Intel Xe discrete GPUs will sell from $199 and upwards

[HEXUS]

And some sites have shared preliminary benchmarks of Intel Ice Lake systems.

Read more.

[DanceswithUnix]

AMD to strike back with new mobile APUs sooner rather than later - depending on the pricing of upcoming Ice Lake portables.

For a budget laptop with an ageing mid range 2500U APU which is sold (and therefore I expect benchmarked here) with single channel ram I think the 2500U did rather well.

[LSG501]

For a budget laptop with an ageing mid range 2500U APU which is sold (and therefore I expect benchmarked here) with single channel ram I think the 2500U did rather well.

Agreed, even more so when you consider the ryzen 5 2500U is 'mid range' and the intel i7-1065G7 is intel's top of the line 15/25W cpu according to yesterdays article and will likely be priced to go up against the ryzen 7 range.... which funnily enough just happens to be missing from the charts...

[Tabbykatze]

For a budget laptop with an ageing mid range 2500U APU which is sold (and therefore I expect benchmarked here) with single channel ram I think the 2500U did rather well.

Agreed, even more so when you consider the ryzen 5 2500U is 'mid range' and the intel i7-1065G7 is intel's top of the line 15/25W cpu according to yesterdays article and will likely be priced to go up against the ryzen 7 range.... which funnily enough just happens to be missing from the charts...

Intel putting themselves on the back foot? Perish the thought that could ever happen.

[LeSWiS]

To build a strong user base from scratch at that price point, Intel gpus will need to beat up the competetion with no debate in the reviews and with amazing drivers support. Intel has a lot of money but it seems like a gargantuan objective from my standing pov. And for Koduri to bring his work from AMD to Intel(by that I mean he said that 'he likens AMD's strategy' that he probably created himself before leaving lol) also seems kind of odd. Wait & see as they say but all of this feels kind of strange with nothing to really grasp from at the moment.

[=assassin=]

It is going to take a long time to build up confidence in their products when the established main players have been doing discrete graphics cards for years. Even if performance was quite competitve at whatever price point they chose for each card, $199 or above is quite alot to take a gamble on when factoring in drivers and future support. If the project is halted in a couple of years, then I doubt they'd still be pushing regular driver updates. Most people will likely just stick with what they know, unless Intel seriously undercut their rivals, which is hard to imagine Intel doing. Perhaps they'll get some Intel fanbois to buy, and people who are after a potential novelty in the future, but initially it will be hard to gain momentum until they've had at least a couple of years under their belt.

[Iota]

amazing drivers support.

factoring in drivers and future support.

I think this would be my own personal issue, driver support has to be there in a similar vein to AMD and Nvidia, which means frequent updates to fix bugs and to provide driver level support for new games and software. If they're serious then they need to get into that more frequently with the Intel on-die graphics to prove they're going to deliver for the dGPU market.

[anselhelm]

Despite the disadvantage of having Raja at the helm, I'm hoping that Intel is able to break into the discrete graphics card market. Why? More competition.

Right now, we essentially have the following situation:

• Nvidia sets the prices and raises these above inflation each year.
• AMD matches those or slightly undercuts them slightly.
• Consumers have a binary choice between the two.

If Intel could actually do a half-decent job at an affordable, mass-market price range (~$200 USD), then they'll be on to something. However, they'll need:

• Provide excellent drivers out of the door.
• To build strong links with game studios, Microsoft, Apple and the Linux community.
• Provide continuing after-market support in terms of frequent driver updates, fixing bugs and adding features.
• To keep prices competitive, relative to the graphics power of the cards.

[preter_s]

Koduri is over-rated. He only knows the high-volume (translation: low price low performance) market that is all he knows. At the end of the day, he is no genius like Jim Keller. He doesn't have anything really to bring to the table.

This was the man who single handedly crippled AMD's top end GPU offerings for what, 2, 3 years by causing them to have nothing to show through the constantly delayed Vega. How? Through his genius of architecting a GPU using HBM memory when it was so bleeding obvious that the industry supply chain couldn't supply much of it, much less at an affordable price. It goes back to the same point - he had nothing original or creative or brilliant to offer except to use the fastest, rarest, most expensive form of memory to compete with Nvidia and to try to close the gap and make him look half-decent.

The fact that Navi is back to using GDDR6 RAM says it all. Lisa Su and the board certainly wasn't happy with him.

So glad AMD is rid of him! He can ruin Intel from within!

[Tabbykatze]

Have to agree with the comments on Koduri.

But could it be he was trying to do too much with too little budget or was he dreaming too big for all the wrong reasons?

We will see what happens when Xe is released.

[DanceswithUnix]

Despite the disadvantage of having Raja at the helm, I'm hoping that Intel is able to break into the discrete graphics card market. Why? More competition.

Sadly I suspect that on the desktop Intel will be competing with AMD at the low end rather than with Nvidia at the high end, doing the consumer no favours but hurting AMD's ability to make new products just at the point things were looking up for them.

[Xlucine]

Sadly I suspect that on the desktop Intel will be competing with AMD at the low end rather than with Nvidia at the high end, doing the consumer no favours but hurting AMD's ability to make new products just at the point things were looking up for them.

I wouldn't put it past intel to burn money selling below cost at the low end just to spite AMD for beating them in the CPU market

[Tabbykatze]

I wouldn't put it past intel to burn money selling below cost at the low end just to spite AMD for beating them in the CPU market

I would put money on them doing that if i were a betting man.

However, i wonder if AMD will have to fully integrate their RTG side of things because the moneys made and spent are solely within RTG. But don't Intel take all of their segments as one big pot so if one one segment is doing well, it can supplement the other?

With AMD, the wins for the CPU side won't be used to bolster the GPU side?

At least that's how I understand it

[DanceswithUnix]

With AMD, the wins for the CPU side won't be used to bolster the GPU side?

I can't imagine why that would be the case. In the past AMD haven't really had any money from either side, so there wasn't an issue.

[excalibur1814]

"Overall the results seem to be quite impressive"

If it's not hitting a 'minimum' of 30fps, then why would anyone buy it for any type of gaming? It's just another load of resources thrown at something that's not needed. Maybe this is a move to completely remove the graphics PU from all of their cpus?

[DanceswithUnix]

Maybe this is a move to completely remove the graphics PU from all of their cpus?

There is now rumours of hybrid crossfire style operation, which would imply that for the normal consumer stuff at least the IGPU stays for a while yet.