 4 @ Rock, Paper, SHOTGUN
- Apple markets the iPhone 7 as a penis in Hong Kong @ The Inquirer
- Ten-year-old Windows Media Player hack is the new black, again @ The Register
Not surprising. I waited over
Not surprising. I waited over a month before the 1060 was actually in stock. What’s up with yields with this node?
its a new proc node its
its a new proc node its pretty common that yields are always iffy.
that’s normal when gpu
that’s normal when gpu transition to new node. i still think the current situation is not as bad as 40nm transition back in 2009/2010.
for, Nvidia, sure, they had
for, Nvidia, sure, they had serious problems at 40nm… Now it seems like it’s AMD with the shorter stick.
There is a lot more FP
There is a lot more FP compute on the RX 480 mainstream SKU than on the GTX 1060! And the RX 480 is on a smaller more densely packed process node so it’s not all the process node’s fault! TSMC’s 16nm node has been around longer so the clocks can go higher because at 16nm the circuits are not as densely packed relative to 14nm. Those DX12/Vulkan titles will make more use of the RX 480’s extra FP compute so that will generate more heat. Maybe GF will be able to with Samsung’s help get that licensed from Samsung 14nm process tweaked and more mature to get the power usage/heat down and allow for higher clocks, but once the DX12/Vulkan titles are in more use let’s wait for more benchmarking to see.
All that Async-compute on AMD’s hardware, in addition to the extra FP compute resources is bound to use a little more power than Nvidia’s SKUs in the same market segment, but the DX12/Vulkan benchmarks are favoring AMD’s GCN/Polaris and earlier GCN SKUs for more improvement over time than any of Nvidia’s SKUs that are regressing/stagnating in performance with the newer Vulkan/DX12 graphics APIs.
Nvidia is having to clock its Pascal mainstream GPUs higher to get the FLOPS higher relative to AMD’s Polaris micro-architecture that gives all that extra FP compute at much lower clock speeds. And with DX12/Vulkan expect that with the new graphics APIs the games will beginning to tap that extra FP compute at lower clocks to give AMD’s RX 480/other GCN SKUs some extra boost over time as the gaming software ecosystem begins to tap into DX12’s/Vulkan’s better usage of AMD’s async-compute ability.
I had to wait awhile after
I had to wait awhile after the release to get my hands on a GTX 1060, so this is no surprise.