-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD VEGA confirmed for 2017 H1
- Thread starter Doom112
- Start date
- Status
beats me!
the Q1 results were actualy very good, AMD was dead on estimate, and they had 18% revenue over last year's Q1.
ppl will buy again after this drop, and sell again in Q2, short term profit is still possible for AMD stock, and it will stabilize in Q3/Q4, when revenue will show large profitability with their new products.
AMD made 70 million loss in operating expenses and 300 million cash went in securities according to lisa ,however, investors know where it went.
$29 million operating loss... $73 million loss in net total. Nothing new because AMD always reported losses. But it is 18% increase year on year revenue growth so they are going in the right direction to improving things. Graphics segment is up 29% year on year and thats just with polaris and not even touching the high end where i was told it was more profitable than mid range lol. Must be seen as AMD not focusing on this made them increase their revenue. Now with Vega on the horizon things will only improve in the graphics segment. Ryzen will only get better so things are looking up for AMD. Love how people like to put negative spins on things.
Hype is the reason AMD share tanked today. If some people do not learn there lesson than you see a worse stock drop than this when AMD releases VEGA.
Soldato
- Joined
- 25 Nov 2011
- Posts
- 20,675
- Location
- The KOP
I'm starting to think those recent Vega 'leaked benchmarks' might be accurate. Though I'd hope it's only for the 'smaller Vega' card. Cos Vega not even beating a 1080 is just poor show after 1 year following the 1000 series launch. One has to wonder what they'll price such a GPU that beats a 1070 though? I'd prefer it to be no more than £400, but knowing AMD, they'll 'slot it around Nvidia' and charge more than what the 1070 costs. I don't think Vega is going to do a Ryzen.
On the whole G-sync front, if DP is right about the cost being so low, then colour me disgusted at people willing to pay so much for something that does exactly the same sort of thing that FreeSync does. But I don't quite think human beings are so hopeless, so I'll believe that the fees, chip-cost and development costs are the reason why G-sync is almost double what a FreeSync monitor costs.
The sad thing is, G-sync is not a bad technology, it's pretty good for what it does. And yet I can't stand it just due to the disgusting pricing of G-sync monitors alone. I'd rather own a FreeSync monitor, even if I have an Nvidia GPU. No way is G-sync worth almost doubling the price of the panel.
On the whole G-sync front, if DP is right about the cost being so low, then colour me disgusted at people willing to pay so much for something that does exactly the same sort of thing that FreeSync does. But I don't quite think human beings are so hopeless, so I'll believe that the fees, chip-cost and development costs are the reason why G-sync is almost double what a FreeSync monitor costs.
The sad thing is, G-sync is not a bad technology, it's pretty good for what it does. And yet I can't stand it just due to the disgusting pricing of G-sync monitors alone. I'd rather own a FreeSync monitor, even if I have an Nvidia GPU. No way is G-sync worth almost doubling the price of the panel.
well i have been looking for this for quite some time, luckily someone posted it on reddit.
cost of 290X GDDR5 vs FuryX HBM1, the cost is 33% more for the memory, but if you facture in the interposer it brings it to about double the price, extra cost on packaging is offset by the pcb cost.
these prices are from way back, my guess HBM2 could be a bit cheaper while 5X is more expensive, best case i see 8GB of HBM2 = 12GB of GDDR5X, worst case = 16GB of GDDR5X
50$ per stack of 4GB of HBM2 isn't that bad, also funny how a 200$ product going through the chain ends up at 600$
, and AMD probably only gets about 100-150$ profit out of it
cost of 290X GDDR5 vs FuryX HBM1, the cost is 33% more for the memory, but if you facture in the interposer it brings it to about double the price, extra cost on packaging is offset by the pcb cost.
these prices are from way back, my guess HBM2 could be a bit cheaper while 5X is more expensive, best case i see 8GB of HBM2 = 12GB of GDDR5X, worst case = 16GB of GDDR5X
50$ per stack of 4GB of HBM2 isn't that bad, also funny how a 200$ product going through the chain ends up at 600$
, and AMD probably only gets about 100-150$ profit out of it
Last edited:
some news from reddit, apprently a new AMD GPU passed a certification in south korean facility.
vega is really coming ?, because im not sure anymore, started to feel like vaporware lately.
It will be here in Q2. If not, then it would be fair to call it vaporware
Nope. It used to be AMD then they branched the GPU section off as Radeon technologies group or RTG for short.
