Interview with Danny Lee from XGI

chavvdarrr

Veteran
http://www.ngohq.com/articles.php?go=read&arc_id=11
egeneration: How was cebit? are there are any particular thoughts and would you describe them?

Danny: CeBIT was great. We were able to meet with worldwide Media and let them preview our new PCIe products and change of XGI?s Marketing Strategy. The change of marketing strategy goes from targeting the high end product marketing to focusing on mainstream and value markets. Furthermore, there is a change in business where XGI is elaborating on its graphics card business in setting up worldwide channels for XGI Volari distribution. In Europe, we have Eagletek and Club3D. In Japan, ASK has a good success selling XGI Volari V3XT. In US, we signed up ASI and D&H as our main distributors for different market segments. We will be announcing a partnership with FIC soon.

Regeneration: What can you tell us about the upcoming products

Danny: The second half of 2005, XGI is pushing for PCIe?s Value and Mainstream Segment products. XG47 is our value segment PCIE with Shader Model 2.0 which targets at 6200 with Turbo Cache-like architecture; we call it eXtreme Cache. XG45 is our mainstream segment PCIE with Shader Model 3.0 which targets at 6600 with flexible dram configuration such as 96/192bit. Both products should have a more visible appearance in Computex 2005.

Regeneration: Maybe you can tell us when to expect the new XG45 and the XG47?

Danny: Like I mentioned in the previous question, in Computex 2005 you will see more visibility of the products. In April/May, XGI will have Media review samples of XG47 ready, and a month delay for XG45 in review samples. Mass production should be 1-2 months after the Media sample reviews.

Regeneration: Are there any particular plans for the Reactor Driver? And when we should expect a driver for the new Windows x64bit?

Danny: We have a total makeover in our new Reactor Driver. Furthermore, we have many new features in our new Reactor Driver. The ?what?s new? of the new Reactor Driver will be released before official product launch. In support of new Windows x64bit, our server Z7 graphic chip supports it. In our next generation PCIe products XG45 and XG47 both support it as well.

Regeneration: Can you give us more details about ?eXtreme Cache??

Danny:








Danny: XGI has the technology to bring the Graphics Memory down to 0. However, the performance won?t be competitive to our competitors. So we don?t recommend that. But our potential clients might be interested in this option because this means they have more playing field in the graphics industry.

Danny: By the same token, we offer flexible dram configuration for the mainstream product XG45 because the potential clients will have better leverage and better competition.





Regeneration: What about upcoming products for the Longhorn WGF?

Danny: WGF2.0 will be in our XG60 series which is coming in H2 of 2006. We are working closely with Microsoft in software compatibility with them.

Regeneration: Will there be a chance to see an AGP version of the new products?

Danny: We do have an AGP bridge designed in case PCIe doesn?t pick up. But we anticipate the chances of using AGP bridge is not high. Furthermore, our competitors are phasing out current AGP products where we are still in production. So XGI has a chance in filling the gap of AGP market if there is a delay in the PCIe market.

Regeneration: Are there any plans for a 512mb card?

Danny: Not for this coming generation products.

Regeneration: What are the plans for the XGI Corporation?

Danny: We plan to target the value and mainstream segment first and make our product available to global consumers. That is why we are pushing distribution worldwide of XGI Volari products. Next, once we consolidated this fundamental market where most of quantity is at, we will move onto the higher end markets where the top performance products are compared. Because we have the Duo Technology know-how, we will be able to implement that onto PCIe where a broader bandwidth is allowed and more suitable.

Danny: Overall, we are trying to catch up to ATi and nVidia in the coming PCIe products, and marching head to head with them in our 60series product that are available in H2 of 2006.

Regeneration: Thank you! and good luck.

96- 128- and 192-bit bus for VG45

PS i know there is topic for XGI at news forum, but looks like i can't post there, not even replies :p
 
Interesting change in strategic direction.

I have always been an advocate of their earlier business model: wow 'em at the high end, and push the brand and technology down. (Similar to the ATI and nvidia models).

However, the barrier to entry is so high (not just physical R&D, but the logistics required in having your architecture compatible with all the games on the market that have never even seen your chip), that at this time, it's basically impossible.

The window has passed.

That being said, I see a small window of opportunity with longhorn. If XGI (or anyone else) can nail...and I mean nail, WGF 2.0 compatibility for the O.S....and have it on the market in time, and be able to price it more aggressively than ATI and nVidia, there is an opportunity to grow your business.

I'm not even talking about game compatibility per-se. I mean be able to run Longhron in its "full glory", and cheaper than everyone else.

If XGI (or someone else) can do this, they can get lots of chips on the market...and then hopefully "force" devs to consider the architecture in new game development.

I can't stress how key it is to have a WGF 2.0 product ready and shipping for the OEM cycle that introduces longhorn though. They really need to nail that...and they should (hopefully) be working very closely with MS to make sure their architecture is "glitch free."

Goo luck, XGI.
 
Personally I never liked XGI attitude. We are going to beat them because we are going to use cheap labour.

I am a bit disturbed at UMC and TMSC getting involved in the graphics industry. UMC has a stake in XGI and TMSC has a stake in ageia.

If I were ATI and Nvidia I would look closely at other fab partners.
 
rwolf said:
If I were ATI and Nvidia I would look closely at other fab partners.
IBM doesn't have enough capacity. The others are insignifiant. And those are just standard risk investments, it's not like they owned more than 33% of those companies...

