Save Zero Dollars By Opting for Intel's iGPU-Crippled GPUs

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Still rocking a 2550K (speed stepping up to 4.5 GHz) on a P67 board. Purchased the CPU for $150 at Micro-center 7 years ago. The MB was cheaper since it does not provide a video port. Perhaps a MB maker could release a cheaper model lacking video ports?
 
Even if AMD mobos were more money, support for AM4 through 2020 means you would amortize that cost more effectively than with Intel products, which seem to come with a new socket just about every new generation of CPUs.
 
Question is why should any supplier purchase an F model if there is absolutely no benefit over the igpu enabled counterpart? Because of stock levels possibly?
 
AMD Phenom chips were sold with disabled cores, but it was often just as likely that they were actually still functional and you could enable them in your BIOS - I got an extra core out of a Phenom II I bought this way.
 


That is only if you do a straight CPU upgrade with no MB change out at the same time. Frankly it works out really well if you bought into the first generation Ryzens especially if you went with a bargain CPU as Zen2 later this year is going to be a really nice upgrade and we may even get one slot with PCIE 4.0 out of the deal. With that said very seldom do people upgrade their CPU's more often than every 4 years so it's somewhat a wash.
 
Possibly marketing allocations as well, such that retailers have to buy so many of more mundane or less desirable variants to get the top-of-the-line models at meaningful levels.
 
The last AMD CPU I bought was an AMD Athlon FX-57 single core!! I was die hard AMD and then intel started just laying into them as far as what I was looking for in a desktop CPU.

I am sorry Intel, you've dropped the ball too much recently, and AMD is bringing back good memories. I've decided, my next desktop will be AMD if I can help it
 


Or in other words.. these are chips with Failed iGPU's that have got recycled and rebadged. AMD use to do that back in the Phenom days but they allows use to do core unlocking and play RNG on a CPU gamble.

 

Unless you upgrade CPUs more often than every four years, AMD sockets being supported for four years makes absolutely no difference since you'll still need a new motherboard by the time you upgrade the CPU. And even if you upgrade CPU before the next socket change, you may still want a new motherboard to fully support updates like PCIe 4.0 and higher power VRM for extra cores and higher clocks anyway. I wouldn't pass on PCIe 4.0 between Ryzen 3 and 500-series chipsets to eliminate possible chipset IO bottlenecks for the foreseeable future.

As far as I am concerned, AMD's four-year sockets may be closer to being a con than a pro due to all the caveats and compromises that creep up on them over time. I don't mind new sockets as long as there are substantial material changes to justify them.

Intel's problem is that it changes sockets for no apparent reason as the CPU's pinout and socket-related features have been fundamentally unchanged since Skylake. (A few pins have been relocated for questionable reasons along the way and that's about it.)
 
Where does the video encode/decode block , you know the builtin vp8/mpg/h265 decoder .. is that still active or is it fused off too?
Cos that may have a big effect on certain software, and falling back to decode in software could be untenable
 

Intel's QuickSync leverages some IGP resources to do its job. No IGP, no QuickSync.
 
These chips are targeted at OEM's. There is significant demand for these chips in the OEM market for multiple reasons. Intel doesn't care about the one CPU you claim you aren't going to buy from them, even though you never would have bought one from them anyway. They're catering to the customers that buys 1000's of their CPU's. OEM's aren't going to care that these chips cost the same. They require a dGPU purchase, so the OEM will be able to make more money per system sale despite the CPU costing the same.
 


You're right, I probably should have qualified that. I am only into HEDT systems (I have 4). The bottom end like the B450 boards, may well be the same. But also consider, I am in Australia, the land of price-gouging and exploitation. Nothing is cheap here.

Even my kid's PCs are at the higher end using only Z370 motherboards, which are averaging $300 each here (that's another 6PCs) - About $200 - $205 USD

Depends on what features you need. For me the ONLY option across all 5 of the main mobo-makers is the X399 Meg Creation. That's $769 AUD.

My Intel X299 Xpower Gaming AC motherboard which has the same features cost me $404 AUD.

Although, I do believe I saw an Intel board for around $1099 recently... might have been an ASUS (which are always feature-poor boards, often with dodgy chipsets).
 
This really just makes me want to buy an AMD Ryzen cpu, instead of shopping around for an Intel CPU, as it seems that a normal business option would be to lower the price of their F-series (defective) CPU's in order to move product; but instead we see Intel is being just as ludicrous as ever with their pricing structures!
 

To be fair, Intel is perpetually out of stock of most SKUs at the moment, which means the market is willing to bear the Intel toll at the moment to get its hands on whatever chips Intel is able to churn out. Since publicly traded companies have to deliver value to shareholders, Intel has every reason to nudge prices up (or not nudge them down on cut-down product variants) at the moment.
 

Intel isn't clueless, it is fulfilling its obligations to shareholders: deliver value by pricing its products up to whatever the market will bear. If Intel has trouble stocking stores' shelves, it means prices still aren't high enough.
 


So? Intel prices a 4 core 4 thread cpu for $130, ryzen offers a cpu that keeps up if not outperforms it for $90, which has overclocking enabled, a great stock cooler, and a far superior igpu. What they need to do is get their head out of their- heatpipe... and do what AMD is doing.
 
I think they're still also able to ride out the previous reputation of having the better gaming performance a bit, and so probably can still get away with charging more.

Sort of like what Sony used to be able to do with home electronics (TVs, etc.) - just riding on their name for a while to justify higher pricing despite the fact that other companies had caught up with them.


Whether my assumption is actually the case or not, I think I'll have to agree with InvalidError - if it delivers the profits to Intel, then it works. If it ever starts adversely affecting their bottom line, then it'll be stupidity to continue the same tactics.
 
Well ... Intel did tell their investors that throughout their 14nm "shortage" they would be able to meet their "revenue targets" (Intel never said they'd meet demand despite the articles).

How to meet revenue targets in face of a shortage? 1) Price increases (done) 2) Sell defective chips that would otherwise have been recycled for the same price as their non-defective chips. (done)

It appears yet another example of Intel throwing its own consumers under the bus for the sake of impressing investors ...

I hope these at least have a little more thermal headroom and thus OC ability, that may at least justify the pricing a little bit.
 

Customers only matter to the extent that they continue buying the products about as fast as they can be manufactured instead of switching to a competitor's, leaving inventory stuck on shelves and warehouses. Shareholders are customer #1 in most other regards.

If you want Intel to stop being excessively greedy, you need a sufficiently large chunk of the market to quit buying Intel to cause Intel to have excess inventory and keep it that way until Intel is forced to drop prices to get that excess inventory moving again.