Associate
- Joined
- 30 Apr 2009
- Posts
- 111
well i have been looking for this for quite some time, luckily someone posted it on reddit.
cost of 290X GDDR5 vs FuryX HBM1, the cost is 33% more for the memory, but if you facture in the interposer it brings it to about double the price, extra cost on packaging is offset by the pcb cost.
these prices are from way back, my guess HBM2 could be a bit cheaper while 5X is more expensive, best case i see 8GB of HBM2 = 12GB of GDDR5X, worst case = 16GB of GDDR5X
50$ per stack of 4GB of HBM2 isn't that bad, also funny how a 200$ product going through the chain ends up at 600$
Not sure if they're quoting prices on release or not (cost vs difference in die size) sure it's not .
Doesn't make sense
Because the stock market is bonkers. AMD met their predicted targets, however they didn't meet the targets predicted by stock market analysts therefore their stocks dropped. Then people panic sold, and it dropped more. Just the stock market doing it's typical nonsense.
Big players sold, that drop triggers automated stop losses from day traders and people that aren't long players.
It cascades down.
AMD also increase EPS directly inline with what Analysts estimated, while Intel dropped.
AMD's increase in shaved off 2/3 of their previous losses on EPS.
Ryzen R7 was on sale for 5 weeks before Q1 ended, that's not much at all; although even then Passmark recorded AMD increased their Market Share by 2% in that time for Processors.
Goldman Sachs is also the ones that had a price target lowered to 10.60.
While other's have kept the same Price Target, or even upped it.
Well Fargo: Increased to $13-15
Merrill Lynch: Increased to $16.50
Barron's: Reiterates Buy
Matt Ramsey: Reiterates Buy $17.00
The last time Goldman advocated for a sell and downgraded the stock, they ended up buying a **** ton of it after the smoke settled down; This was also after they had to cover their losses when NVIDIA dropped from $120 > $99
As for Ryzen R5, R3, Vega, and Naples; we won't be seeing how well those perform until Q3-Q4.
Ryzen R5 has barely been out, not included in Q1 earnings; we might see that and R7 for Q2 earnings call.
Vega and R3 launching end of Q2, and availability might not be decent until Q3, so once again need to look to Q4 before we can see how well it actually performed for them
the value was built on hype mostly, now it adjusted, and will climb back up.
it's looking good for naple, seeing how Ryzen performs.
the unknown is Vega, although i dont think it would shake up their stock that much even it's not good.
personally i think their biggest product will be raven ridge, but thats for 1H 2018
well i have been looking for this for quite some time, luckily someone posted it on reddit.
cost of 290X GDDR5 vs FuryX HBM1, the cost is 33% more for the memory, but if you facture in the interposer it brings it to about double the price, extra cost on packaging is offset by the pcb cost.
these prices are from way back, my guess HBM2 could be a bit cheaper while 5X is more expensive, best case i see 8GB of HBM2 = 12GB of GDDR5X, worst case = 16GB of GDDR5X
50$ per stack of 4GB of HBM2 isn't that bad, also funny how a 200$ product going through the chain ends up at 600$
Those costs are off a bit (even taking the year into account) - as I mentioned before I can order GDDR5 at $2 for one of those modules - which would come to $32 however if I bought in quantities of 1000 with minimum order quantity of 5000 - which I assume a big company would do something like that it comes down to $0.50 per module (+/- a few cents).
The HBM would cost me a minimum of $56 for that setup but they don't really sell it like that - AFAIK unless you are a big company you still have go through a design company who builds a product on an interposer for you and with HBM2 probably more like $60 per stack (no idea what it would cost AMD).
Last edited:
Soldato
- Joined
- 28 Sep 2014
- Posts
- 3,574
- Location
- Scotland
also funny how a 200$ product going through the chain ends up at 600$
Actually Fury X launch price was $649 but AMD didn't get $100-$150 profit out of it. Accorded to both Gibbo and also Lisa Su back to AMD Financial Analyst Day 2015 said both CPU and GPU's profit margin is in mid single digit that around 5% or 6%.
Using margin calculator below showed AMD made a $9.89 profit at 5% margin or $12 profit at 6% margin on Fury X.
https://www.omnicalculator.com/business/margin
I never understood how AMD business model operated on very low profit margins. Still shocked at poor AMD Q1 2017 result showed revenue still flat with very low profit margins after sold 1 million Ryzen 7 CPUs made AMD a peanut profit of about $10 to 12 million and still lose lots of money posted $73 million loss. Pathetic.
Absolutely make no sense.
- Status