Uttar
 
Their Business model will *Never* Work.

Not in a million years. They have just doomed thmselves to abject failure. You cannot target the Low end and Value segment and sell Squat. Its just not going to happen.

How did ATi come back and take the lead over from Nvidia even if only by a small percentage?? They did it by Kicking the crap out of the High end and making every single writer, website, and magazine take notice. The buzz rolled in and ATi as a high performance Quality brand permiated the market. The "Radeon" name has become synonumous with Quality and Performance. When someone goes into a store and wants a card the guy says "get this 9200 or 9600 they are ATi's and they ROCK dude!!".

Thats how it happens.

If you release some card that no one knows about or associates the name "Volari" with a laughable overblown, overhyped underperforming clucky piece of hardware no one is going to buy it exccept someone in Korea at a
hack shop.

Volari is a dead brand. Plain and simple. They need to get a New Brand name and get something done thats worth while and without 100 IQ hacks just to make the thing run at 1/2 the speed of the closest copetitor.
 
If people decisions on their budget cards is made by $500 class performance, then human kind is doomed (together with xgi).
Im not so pesimistic in this, maybe if someone will come up with really solid cheap card from beginning, it would reverse the whole effect to something like: "look at those stupid big two playing race with too expensive cards which then cut in half and half gain, but here is one special budget solution fiting your needs better and for less".
 
:oops:

I dont know what that "RUMA architechture" is supposed to mean but I think it sums up whole xgi pretty well...

considering RUMA = UGLY in finnish!
 
Fodder said:
Hellbinder said:
Their Business model will *Never* Work.
It was working quite well for PowerVR/STMicro in the Kyro days.
Is that how they got to be such big players today? ;)

BTW Mendel- "RUMA = UGLY in finnish"
rofl.gif
rofl.gif
rofl.gif


Yeah, that fits! :LOL:
 
Putas said:
rwolf said:
Personally I never liked XGI attitude. We are going to beat them because we are going to use cheap labour.

What is wrong with that?

Because it drives down the standard of living for everyone else. Why not pull asians up to the same standard of living as the western world?
 
emphasis on cheap labour, the only way these people are going to make any money is we buy free trade graphics cards.
 
rwolf said:
Because it drives down the standard of living for everyone else. Why not pull asians up to the same standard of living as the western world?

in order to have a good standard of living we need them to have an inferior standard of living.
 
Fodder said:
It was working quite well for PowerVR/STMicro in the Kyro days.
You're joking, right?

Kyro (and the few previous PowerVR generations) were never anything more than extreme fringe market players, a flyspeck on the market.
 
Fodder said:
Hellbinder said:
Their Business model will *Never* Work.
It was working quite well for PowerVR/STMicro in the Kyro days.

No actually, it wasnt. If it was ST microelectronics would not have cut that leg of their business. They only sold a small ammount world wide. They did have a good thing going with SEGA though.

Besides whioch the PowerVR "thing" has a built in fanbase and "vibe" going for it becuase of its unique approach and its one of the granddaddys.

Volari/XGI has none of that going for them.
 
Putas said:
If people decisions on their budget cards is made by $500 class performance, then human kind is doomed (together with xgi).
Im not so pesimistic in this, maybe if someone will come up with really solid cheap card from beginning, it would reverse the whole effect to something like: "look at those stupid big two playing race with too expensive cards which then cut in half and half gain, but here is one special budget solution fiting your needs better and for less".

This is how its done unless you get a MOBO with a Built in Graphics chip ala Intel, S3 Uni Chrome, Geforce MX etc...

People *do* make buying decisions based off 500$ Classed Performance. They do it because thats how the brand name and reputation are established throughout the marketplace. Then the Clerk at the store or the name recognition does the rest. Its all trickledown sales power. The Big guns generate interest in the little guns because the big Guns generate the press. Then the press and the infrastructure does the rest as long as the products are competative.
 
"The halo effect". A nice succinct easily understandable label.

Not that it can't be overcome, but doing so also requires assets these companies usually don't have.
 
Considering the difference in sheer size between them and the big guys, I think this is the best that they can do ATM.They could have any business model, they don`t have the R&D potential to develop a high-end chip IMO.If this line goes OK and they actually become profitable, until Longhorn and WGF they could build up on the hypothetical succes of this gen with a refresh and in the meantime work on a fully featured WGF2.0 architecture that could be near the big-boys.Then again, they can suck badly and die, but that`s a worst case scenario IMHO
 
Taipei, Taiwan, April 11th, 2005: XGI Technology, Inc. today releases its source code for Linux. To support the large and growing Linux community, XGI Technology released the source code to X.Org, XFree86.Org, and Kernal.Org for its desktop, notebook, and server products.

"The Linux community is growing strong in today's technology market. To incorporate the newest Linux development to coinciding with XGI's graphics innovations, we have decided to open our source code for Linux," said Mr. Derek LinXGI's software VP. "Our goal is to support all Linux developers so that our graphics processors can run on nearly all Linux platforms; X.Org, XFree86.Org, and Kernal.Org are the vehicles to achieve this goal."


XGI supports X.Org 6.8.2nd XFree86 4.4.0. The open source supports package installation for IA32, AMD64 and EM64T, and Frame buffer device (FBDev) supportfor kernel version 2.4.X.
 
Back
Top