About the Nano..... WCCFtech shared this:
First, "~900MHz for gaming", but assuming that the card is running cool to begin with, for a 30-second benchmark run, before temps rise to 75C when it starts to throttle... (I'm guessing that it throttles all the way down to maybe 600MHz or so when hitting 85C).
Then "Not enough cooling?" FAR FROM TRUTH?!??!??? Perhaps only for the short benchmarks, or 3DMark where there's a cool-down period in between the short tests...
900MHz will be the absolute max that 175W PCIE supply could ever dream of powering, no matter what the temperature is - unless it's a super simplistic non-demanding game that can run at 1000MHz without using much power.
AMD's PR BS at its finest!
"~75C at typical loads" stated by the PR, usually translates to maybe "~70C at modern popular game loads", or less.
For that tiny fan to keep the HUGE 599mm^2 chip at less than 70C so that power leakage caused by heat doesn't waste too many watts, would probably mean no more than 700-800MHz for sustained gaming of more than say, 15-20 minutes. It has to be that much slower than the WATERCOOLED Fury X, for anything to make sense.
Quote:Quote:Q1- Will the R9 Nano be offered as a reference model only or will AIBs be allowed to offer custom models?
Yes, 3+ months later.
Q2- Can you talk a little about overclocking on the Radeon R9 Nano?
Sort of, because of board power, it may be limited. Targeted at power efficiency. Can overclock, increase power limit, increase clock, but limited by 175W total board power. Wide engine ran at low speed for more efficiency. Up to 1000MHz means it optimizes the frequency based on needs and applications. ~900MHz for gaming. FurMark would likely be lower. 500MHz Memory, can’t modify it officially, won’t be officially supported to be overclocked either.
Q3- In your hot chips 2015 presentation, you showed the Fiji chip has 4 ACE Engines compared to 8 as shown in the previous slides, is it actually 4 ACEs on Fiji?
Q4- Price?
$649
Throttled at 85C. ~75 C at typical loads. 175W board power.
Cooler looks less than it actually is. More efficient and advanced than it looks. Fans mounted into shroud No 3D CAD drawing for Nano. Can’t remove shroud without removing fan, though fan mounts can be made!
Not enough cooling? FAR FROM TRUTH
Dedicated heatpipe for VRM’s. Industry first for OEM. Fins are horizontal. Helps exhaust air out the back of the card and maintain positive air pressure even with an axial solution. Hybrid vapor chamber/heatpipes, Flattened heatpipes on the whole underside of the heatsink. Vapor chamber over ASIC itself.
September 7th availability
Read more: http://wccftech.com/amd-radeon-r9-nano-unleashed-649-fastest-miniitx-card-full-fiji-gpu-4-gb-hbm-performance-faster-gtx-980/#ixzz3kPmEGcim
First, "~900MHz for gaming", but assuming that the card is running cool to begin with, for a 30-second benchmark run, before temps rise to 75C when it starts to throttle... (I'm guessing that it throttles all the way down to maybe 600MHz or so when hitting 85C).
Then "Not enough cooling?" FAR FROM TRUTH?!??!??? Perhaps only for the short benchmarks, or 3DMark where there's a cool-down period in between the short tests...
900MHz will be the absolute max that 175W PCIE supply could ever dream of powering, no matter what the temperature is - unless it's a super simplistic non-demanding game that can run at 1000MHz without using much power.
AMD's PR BS at its finest!
"~75C at typical loads" stated by the PR, usually translates to maybe "~70C at modern popular game loads", or less.
For that tiny fan to keep the HUGE 599mm^2 chip at less than 70C so that power leakage caused by heat doesn't waste too many watts, would probably mean no more than 700-800MHz for sustained gaming of more than say, 15-20 minutes. It has to be that much slower than the WATERCOOLED Fury X, for anything to make sense.
Ok with science that the big bang theory requires that fundamental scientific laws do not exist for the first few minutes, but not ok for the creator to defy these laws... 